diff --git "a/checkpoint-6000/trainer_state.json" "b/checkpoint-6000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6000/trainer_state.json"
@@ -0,0 +1,42033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2770417664272413,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00021282255919127428,
+      "grad_norm": 5.140412253934901,
+      "learning_rate": 7.092198581560284e-08,
+      "loss": 11.0108,
+      "step": 1
+    },
+    {
+      "epoch": 0.00042564511838254855,
+      "grad_norm": 5.157597725266925,
+      "learning_rate": 1.4184397163120568e-07,
+      "loss": 11.0076,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006384676775738228,
+      "grad_norm": 5.35230126282656,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 11.0076,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008512902367650971,
+      "grad_norm": 5.244705527161708,
+      "learning_rate": 2.8368794326241136e-07,
+      "loss": 11.0066,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010641127959563714,
+      "grad_norm": 5.046074166444064,
+      "learning_rate": 3.546099290780142e-07,
+      "loss": 11.006,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012769353551476455,
+      "grad_norm": 5.335750046320456,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 11.0092,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014897579143389199,
+      "grad_norm": 5.237629005731343,
+      "learning_rate": 4.964539007092199e-07,
+      "loss": 11.0077,
+      "step": 7
+    },
+    {
+      "epoch": 0.0017025804735301942,
+      "grad_norm": 4.878807761693661,
+      "learning_rate": 5.673758865248227e-07,
+      "loss": 11.0065,
+      "step": 8
+    },
+    {
+      "epoch": 0.0019154030327214685,
+      "grad_norm": 5.516115331357534,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 11.0094,
+      "step": 9
+    },
+    {
+      "epoch": 0.002128225591912743,
+      "grad_norm": 5.1289953849170855,
+      "learning_rate": 7.092198581560284e-07,
+      "loss": 11.0076,
+      "step": 10
+    },
+    {
+      "epoch": 0.002341048151104017,
+      "grad_norm": 5.580830058715931,
+      "learning_rate": 7.801418439716313e-07,
+      "loss": 11.0075,
+      "step": 11
+    },
+    {
+      "epoch": 0.002553870710295291,
+      "grad_norm": 5.359132454829497,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 11.0012,
+      "step": 12
+    },
+    {
+      "epoch": 0.0027666932694865656,
+      "grad_norm": 5.901197161699977,
+      "learning_rate": 9.219858156028369e-07,
+      "loss": 11.002,
+      "step": 13
+    },
+    {
+      "epoch": 0.0029795158286778397,
+      "grad_norm": 5.424402150323474,
+      "learning_rate": 9.929078014184399e-07,
+      "loss": 10.9987,
+      "step": 14
+    },
+    {
+      "epoch": 0.0031923383878691143,
+      "grad_norm": 5.251558136315041,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 11.0003,
+      "step": 15
+    },
+    {
+      "epoch": 0.0034051609470603884,
+      "grad_norm": 5.6753276171778095,
+      "learning_rate": 1.1347517730496454e-06,
+      "loss": 11.0005,
+      "step": 16
+    },
+    {
+      "epoch": 0.0036179835062516625,
+      "grad_norm": 4.894432170293718,
+      "learning_rate": 1.2056737588652482e-06,
+      "loss": 10.9769,
+      "step": 17
+    },
+    {
+      "epoch": 0.003830806065442937,
+      "grad_norm": 4.996529267148075,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 10.9758,
+      "step": 18
+    },
+    {
+      "epoch": 0.004043628624634211,
+      "grad_norm": 5.863808156657381,
+      "learning_rate": 1.347517730496454e-06,
+      "loss": 10.9729,
+      "step": 19
+    },
+    {
+      "epoch": 0.004256451183825486,
+      "grad_norm": 5.627562290676342,
+      "learning_rate": 1.4184397163120568e-06,
+      "loss": 10.9681,
+      "step": 20
+    },
+    {
+      "epoch": 0.004469273743016759,
+      "grad_norm": 5.165366839102104,
+      "learning_rate": 1.4893617021276598e-06,
+      "loss": 10.9668,
+      "step": 21
+    },
+    {
+      "epoch": 0.004682096302208034,
+      "grad_norm": 5.007773725439582,
+      "learning_rate": 1.5602836879432626e-06,
+      "loss": 10.9671,
+      "step": 22
+    },
+    {
+      "epoch": 0.0048949188613993085,
+      "grad_norm": 5.23160351496433,
+      "learning_rate": 1.6312056737588652e-06,
+      "loss": 10.8924,
+      "step": 23
+    },
+    {
+      "epoch": 0.005107741420590582,
+      "grad_norm": 5.69076565825501,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 10.8859,
+      "step": 24
+    },
+    {
+      "epoch": 0.005320563979781857,
+      "grad_norm": 4.984989282347749,
+      "learning_rate": 1.7730496453900712e-06,
+      "loss": 10.8889,
+      "step": 25
+    },
+    {
+      "epoch": 0.005533386538973131,
+      "grad_norm": 4.8900764676672255,
+      "learning_rate": 1.8439716312056737e-06,
+      "loss": 10.8877,
+      "step": 26
+    },
+    {
+      "epoch": 0.005746209098164406,
+      "grad_norm": 5.119422554509663,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 10.8759,
+      "step": 27
+    },
+    {
+      "epoch": 0.0059590316573556795,
+      "grad_norm": 5.328533621660562,
+      "learning_rate": 1.9858156028368797e-06,
+      "loss": 10.8585,
+      "step": 28
+    },
+    {
+      "epoch": 0.006171854216546954,
+      "grad_norm": 4.789994198398988,
+      "learning_rate": 2.0567375886524823e-06,
+      "loss": 10.8642,
+      "step": 29
+    },
+    {
+      "epoch": 0.006384676775738229,
+      "grad_norm": 5.322140282867049,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 10.8463,
+      "step": 30
+    },
+    {
+      "epoch": 0.006597499334929502,
+      "grad_norm": 4.878524757680411,
+      "learning_rate": 2.1985815602836883e-06,
+      "loss": 10.6805,
+      "step": 31
+    },
+    {
+      "epoch": 0.006810321894120777,
+      "grad_norm": 3.994653843389885,
+      "learning_rate": 2.269503546099291e-06,
+      "loss": 10.64,
+      "step": 32
+    },
+    {
+      "epoch": 0.007023144453312051,
+      "grad_norm": 4.084409595371099,
+      "learning_rate": 2.3404255319148935e-06,
+      "loss": 10.618,
+      "step": 33
+    },
+    {
+      "epoch": 0.007235967012503325,
+      "grad_norm": 3.739966263593851,
+      "learning_rate": 2.4113475177304965e-06,
+      "loss": 10.6442,
+      "step": 34
+    },
+    {
+      "epoch": 0.0074487895716946,
+      "grad_norm": 4.04512335327713,
+      "learning_rate": 2.4822695035460995e-06,
+      "loss": 10.5989,
+      "step": 35
+    },
+    {
+      "epoch": 0.007661612130885874,
+      "grad_norm": 3.763220757700057,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 10.617,
+      "step": 36
+    },
+    {
+      "epoch": 0.007874434690077148,
+      "grad_norm": 3.683474027596859,
+      "learning_rate": 2.624113475177305e-06,
+      "loss": 10.5965,
+      "step": 37
+    },
+    {
+      "epoch": 0.008087257249268422,
+      "grad_norm": 3.6736028245269323,
+      "learning_rate": 2.695035460992908e-06,
+      "loss": 10.5735,
+      "step": 38
+    },
+    {
+      "epoch": 0.008300079808459697,
+      "grad_norm": 3.9583691064503395,
+      "learning_rate": 2.7659574468085106e-06,
+      "loss": 10.5151,
+      "step": 39
+    },
+    {
+      "epoch": 0.008512902367650971,
+      "grad_norm": 3.837266654676182,
+      "learning_rate": 2.8368794326241136e-06,
+      "loss": 10.4909,
+      "step": 40
+    },
+    {
+      "epoch": 0.008725724926842246,
+      "grad_norm": 3.7134425342939053,
+      "learning_rate": 2.9078014184397166e-06,
+      "loss": 10.4398,
+      "step": 41
+    },
+    {
+      "epoch": 0.008938547486033519,
+      "grad_norm": 3.632661459189256,
+      "learning_rate": 2.9787234042553196e-06,
+      "loss": 10.3376,
+      "step": 42
+    },
+    {
+      "epoch": 0.009151370045224793,
+      "grad_norm": 3.415743415638429,
+      "learning_rate": 3.049645390070922e-06,
+      "loss": 10.2726,
+      "step": 43
+    },
+    {
+      "epoch": 0.009364192604416068,
+      "grad_norm": 3.38220052890188,
+      "learning_rate": 3.120567375886525e-06,
+      "loss": 10.1824,
+      "step": 44
+    },
+    {
+      "epoch": 0.009577015163607342,
+      "grad_norm": 3.281845166838853,
+      "learning_rate": 3.1914893617021277e-06,
+      "loss": 10.1593,
+      "step": 45
+    },
+    {
+      "epoch": 0.009789837722798617,
+      "grad_norm": 3.1961272383133688,
+      "learning_rate": 3.2624113475177303e-06,
+      "loss": 10.1396,
+      "step": 46
+    },
+    {
+      "epoch": 0.010002660281989892,
+      "grad_norm": 3.1322155733562367,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 10.1077,
+      "step": 47
+    },
+    {
+      "epoch": 0.010215482841181164,
+      "grad_norm": 3.05229277817991,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 10.0817,
+      "step": 48
+    },
+    {
+      "epoch": 0.010428305400372439,
+      "grad_norm": 2.985964541284016,
+      "learning_rate": 3.4751773049645393e-06,
+      "loss": 10.0512,
+      "step": 49
+    },
+    {
+      "epoch": 0.010641127959563713,
+      "grad_norm": 2.8757021964648906,
+      "learning_rate": 3.5460992907801423e-06,
+      "loss": 10.0193,
+      "step": 50
+    },
+    {
+      "epoch": 0.010853950518754988,
+      "grad_norm": 2.8013526295393927,
+      "learning_rate": 3.6170212765957445e-06,
+      "loss": 9.9775,
+      "step": 51
+    },
+    {
+      "epoch": 0.011066773077946263,
+      "grad_norm": 2.798406483870758,
+      "learning_rate": 3.6879432624113475e-06,
+      "loss": 9.9943,
+      "step": 52
+    },
+    {
+      "epoch": 0.011279595637137537,
+      "grad_norm": 2.724951005355275,
+      "learning_rate": 3.7588652482269505e-06,
+      "loss": 9.9251,
+      "step": 53
+    },
+    {
+      "epoch": 0.011492418196328812,
+      "grad_norm": 2.660993827751493,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 9.864,
+      "step": 54
+    },
+    {
+      "epoch": 0.011705240755520084,
+      "grad_norm": 2.5691658689613046,
+      "learning_rate": 3.9007092198581565e-06,
+      "loss": 9.8447,
+      "step": 55
+    },
+    {
+      "epoch": 0.011918063314711359,
+      "grad_norm": 2.5074595309162544,
+      "learning_rate": 3.9716312056737595e-06,
+      "loss": 9.8079,
+      "step": 56
+    },
+    {
+      "epoch": 0.012130885873902634,
+      "grad_norm": 2.4787093421395396,
+      "learning_rate": 4.0425531914893625e-06,
+      "loss": 9.8143,
+      "step": 57
+    },
+    {
+      "epoch": 0.012343708433093908,
+      "grad_norm": 2.395837764269464,
+      "learning_rate": 4.113475177304965e-06,
+      "loss": 9.7722,
+      "step": 58
+    },
+    {
+      "epoch": 0.012556530992285183,
+      "grad_norm": 2.32816295645428,
+      "learning_rate": 4.184397163120568e-06,
+      "loss": 9.7642,
+      "step": 59
+    },
+    {
+      "epoch": 0.012769353551476457,
+      "grad_norm": 2.3035986433737015,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 9.7097,
+      "step": 60
+    },
+    {
+      "epoch": 0.01298217611066773,
+      "grad_norm": 2.2212541529584056,
+      "learning_rate": 4.326241134751774e-06,
+      "loss": 9.6321,
+      "step": 61
+    },
+    {
+      "epoch": 0.013194998669859005,
+      "grad_norm": 2.1842239112940747,
+      "learning_rate": 4.397163120567377e-06,
+      "loss": 9.6229,
+      "step": 62
+    },
+    {
+      "epoch": 0.013407821229050279,
+      "grad_norm": 2.1169432708023166,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 9.5978,
+      "step": 63
+    },
+    {
+      "epoch": 0.013620643788241554,
+      "grad_norm": 2.1245233467642892,
+      "learning_rate": 4.539007092198582e-06,
+      "loss": 9.5763,
+      "step": 64
+    },
+    {
+      "epoch": 0.013833466347432828,
+      "grad_norm": 2.1389592114304796,
+      "learning_rate": 4.609929078014184e-06,
+      "loss": 9.5107,
+      "step": 65
+    },
+    {
+      "epoch": 0.014046288906624103,
+      "grad_norm": 2.1055727257614767,
+      "learning_rate": 4.680851063829787e-06,
+      "loss": 9.5152,
+      "step": 66
+    },
+    {
+      "epoch": 0.014259111465815377,
+      "grad_norm": 2.10852581868414,
+      "learning_rate": 4.75177304964539e-06,
+      "loss": 9.4513,
+      "step": 67
+    },
+    {
+      "epoch": 0.01447193402500665,
+      "grad_norm": 2.065636522367971,
+      "learning_rate": 4.822695035460993e-06,
+      "loss": 9.4653,
+      "step": 68
+    },
+    {
+      "epoch": 0.014684756584197925,
+      "grad_norm": 2.008640659431026,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 9.4381,
+      "step": 69
+    },
+    {
+      "epoch": 0.0148975791433892,
+      "grad_norm": 1.9247132455800986,
+      "learning_rate": 4.964539007092199e-06,
+      "loss": 9.4408,
+      "step": 70
+    },
+    {
+      "epoch": 0.015110401702580474,
+      "grad_norm": 1.914205492489419,
+      "learning_rate": 5.035460992907802e-06,
+      "loss": 9.3993,
+      "step": 71
+    },
+    {
+      "epoch": 0.015323224261771748,
+      "grad_norm": 1.8821089753333557,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 9.3807,
+      "step": 72
+    },
+    {
+      "epoch": 0.015536046820963023,
+      "grad_norm": 1.878933338464255,
+      "learning_rate": 5.177304964539007e-06,
+      "loss": 9.3763,
+      "step": 73
+    },
+    {
+      "epoch": 0.015748869380154296,
+      "grad_norm": 1.844016304429905,
+      "learning_rate": 5.24822695035461e-06,
+      "loss": 9.3847,
+      "step": 74
+    },
+    {
+      "epoch": 0.01596169193934557,
+      "grad_norm": 1.8661800798569947,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 9.344,
+      "step": 75
+    },
+    {
+      "epoch": 0.016174514498536845,
+      "grad_norm": 1.8239076915367878,
+      "learning_rate": 5.390070921985816e-06,
+      "loss": 9.3152,
+      "step": 76
+    },
+    {
+      "epoch": 0.01638733705772812,
+      "grad_norm": 1.827017208385593,
+      "learning_rate": 5.460992907801419e-06,
+      "loss": 9.3211,
+      "step": 77
+    },
+    {
+      "epoch": 0.016600159616919394,
+      "grad_norm": 1.8262460118479378,
+      "learning_rate": 5.531914893617021e-06,
+      "loss": 9.2418,
+      "step": 78
+    },
+    {
+      "epoch": 0.01681298217611067,
+      "grad_norm": 1.7943394110365032,
+      "learning_rate": 5.602836879432624e-06,
+      "loss": 9.2441,
+      "step": 79
+    },
+    {
+      "epoch": 0.017025804735301943,
+      "grad_norm": 1.8265125332123977,
+      "learning_rate": 5.673758865248227e-06,
+      "loss": 9.263,
+      "step": 80
+    },
+    {
+      "epoch": 0.017238627294493217,
+      "grad_norm": 1.8362212621632064,
+      "learning_rate": 5.74468085106383e-06,
+      "loss": 9.2276,
+      "step": 81
+    },
+    {
+      "epoch": 0.017451449853684492,
+      "grad_norm": 1.8502519342703758,
+      "learning_rate": 5.815602836879433e-06,
+      "loss": 9.2092,
+      "step": 82
+    },
+    {
+      "epoch": 0.017664272412875767,
+      "grad_norm": 1.8129903931941647,
+      "learning_rate": 5.886524822695036e-06,
+      "loss": 9.1616,
+      "step": 83
+    },
+    {
+      "epoch": 0.017877094972067038,
+      "grad_norm": 1.7757102246424927,
+      "learning_rate": 5.957446808510639e-06,
+      "loss": 9.2054,
+      "step": 84
+    },
+    {
+      "epoch": 0.018089917531258312,
+      "grad_norm": 1.8045393007443733,
+      "learning_rate": 6.028368794326241e-06,
+      "loss": 9.0991,
+      "step": 85
+    },
+    {
+      "epoch": 0.018302740090449587,
+      "grad_norm": 1.733910315364888,
+      "learning_rate": 6.099290780141844e-06,
+      "loss": 9.115,
+      "step": 86
+    },
+    {
+      "epoch": 0.01851556264964086,
+      "grad_norm": 1.7330820592615432,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 9.1967,
+      "step": 87
+    },
+    {
+      "epoch": 0.018728385208832136,
+      "grad_norm": 1.7134665928564397,
+      "learning_rate": 6.24113475177305e-06,
+      "loss": 9.1427,
+      "step": 88
+    },
+    {
+      "epoch": 0.01894120776802341,
+      "grad_norm": 1.7073983238574284,
+      "learning_rate": 6.3120567375886525e-06,
+      "loss": 9.047,
+      "step": 89
+    },
+    {
+      "epoch": 0.019154030327214685,
+      "grad_norm": 1.6989953628193428,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 9.1066,
+      "step": 90
+    },
+    {
+      "epoch": 0.01936685288640596,
+      "grad_norm": 1.735017631934819,
+      "learning_rate": 6.453900709219858e-06,
+      "loss": 9.0485,
+      "step": 91
+    },
+    {
+      "epoch": 0.019579675445597234,
+      "grad_norm": 1.7053356200992407,
+      "learning_rate": 6.524822695035461e-06,
+      "loss": 9.0803,
+      "step": 92
+    },
+    {
+      "epoch": 0.01979249800478851,
+      "grad_norm": 1.7149345281973412,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 9.0985,
+      "step": 93
+    },
+    {
+      "epoch": 0.020005320563979783,
+      "grad_norm": 1.7649915623682806,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 8.9858,
+      "step": 94
+    },
+    {
+      "epoch": 0.020218143123171058,
+      "grad_norm": 1.7050396888974504,
+      "learning_rate": 6.73758865248227e-06,
+      "loss": 9.0442,
+      "step": 95
+    },
+    {
+      "epoch": 0.02043096568236233,
+      "grad_norm": 1.7165570743402758,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 9.0176,
+      "step": 96
+    },
+    {
+      "epoch": 0.020643788241553603,
+      "grad_norm": 1.6914594725157326,
+      "learning_rate": 6.879432624113476e-06,
+      "loss": 9.0056,
+      "step": 97
+    },
+    {
+      "epoch": 0.020856610800744878,
+      "grad_norm": 1.6893031616506453,
+      "learning_rate": 6.950354609929079e-06,
+      "loss": 8.9851,
+      "step": 98
+    },
+    {
+      "epoch": 0.021069433359936152,
+      "grad_norm": 1.6623619224515362,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 8.9676,
+      "step": 99
+    },
+    {
+      "epoch": 0.021282255919127427,
+      "grad_norm": 1.6862398154912932,
+      "learning_rate": 7.092198581560285e-06,
+      "loss": 8.9871,
+      "step": 100
+    },
+    {
+      "epoch": 0.0214950784783187,
+      "grad_norm": 1.682519580458688,
+      "learning_rate": 7.163120567375888e-06,
+      "loss": 8.9527,
+      "step": 101
+    },
+    {
+      "epoch": 0.021707901037509976,
+      "grad_norm": 1.6820885201225904,
+      "learning_rate": 7.234042553191489e-06,
+      "loss": 8.9118,
+      "step": 102
+    },
+    {
+      "epoch": 0.02192072359670125,
+      "grad_norm": 1.6643111797231578,
+      "learning_rate": 7.304964539007092e-06,
+      "loss": 8.937,
+      "step": 103
+    },
+    {
+      "epoch": 0.022133546155892525,
+      "grad_norm": 1.6600183899977525,
+      "learning_rate": 7.375886524822695e-06,
+      "loss": 8.8916,
+      "step": 104
+    },
+    {
+      "epoch": 0.0223463687150838,
+      "grad_norm": 1.6533527105070631,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 8.8745,
+      "step": 105
+    },
+    {
+      "epoch": 0.022559191274275074,
+      "grad_norm": 1.6146593625977317,
+      "learning_rate": 7.517730496453901e-06,
+      "loss": 8.8618,
+      "step": 106
+    },
+    {
+      "epoch": 0.02277201383346635,
+      "grad_norm": 1.6367604147843513,
+      "learning_rate": 7.588652482269504e-06,
+      "loss": 8.8552,
+      "step": 107
+    },
+    {
+      "epoch": 0.022984836392657623,
+      "grad_norm": 1.6188967197207906,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 8.845,
+      "step": 108
+    },
+    {
+      "epoch": 0.023197658951848894,
+      "grad_norm": 1.6507762880218184,
+      "learning_rate": 7.73049645390071e-06,
+      "loss": 8.8165,
+      "step": 109
+    },
+    {
+      "epoch": 0.02341048151104017,
+      "grad_norm": 1.6075882363558114,
+      "learning_rate": 7.801418439716313e-06,
+      "loss": 8.7727,
+      "step": 110
+    },
+    {
+      "epoch": 0.023623304070231443,
+      "grad_norm": 1.6022856615023926,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 8.7728,
+      "step": 111
+    },
+    {
+      "epoch": 0.023836126629422718,
+      "grad_norm": 1.5802848326070125,
+      "learning_rate": 7.943262411347519e-06,
+      "loss": 8.7732,
+      "step": 112
+    },
+    {
+      "epoch": 0.024048949188613993,
+      "grad_norm": 1.5554629915743439,
+      "learning_rate": 8.014184397163122e-06,
+      "loss": 8.7755,
+      "step": 113
+    },
+    {
+      "epoch": 0.024261771747805267,
+      "grad_norm": 1.5579402143056102,
+      "learning_rate": 8.085106382978725e-06,
+      "loss": 8.7194,
+      "step": 114
+    },
+    {
+      "epoch": 0.02447459430699654,
+      "grad_norm": 1.5754204874263333,
+      "learning_rate": 8.156028368794328e-06,
+      "loss": 8.7197,
+      "step": 115
+    },
+    {
+      "epoch": 0.024687416866187816,
+      "grad_norm": 1.5667237397512201,
+      "learning_rate": 8.22695035460993e-06,
+      "loss": 8.7722,
+      "step": 116
+    },
+    {
+      "epoch": 0.02490023942537909,
+      "grad_norm": 1.5230352621452128,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 8.7351,
+      "step": 117
+    },
+    {
+      "epoch": 0.025113061984570365,
+      "grad_norm": 1.5323159813249516,
+      "learning_rate": 8.368794326241135e-06,
+      "loss": 8.7206,
+      "step": 118
+    },
+    {
+      "epoch": 0.02532588454376164,
+      "grad_norm": 1.511112466876129,
+      "learning_rate": 8.439716312056738e-06,
+      "loss": 8.6591,
+      "step": 119
+    },
+    {
+      "epoch": 0.025538707102952914,
+      "grad_norm": 1.536949692312784,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 8.7061,
+      "step": 120
+    },
+    {
+      "epoch": 0.02575152966214419,
+      "grad_norm": 1.5238118625693244,
+      "learning_rate": 8.581560283687944e-06,
+      "loss": 8.6661,
+      "step": 121
+    },
+    {
+      "epoch": 0.02596435222133546,
+      "grad_norm": 1.5136382130337027,
+      "learning_rate": 8.652482269503547e-06,
+      "loss": 8.6764,
+      "step": 122
+    },
+    {
+      "epoch": 0.026177174780526734,
+      "grad_norm": 1.5073473638724457,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 8.6216,
+      "step": 123
+    },
+    {
+      "epoch": 0.02638999733971801,
+      "grad_norm": 1.4717618115344013,
+      "learning_rate": 8.794326241134753e-06,
+      "loss": 8.6378,
+      "step": 124
+    },
+    {
+      "epoch": 0.026602819898909284,
+      "grad_norm": 1.5026013905251554,
+      "learning_rate": 8.865248226950355e-06,
+      "loss": 8.5937,
+      "step": 125
+    },
+    {
+      "epoch": 0.026815642458100558,
+      "grad_norm": 1.4795654402452285,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 8.6282,
+      "step": 126
+    },
+    {
+      "epoch": 0.027028465017291833,
+      "grad_norm": 1.4682025195647819,
+      "learning_rate": 9.00709219858156e-06,
+      "loss": 8.5683,
+      "step": 127
+    },
+    {
+      "epoch": 0.027241287576483107,
+      "grad_norm": 1.4883895695119789,
+      "learning_rate": 9.078014184397164e-06,
+      "loss": 8.6228,
+      "step": 128
+    },
+    {
+      "epoch": 0.027454110135674382,
+      "grad_norm": 1.4594270307270611,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 8.6029,
+      "step": 129
+    },
+    {
+      "epoch": 0.027666932694865656,
+      "grad_norm": 1.4730469568331355,
+      "learning_rate": 9.219858156028368e-06,
+      "loss": 8.549,
+      "step": 130
+    },
+    {
+      "epoch": 0.02787975525405693,
+      "grad_norm": 1.4649838817501863,
+      "learning_rate": 9.29078014184397e-06,
+      "loss": 8.5825,
+      "step": 131
+    },
+    {
+      "epoch": 0.028092577813248205,
+      "grad_norm": 1.4678581779917808,
+      "learning_rate": 9.361702127659574e-06,
+      "loss": 8.5739,
+      "step": 132
+    },
+    {
+      "epoch": 0.02830540037243948,
+      "grad_norm": 1.466879951910906,
+      "learning_rate": 9.432624113475177e-06,
+      "loss": 8.4957,
+      "step": 133
+    },
+    {
+      "epoch": 0.028518222931630755,
+      "grad_norm": 1.4382688029704138,
+      "learning_rate": 9.50354609929078e-06,
+      "loss": 8.5106,
+      "step": 134
+    },
+    {
+      "epoch": 0.028731045490822026,
+      "grad_norm": 1.444932548605664,
+      "learning_rate": 9.574468085106383e-06,
+      "loss": 8.4801,
+      "step": 135
+    },
+    {
+      "epoch": 0.0289438680500133,
+      "grad_norm": 1.4337827761040187,
+      "learning_rate": 9.645390070921986e-06,
+      "loss": 8.5241,
+      "step": 136
+    },
+    {
+      "epoch": 0.029156690609204575,
+      "grad_norm": 1.414181424714178,
+      "learning_rate": 9.716312056737589e-06,
+      "loss": 8.446,
+      "step": 137
+    },
+    {
+      "epoch": 0.02936951316839585,
+      "grad_norm": 1.4248471545730776,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 8.412,
+      "step": 138
+    },
+    {
+      "epoch": 0.029582335727587124,
+      "grad_norm": 1.4140685230784498,
+      "learning_rate": 9.858156028368795e-06,
+      "loss": 8.4284,
+      "step": 139
+    },
+    {
+      "epoch": 0.0297951582867784,
+      "grad_norm": 1.4111917914863266,
+      "learning_rate": 9.929078014184398e-06,
+      "loss": 8.4252,
+      "step": 140
+    },
+    {
+      "epoch": 0.030007980845969673,
+      "grad_norm": 1.375879728261706,
+      "learning_rate": 1e-05,
+      "loss": 8.3982,
+      "step": 141
+    },
+    {
+      "epoch": 0.030220803405160947,
+      "grad_norm": 1.3868004736544457,
+      "learning_rate": 1.0070921985815604e-05,
+      "loss": 8.418,
+      "step": 142
+    },
+    {
+      "epoch": 0.030433625964352222,
+      "grad_norm": 1.3838702035329888,
+      "learning_rate": 1.0141843971631205e-05,
+      "loss": 8.4469,
+      "step": 143
+    },
+    {
+      "epoch": 0.030646448523543496,
+      "grad_norm": 1.3896670719791104,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 8.3934,
+      "step": 144
+    },
+    {
+      "epoch": 0.03085927108273477,
+      "grad_norm": 1.3668956393785472,
+      "learning_rate": 1.0283687943262411e-05,
+      "loss": 8.3627,
+      "step": 145
+    },
+    {
+      "epoch": 0.031072093641926046,
+      "grad_norm": 1.3610532173304597,
+      "learning_rate": 1.0354609929078014e-05,
+      "loss": 8.3524,
+      "step": 146
+    },
+    {
+      "epoch": 0.03128491620111732,
+      "grad_norm": 1.3514943225907783,
+      "learning_rate": 1.0425531914893617e-05,
+      "loss": 8.3614,
+      "step": 147
+    },
+    {
+      "epoch": 0.03149773876030859,
+      "grad_norm": 1.3443253875208467,
+      "learning_rate": 1.049645390070922e-05,
+      "loss": 8.3238,
+      "step": 148
+    },
+    {
+      "epoch": 0.03171056131949987,
+      "grad_norm": 1.3692420818766269,
+      "learning_rate": 1.0567375886524823e-05,
+      "loss": 8.3205,
+      "step": 149
+    },
+    {
+      "epoch": 0.03192338387869114,
+      "grad_norm": 1.3259399241483978,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 8.3315,
+      "step": 150
+    },
+    {
+      "epoch": 0.03213620643788242,
+      "grad_norm": 1.3235477375737166,
+      "learning_rate": 1.0709219858156029e-05,
+      "loss": 8.2629,
+      "step": 151
+    },
+    {
+      "epoch": 0.03234902899707369,
+      "grad_norm": 1.3071336712970918,
+      "learning_rate": 1.0780141843971632e-05,
+      "loss": 8.3011,
+      "step": 152
+    },
+    {
+      "epoch": 0.03256185155626497,
+      "grad_norm": 1.312152623056875,
+      "learning_rate": 1.0851063829787235e-05,
+      "loss": 8.2697,
+      "step": 153
+    },
+    {
+      "epoch": 0.03277467411545624,
+      "grad_norm": 1.2885337997004542,
+      "learning_rate": 1.0921985815602838e-05,
+      "loss": 8.2417,
+      "step": 154
+    },
+    {
+      "epoch": 0.03298749667464751,
+      "grad_norm": 1.2801756429300917,
+      "learning_rate": 1.0992907801418441e-05,
+      "loss": 8.2314,
+      "step": 155
+    },
+    {
+      "epoch": 0.03320031923383879,
+      "grad_norm": 1.2996004014468072,
+      "learning_rate": 1.1063829787234042e-05,
+      "loss": 8.2845,
+      "step": 156
+    },
+    {
+      "epoch": 0.03341314179303006,
+      "grad_norm": 1.29702982956657,
+      "learning_rate": 1.1134751773049645e-05,
+      "loss": 8.209,
+      "step": 157
+    },
+    {
+      "epoch": 0.03362596435222134,
+      "grad_norm": 1.2953987975999832,
+      "learning_rate": 1.1205673758865248e-05,
+      "loss": 8.2611,
+      "step": 158
+    },
+    {
+      "epoch": 0.03383878691141261,
+      "grad_norm": 1.2682617028805188,
+      "learning_rate": 1.1276595744680851e-05,
+      "loss": 8.1605,
+      "step": 159
+    },
+    {
+      "epoch": 0.034051609470603886,
+      "grad_norm": 1.2696539166450465,
+      "learning_rate": 1.1347517730496454e-05,
+      "loss": 8.2389,
+      "step": 160
+    },
+    {
+      "epoch": 0.03426443202979516,
+      "grad_norm": 1.2535663454368084,
+      "learning_rate": 1.1418439716312057e-05,
+      "loss": 8.1903,
+      "step": 161
+    },
+    {
+      "epoch": 0.034477254588986435,
+      "grad_norm": 1.2510738503404535,
+      "learning_rate": 1.148936170212766e-05,
+      "loss": 8.1523,
+      "step": 162
+    },
+    {
+      "epoch": 0.034690077148177706,
+      "grad_norm": 1.2791730744552423,
+      "learning_rate": 1.1560283687943263e-05,
+      "loss": 8.1721,
+      "step": 163
+    },
+    {
+      "epoch": 0.034902899707368984,
+      "grad_norm": 1.2360799908004023,
+      "learning_rate": 1.1631205673758866e-05,
+      "loss": 8.1484,
+      "step": 164
+    },
+    {
+      "epoch": 0.035115722266560255,
+      "grad_norm": 1.2388063344186366,
+      "learning_rate": 1.170212765957447e-05,
+      "loss": 8.1602,
+      "step": 165
+    },
+    {
+      "epoch": 0.03532854482575153,
+      "grad_norm": 1.2260063802009897,
+      "learning_rate": 1.1773049645390072e-05,
+      "loss": 8.1135,
+      "step": 166
+    },
+    {
+      "epoch": 0.035541367384942804,
+      "grad_norm": 1.2387138762822183,
+      "learning_rate": 1.1843971631205675e-05,
+      "loss": 8.0973,
+      "step": 167
+    },
+    {
+      "epoch": 0.035754189944134075,
+      "grad_norm": 1.2375961137870417,
+      "learning_rate": 1.1914893617021278e-05,
+      "loss": 8.1044,
+      "step": 168
+    },
+    {
+      "epoch": 0.03596701250332535,
+      "grad_norm": 1.2296002311922016,
+      "learning_rate": 1.1985815602836881e-05,
+      "loss": 8.0734,
+      "step": 169
+    },
+    {
+      "epoch": 0.036179835062516624,
+      "grad_norm": 1.218999785699375,
+      "learning_rate": 1.2056737588652483e-05,
+      "loss": 8.0691,
+      "step": 170
+    },
+    {
+      "epoch": 0.0363926576217079,
+      "grad_norm": 1.200795530219513,
+      "learning_rate": 1.2127659574468086e-05,
+      "loss": 8.0582,
+      "step": 171
+    },
+    {
+      "epoch": 0.03660548018089917,
+      "grad_norm": 1.2018938557179464,
+      "learning_rate": 1.2198581560283689e-05,
+      "loss": 7.9799,
+      "step": 172
+    },
+    {
+      "epoch": 0.03681830274009045,
+      "grad_norm": 1.2059230234817642,
+      "learning_rate": 1.2269503546099292e-05,
+      "loss": 8.0462,
+      "step": 173
+    },
+    {
+      "epoch": 0.03703112529928172,
+      "grad_norm": 1.203836725307771,
+      "learning_rate": 1.2340425531914895e-05,
+      "loss": 8.0504,
+      "step": 174
+    },
+    {
+      "epoch": 0.037243947858473,
+      "grad_norm": 1.188888859948644,
+      "learning_rate": 1.2411347517730498e-05,
+      "loss": 7.9974,
+      "step": 175
+    },
+    {
+      "epoch": 0.03745677041766427,
+      "grad_norm": 1.2012727063417539,
+      "learning_rate": 1.24822695035461e-05,
+      "loss": 8.0523,
+      "step": 176
+    },
+    {
+      "epoch": 0.03766959297685555,
+      "grad_norm": 1.1795360541734434,
+      "learning_rate": 1.2553191489361702e-05,
+      "loss": 7.9666,
+      "step": 177
+    },
+    {
+      "epoch": 0.03788241553604682,
+      "grad_norm": 1.1598263709389438,
+      "learning_rate": 1.2624113475177305e-05,
+      "loss": 7.9742,
+      "step": 178
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 1.197259335506373,
+      "learning_rate": 1.2695035460992908e-05,
+      "loss": 7.9882,
+      "step": 179
+    },
+    {
+      "epoch": 0.03830806065442937,
+      "grad_norm": 1.151697294984388,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 7.9576,
+      "step": 180
+    },
+    {
+      "epoch": 0.03852088321362064,
+      "grad_norm": 1.1592657518523581,
+      "learning_rate": 1.2836879432624114e-05,
+      "loss": 7.8888,
+      "step": 181
+    },
+    {
+      "epoch": 0.03873370577281192,
+      "grad_norm": 1.1515070824638616,
+      "learning_rate": 1.2907801418439715e-05,
+      "loss": 7.9112,
+      "step": 182
+    },
+    {
+      "epoch": 0.03894652833200319,
+      "grad_norm": 1.123179168358839,
+      "learning_rate": 1.2978723404255318e-05,
+      "loss": 7.9097,
+      "step": 183
+    },
+    {
+      "epoch": 0.03915935089119447,
+      "grad_norm": 1.1323386074380433,
+      "learning_rate": 1.3049645390070921e-05,
+      "loss": 7.87,
+      "step": 184
+    },
+    {
+      "epoch": 0.03937217345038574,
+      "grad_norm": 1.106364352396766,
+      "learning_rate": 1.3120567375886524e-05,
+      "loss": 7.8738,
+      "step": 185
+    },
+    {
+      "epoch": 0.03958499600957702,
+      "grad_norm": 1.1125356603334557,
+      "learning_rate": 1.3191489361702127e-05,
+      "loss": 7.827,
+      "step": 186
+    },
+    {
+      "epoch": 0.03979781856876829,
+      "grad_norm": 1.0873386773238216,
+      "learning_rate": 1.326241134751773e-05,
+      "loss": 7.8063,
+      "step": 187
+    },
+    {
+      "epoch": 0.040010641127959566,
+      "grad_norm": 1.092218415265519,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 7.8259,
+      "step": 188
+    },
+    {
+      "epoch": 0.04022346368715084,
+      "grad_norm": 1.0867716498041,
+      "learning_rate": 1.3404255319148936e-05,
+      "loss": 7.9025,
+      "step": 189
+    },
+    {
+      "epoch": 0.040436286246342115,
+      "grad_norm": 1.0785720037636046,
+      "learning_rate": 1.347517730496454e-05,
+      "loss": 7.8339,
+      "step": 190
+    },
+    {
+      "epoch": 0.040649108805533386,
+      "grad_norm": 1.0630176586011884,
+      "learning_rate": 1.3546099290780142e-05,
+      "loss": 7.7367,
+      "step": 191
+    },
+    {
+      "epoch": 0.04086193136472466,
+      "grad_norm": 1.0788652841832922,
+      "learning_rate": 1.3617021276595745e-05,
+      "loss": 7.8138,
+      "step": 192
+    },
+    {
+      "epoch": 0.041074753923915935,
+      "grad_norm": 1.0776611366695936,
+      "learning_rate": 1.3687943262411348e-05,
+      "loss": 7.7408,
+      "step": 193
+    },
+    {
+      "epoch": 0.041287576483107206,
+      "grad_norm": 1.044984878403835,
+      "learning_rate": 1.3758865248226951e-05,
+      "loss": 7.7872,
+      "step": 194
+    },
+    {
+      "epoch": 0.041500399042298484,
+      "grad_norm": 1.0369145052449924,
+      "learning_rate": 1.3829787234042554e-05,
+      "loss": 7.7722,
+      "step": 195
+    },
+    {
+      "epoch": 0.041713221601489756,
+      "grad_norm": 1.0907809821830945,
+      "learning_rate": 1.3900709219858157e-05,
+      "loss": 7.7535,
+      "step": 196
+    },
+    {
+      "epoch": 0.041926044160681034,
+      "grad_norm": 1.062723401214868,
+      "learning_rate": 1.397163120567376e-05,
+      "loss": 7.6995,
+      "step": 197
+    },
+    {
+      "epoch": 0.042138866719872305,
+      "grad_norm": 1.0248594602592236,
+      "learning_rate": 1.4042553191489363e-05,
+      "loss": 7.6871,
+      "step": 198
+    },
+    {
+      "epoch": 0.04235168927906358,
+      "grad_norm": 1.0177779200568362,
+      "learning_rate": 1.4113475177304966e-05,
+      "loss": 7.7036,
+      "step": 199
+    },
+    {
+      "epoch": 0.042564511838254854,
+      "grad_norm": 1.0134898084561892,
+      "learning_rate": 1.418439716312057e-05,
+      "loss": 7.7011,
+      "step": 200
+    },
+    {
+      "epoch": 0.04277733439744613,
+      "grad_norm": 1.0084465696767226,
+      "learning_rate": 1.4255319148936172e-05,
+      "loss": 7.7066,
+      "step": 201
+    },
+    {
+      "epoch": 0.0429901569566374,
+      "grad_norm": 0.9831592845564199,
+      "learning_rate": 1.4326241134751775e-05,
+      "loss": 7.6932,
+      "step": 202
+    },
+    {
+      "epoch": 0.04320297951582868,
+      "grad_norm": 0.9899847086973093,
+      "learning_rate": 1.4397163120567378e-05,
+      "loss": 7.6151,
+      "step": 203
+    },
+    {
+      "epoch": 0.04341580207501995,
+      "grad_norm": 0.9786872594197019,
+      "learning_rate": 1.4468085106382978e-05,
+      "loss": 7.6326,
+      "step": 204
+    },
+    {
+      "epoch": 0.04362862463421122,
+      "grad_norm": 0.9841234950765124,
+      "learning_rate": 1.4539007092198581e-05,
+      "loss": 7.6439,
+      "step": 205
+    },
+    {
+      "epoch": 0.0438414471934025,
+      "grad_norm": 0.9811230941831721,
+      "learning_rate": 1.4609929078014184e-05,
+      "loss": 7.6117,
+      "step": 206
+    },
+    {
+      "epoch": 0.04405426975259377,
+      "grad_norm": 0.9760876296375166,
+      "learning_rate": 1.4680851063829787e-05,
+      "loss": 7.6025,
+      "step": 207
+    },
+    {
+      "epoch": 0.04426709231178505,
+      "grad_norm": 1.002329526915382,
+      "learning_rate": 1.475177304964539e-05,
+      "loss": 7.5929,
+      "step": 208
+    },
+    {
+      "epoch": 0.04447991487097632,
+      "grad_norm": 1.113103060681469,
+      "learning_rate": 1.4822695035460993e-05,
+      "loss": 7.5592,
+      "step": 209
+    },
+    {
+      "epoch": 0.0446927374301676,
+      "grad_norm": 1.1460247229654037,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 7.5496,
+      "step": 210
+    },
+    {
+      "epoch": 0.04490555998935887,
+      "grad_norm": 1.009134107715759,
+      "learning_rate": 1.4964539007092199e-05,
+      "loss": 7.5447,
+      "step": 211
+    },
+    {
+      "epoch": 0.04511838254855015,
+      "grad_norm": 1.0027382846464932,
+      "learning_rate": 1.5035460992907802e-05,
+      "loss": 7.5697,
+      "step": 212
+    },
+    {
+      "epoch": 0.04533120510774142,
+      "grad_norm": 1.2250005956445211,
+      "learning_rate": 1.5106382978723405e-05,
+      "loss": 7.5259,
+      "step": 213
+    },
+    {
+      "epoch": 0.0455440276669327,
+      "grad_norm": 1.0786076272192184,
+      "learning_rate": 1.5177304964539008e-05,
+      "loss": 7.4952,
+      "step": 214
+    },
+    {
+      "epoch": 0.04575685022612397,
+      "grad_norm": 0.9115678253053698,
+      "learning_rate": 1.5248226950354611e-05,
+      "loss": 7.4628,
+      "step": 215
+    },
+    {
+      "epoch": 0.045969672785315246,
+      "grad_norm": 1.0150230603573522,
+      "learning_rate": 1.5319148936170214e-05,
+      "loss": 7.4584,
+      "step": 216
+    },
+    {
+      "epoch": 0.04618249534450652,
+      "grad_norm": 1.189619936625802,
+      "learning_rate": 1.5390070921985817e-05,
+      "loss": 7.436,
+      "step": 217
+    },
+    {
+      "epoch": 0.04639531790369779,
+      "grad_norm": 0.9603212874022666,
+      "learning_rate": 1.546099290780142e-05,
+      "loss": 7.4205,
+      "step": 218
+    },
+    {
+      "epoch": 0.04660814046288907,
+      "grad_norm": 0.9564286047797671,
+      "learning_rate": 1.5531914893617023e-05,
+      "loss": 7.4247,
+      "step": 219
+    },
+    {
+      "epoch": 0.04682096302208034,
+      "grad_norm": 1.1789189108894647,
+      "learning_rate": 1.5602836879432626e-05,
+      "loss": 7.4501,
+      "step": 220
+    },
+    {
+      "epoch": 0.047033785581271616,
+      "grad_norm": 0.9238442589642779,
+      "learning_rate": 1.567375886524823e-05,
+      "loss": 7.4157,
+      "step": 221
+    },
+    {
+      "epoch": 0.04724660814046289,
+      "grad_norm": 0.8880870395459604,
+      "learning_rate": 1.5744680851063832e-05,
+      "loss": 7.4193,
+      "step": 222
+    },
+    {
+      "epoch": 0.047459430699654165,
+      "grad_norm": 1.080644813699659,
+      "learning_rate": 1.5815602836879435e-05,
+      "loss": 7.3783,
+      "step": 223
+    },
+    {
+      "epoch": 0.047672253258845436,
+      "grad_norm": 0.9902158942369448,
+      "learning_rate": 1.5886524822695038e-05,
+      "loss": 7.3838,
+      "step": 224
+    },
+    {
+      "epoch": 0.047885075818036714,
+      "grad_norm": 0.8738535135547402,
+      "learning_rate": 1.595744680851064e-05,
+      "loss": 7.4029,
+      "step": 225
+    },
+    {
+      "epoch": 0.048097898377227985,
+      "grad_norm": 0.9325148202953346,
+      "learning_rate": 1.6028368794326244e-05,
+      "loss": 7.3866,
+      "step": 226
+    },
+    {
+      "epoch": 0.04831072093641926,
+      "grad_norm": 1.0055645778341522,
+      "learning_rate": 1.6099290780141847e-05,
+      "loss": 7.3328,
+      "step": 227
+    },
+    {
+      "epoch": 0.048523543495610534,
+      "grad_norm": 1.2936647365320946,
+      "learning_rate": 1.617021276595745e-05,
+      "loss": 7.3387,
+      "step": 228
+    },
+    {
+      "epoch": 0.04873636605480181,
+      "grad_norm": 1.0798239094614512,
+      "learning_rate": 1.6241134751773053e-05,
+      "loss": 7.3279,
+      "step": 229
+    },
+    {
+      "epoch": 0.04894918861399308,
+      "grad_norm": 1.1795921704513437,
+      "learning_rate": 1.6312056737588656e-05,
+      "loss": 7.3828,
+      "step": 230
+    },
+    {
+      "epoch": 0.049162011173184354,
+      "grad_norm": 1.1666381002572075,
+      "learning_rate": 1.6382978723404255e-05,
+      "loss": 7.2723,
+      "step": 231
+    },
+    {
+      "epoch": 0.04937483373237563,
+      "grad_norm": 0.8854222573088293,
+      "learning_rate": 1.645390070921986e-05,
+      "loss": 7.3173,
+      "step": 232
+    },
+    {
+      "epoch": 0.0495876562915669,
+      "grad_norm": 0.9644140219987087,
+      "learning_rate": 1.652482269503546e-05,
+      "loss": 7.2794,
+      "step": 233
+    },
+    {
+      "epoch": 0.04980047885075818,
+      "grad_norm": 1.4375072771287478,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 7.3039,
+      "step": 234
+    },
+    {
+      "epoch": 0.05001330140994945,
+      "grad_norm": 1.085911249729322,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 7.2764,
+      "step": 235
+    },
+    {
+      "epoch": 0.05022612396914073,
+      "grad_norm": 0.987481060718086,
+      "learning_rate": 1.673758865248227e-05,
+      "loss": 7.2122,
+      "step": 236
+    },
+    {
+      "epoch": 0.050438946528332,
+      "grad_norm": 0.9816123808242504,
+      "learning_rate": 1.6808510638297873e-05,
+      "loss": 7.2341,
+      "step": 237
+    },
+    {
+      "epoch": 0.05065176908752328,
+      "grad_norm": 0.9938582587840539,
+      "learning_rate": 1.6879432624113476e-05,
+      "loss": 7.2261,
+      "step": 238
+    },
+    {
+      "epoch": 0.05086459164671455,
+      "grad_norm": 0.9943938530332892,
+      "learning_rate": 1.695035460992908e-05,
+      "loss": 7.2339,
+      "step": 239
+    },
+    {
+      "epoch": 0.05107741420590583,
+      "grad_norm": 0.975743290380457,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 7.2053,
+      "step": 240
+    },
+    {
+      "epoch": 0.0512902367650971,
+      "grad_norm": 1.4418553709725543,
+      "learning_rate": 1.7092198581560285e-05,
+      "loss": 7.2044,
+      "step": 241
+    },
+    {
+      "epoch": 0.05150305932428838,
+      "grad_norm": 1.2508769053157591,
+      "learning_rate": 1.716312056737589e-05,
+      "loss": 7.1497,
+      "step": 242
+    },
+    {
+      "epoch": 0.05171588188347965,
+      "grad_norm": 0.8519511434925519,
+      "learning_rate": 1.723404255319149e-05,
+      "loss": 7.1799,
+      "step": 243
+    },
+    {
+      "epoch": 0.05192870444267092,
+      "grad_norm": 0.8093993647109643,
+      "learning_rate": 1.7304964539007094e-05,
+      "loss": 7.2088,
+      "step": 244
+    },
+    {
+      "epoch": 0.0521415270018622,
+      "grad_norm": 0.9389100957172557,
+      "learning_rate": 1.7375886524822697e-05,
+      "loss": 7.2077,
+      "step": 245
+    },
+    {
+      "epoch": 0.05235434956105347,
+      "grad_norm": 1.0314608971152408,
+      "learning_rate": 1.74468085106383e-05,
+      "loss": 7.1993,
+      "step": 246
+    },
+    {
+      "epoch": 0.05256717212024475,
+      "grad_norm": 1.0672691729163957,
+      "learning_rate": 1.7517730496453903e-05,
+      "loss": 7.1351,
+      "step": 247
+    },
+    {
+      "epoch": 0.05277999467943602,
+      "grad_norm": 0.8888564120735334,
+      "learning_rate": 1.7588652482269506e-05,
+      "loss": 7.1209,
+      "step": 248
+    },
+    {
+      "epoch": 0.052992817238627296,
+      "grad_norm": 1.5604024888242365,
+      "learning_rate": 1.7659574468085106e-05,
+      "loss": 7.1045,
+      "step": 249
+    },
+    {
+      "epoch": 0.05320563979781857,
+      "grad_norm": 1.4802873858635037,
+      "learning_rate": 1.773049645390071e-05,
+      "loss": 7.1335,
+      "step": 250
+    },
+    {
+      "epoch": 0.053418462357009845,
+      "grad_norm": 1.013477926018591,
+      "learning_rate": 1.7801418439716312e-05,
+      "loss": 7.1092,
+      "step": 251
+    },
+    {
+      "epoch": 0.053631284916201116,
+      "grad_norm": 1.3692131316491678,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 7.094,
+      "step": 252
+    },
+    {
+      "epoch": 0.053844107475392394,
+      "grad_norm": 1.2589198390809315,
+      "learning_rate": 1.7943262411347518e-05,
+      "loss": 7.0576,
+      "step": 253
+    },
+    {
+      "epoch": 0.054056930034583665,
+      "grad_norm": 1.4007917238018461,
+      "learning_rate": 1.801418439716312e-05,
+      "loss": 7.1136,
+      "step": 254
+    },
+    {
+      "epoch": 0.05426975259377494,
+      "grad_norm": 1.5278901597911807,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 7.0749,
+      "step": 255
+    },
+    {
+      "epoch": 0.054482575152966214,
+      "grad_norm": 1.0516147454595859,
+      "learning_rate": 1.8156028368794327e-05,
+      "loss": 7.0947,
+      "step": 256
+    },
+    {
+      "epoch": 0.054695397712157486,
+      "grad_norm": 1.448909384763878,
+      "learning_rate": 1.822695035460993e-05,
+      "loss": 7.0406,
+      "step": 257
+    },
+    {
+      "epoch": 0.054908220271348764,
+      "grad_norm": 1.3185597984441393,
+      "learning_rate": 1.8297872340425533e-05,
+      "loss": 7.0495,
+      "step": 258
+    },
+    {
+      "epoch": 0.055121042830540035,
+      "grad_norm": 1.6057359929130517,
+      "learning_rate": 1.8368794326241136e-05,
+      "loss": 7.0382,
+      "step": 259
+    },
+    {
+      "epoch": 0.05533386538973131,
+      "grad_norm": 1.1811869905171015,
+      "learning_rate": 1.8439716312056736e-05,
+      "loss": 7.0716,
+      "step": 260
+    },
+    {
+      "epoch": 0.055546687948922584,
+      "grad_norm": 1.168923109920534,
+      "learning_rate": 1.851063829787234e-05,
+      "loss": 7.0339,
+      "step": 261
+    },
+    {
+      "epoch": 0.05575951050811386,
+      "grad_norm": 2.011005938491588,
+      "learning_rate": 1.858156028368794e-05,
+      "loss": 7.0301,
+      "step": 262
+    },
+    {
+      "epoch": 0.05597233306730513,
+      "grad_norm": 0.984221472358444,
+      "learning_rate": 1.8652482269503545e-05,
+      "loss": 7.038,
+      "step": 263
+    },
+    {
+      "epoch": 0.05618515562649641,
+      "grad_norm": 1.5955355087650673,
+      "learning_rate": 1.8723404255319148e-05,
+      "loss": 6.956,
+      "step": 264
+    },
+    {
+      "epoch": 0.05639797818568768,
+      "grad_norm": 1.049980291123236,
+      "learning_rate": 1.879432624113475e-05,
+      "loss": 6.9807,
+      "step": 265
+    },
+    {
+      "epoch": 0.05661080074487896,
+      "grad_norm": 1.1897738625067185,
+      "learning_rate": 1.8865248226950354e-05,
+      "loss": 6.9967,
+      "step": 266
+    },
+    {
+      "epoch": 0.05682362330407023,
+      "grad_norm": 2.500268730807578,
+      "learning_rate": 1.8936170212765957e-05,
+      "loss": 7.0236,
+      "step": 267
+    },
+    {
+      "epoch": 0.05703644586326151,
+      "grad_norm": 0.7307747432930257,
+      "learning_rate": 1.900709219858156e-05,
+      "loss": 6.9649,
+      "step": 268
+    },
+    {
+      "epoch": 0.05724926842245278,
+      "grad_norm": 2.495244048496352,
+      "learning_rate": 1.9078014184397163e-05,
+      "loss": 6.9493,
+      "step": 269
+    },
+    {
+      "epoch": 0.05746209098164405,
+      "grad_norm": 1.6903526996500675,
+      "learning_rate": 1.9148936170212766e-05,
+      "loss": 6.9446,
+      "step": 270
+    },
+    {
+      "epoch": 0.05767491354083533,
+      "grad_norm": 1.260237882348046,
+      "learning_rate": 1.921985815602837e-05,
+      "loss": 6.9196,
+      "step": 271
+    },
+    {
+      "epoch": 0.0578877361000266,
+      "grad_norm": 1.5202240830064127,
+      "learning_rate": 1.929078014184397e-05,
+      "loss": 6.9056,
+      "step": 272
+    },
+    {
+      "epoch": 0.05810055865921788,
+      "grad_norm": 1.9965691930770237,
+      "learning_rate": 1.9361702127659575e-05,
+      "loss": 6.9298,
+      "step": 273
+    },
+    {
+      "epoch": 0.05831338121840915,
+      "grad_norm": 2.1385578111528227,
+      "learning_rate": 1.9432624113475178e-05,
+      "loss": 6.9225,
+      "step": 274
+    },
+    {
+      "epoch": 0.05852620377760043,
+      "grad_norm": 0.7840638208732541,
+      "learning_rate": 1.950354609929078e-05,
+      "loss": 6.896,
+      "step": 275
+    },
+    {
+      "epoch": 0.0587390263367917,
+      "grad_norm": 1.2201132183991665,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 6.9099,
+      "step": 276
+    },
+    {
+      "epoch": 0.058951848895982976,
+      "grad_norm": 2.415376982754387,
+      "learning_rate": 1.9645390070921987e-05,
+      "loss": 6.9058,
+      "step": 277
+    },
+    {
+      "epoch": 0.05916467145517425,
+      "grad_norm": 1.6609875952038415,
+      "learning_rate": 1.971631205673759e-05,
+      "loss": 6.8868,
+      "step": 278
+    },
+    {
+      "epoch": 0.059377494014365526,
+      "grad_norm": 1.8454343720607163,
+      "learning_rate": 1.9787234042553193e-05,
+      "loss": 6.879,
+      "step": 279
+    },
+    {
+      "epoch": 0.0595903165735568,
+      "grad_norm": 0.9913499843580947,
+      "learning_rate": 1.9858156028368796e-05,
+      "loss": 6.8626,
+      "step": 280
+    },
+    {
+      "epoch": 0.059803139132748075,
+      "grad_norm": 3.0750237656516974,
+      "learning_rate": 1.99290780141844e-05,
+      "loss": 6.8869,
+      "step": 281
+    },
+    {
+      "epoch": 0.060015961691939346,
+      "grad_norm": 2.470934216521558,
+      "learning_rate": 2e-05,
+      "loss": 6.858,
+      "step": 282
+    },
+    {
+      "epoch": 0.06022878425113062,
+      "grad_norm": 1.582891225391396,
+      "learning_rate": 2.0070921985815605e-05,
+      "loss": 6.8621,
+      "step": 283
+    },
+    {
+      "epoch": 0.060441606810321895,
+      "grad_norm": 1.536248772317487,
+      "learning_rate": 2.0141843971631208e-05,
+      "loss": 6.9056,
+      "step": 284
+    },
+    {
+      "epoch": 0.060654429369513166,
+      "grad_norm": 1.631945280485595,
+      "learning_rate": 2.0212765957446807e-05,
+      "loss": 6.817,
+      "step": 285
+    },
+    {
+      "epoch": 0.060867251928704444,
+      "grad_norm": 1.815508059309587,
+      "learning_rate": 2.028368794326241e-05,
+      "loss": 6.8637,
+      "step": 286
+    },
+    {
+      "epoch": 0.061080074487895715,
+      "grad_norm": 1.9918753165448742,
+      "learning_rate": 2.0354609929078013e-05,
+      "loss": 6.8609,
+      "step": 287
+    },
+    {
+      "epoch": 0.06129289704708699,
+      "grad_norm": 1.893411367651298,
+      "learning_rate": 2.0425531914893616e-05,
+      "loss": 6.8648,
+      "step": 288
+    },
+    {
+      "epoch": 0.061505719606278264,
+      "grad_norm": 1.4652764655467514,
+      "learning_rate": 2.049645390070922e-05,
+      "loss": 6.8653,
+      "step": 289
+    },
+    {
+      "epoch": 0.06171854216546954,
+      "grad_norm": 2.3785428028754265,
+      "learning_rate": 2.0567375886524822e-05,
+      "loss": 6.8061,
+      "step": 290
+    },
+    {
+      "epoch": 0.06193136472466081,
+      "grad_norm": 1.672966873753503,
+      "learning_rate": 2.0638297872340425e-05,
+      "loss": 6.8047,
+      "step": 291
+    },
+    {
+      "epoch": 0.06214418728385209,
+      "grad_norm": 1.1347151808601257,
+      "learning_rate": 2.0709219858156028e-05,
+      "loss": 6.7963,
+      "step": 292
+    },
+    {
+      "epoch": 0.06235700984304336,
+      "grad_norm": 2.0879248801483397,
+      "learning_rate": 2.078014184397163e-05,
+      "loss": 6.8512,
+      "step": 293
+    },
+    {
+      "epoch": 0.06256983240223464,
+      "grad_norm": 2.3615984542740565,
+      "learning_rate": 2.0851063829787234e-05,
+      "loss": 6.816,
+      "step": 294
+    },
+    {
+      "epoch": 0.0627826549614259,
+      "grad_norm": 1.1438209792695726,
+      "learning_rate": 2.0921985815602837e-05,
+      "loss": 6.7666,
+      "step": 295
+    },
+    {
+      "epoch": 0.06299547752061718,
+      "grad_norm": 1.731041018987362,
+      "learning_rate": 2.099290780141844e-05,
+      "loss": 6.7761,
+      "step": 296
+    },
+    {
+      "epoch": 0.06320830007980846,
+      "grad_norm": 1.9694340909460615,
+      "learning_rate": 2.1063829787234043e-05,
+      "loss": 6.7703,
+      "step": 297
+    },
+    {
+      "epoch": 0.06342112263899974,
+      "grad_norm": 1.988080086079872,
+      "learning_rate": 2.1134751773049646e-05,
+      "loss": 6.7783,
+      "step": 298
+    },
+    {
+      "epoch": 0.063633945198191,
+      "grad_norm": 1.8332186715864827,
+      "learning_rate": 2.120567375886525e-05,
+      "loss": 6.7958,
+      "step": 299
+    },
+    {
+      "epoch": 0.06384676775738228,
+      "grad_norm": 1.3889962145308627,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 6.7818,
+      "step": 300
+    },
+    {
+      "epoch": 0.06405959031657356,
+      "grad_norm": 2.8544456020492324,
+      "learning_rate": 2.1347517730496455e-05,
+      "loss": 6.7817,
+      "step": 301
+    },
+    {
+      "epoch": 0.06427241287576484,
+      "grad_norm": 1.188799358553835,
+      "learning_rate": 2.1418439716312058e-05,
+      "loss": 6.7603,
+      "step": 302
+    },
+    {
+      "epoch": 0.0644852354349561,
+      "grad_norm": 1.3386076895966692,
+      "learning_rate": 2.148936170212766e-05,
+      "loss": 6.7528,
+      "step": 303
+    },
+    {
+      "epoch": 0.06469805799414738,
+      "grad_norm": 2.868261994371003,
+      "learning_rate": 2.1560283687943264e-05,
+      "loss": 6.7248,
+      "step": 304
+    },
+    {
+      "epoch": 0.06491088055333866,
+      "grad_norm": 1.3709663013056754,
+      "learning_rate": 2.1631205673758867e-05,
+      "loss": 6.713,
+      "step": 305
+    },
+    {
+      "epoch": 0.06512370311252993,
+      "grad_norm": 2.5428913066606174,
+      "learning_rate": 2.170212765957447e-05,
+      "loss": 6.761,
+      "step": 306
+    },
+    {
+      "epoch": 0.0653365256717212,
+      "grad_norm": 1.3125182668754987,
+      "learning_rate": 2.1773049645390073e-05,
+      "loss": 6.7359,
+      "step": 307
+    },
+    {
+      "epoch": 0.06554934823091248,
+      "grad_norm": 3.2329659577273704,
+      "learning_rate": 2.1843971631205676e-05,
+      "loss": 6.7166,
+      "step": 308
+    },
+    {
+      "epoch": 0.06576217079010375,
+      "grad_norm": 2.210205435103367,
+      "learning_rate": 2.191489361702128e-05,
+      "loss": 6.7401,
+      "step": 309
+    },
+    {
+      "epoch": 0.06597499334929502,
+      "grad_norm": 2.595002205180199,
+      "learning_rate": 2.1985815602836882e-05,
+      "loss": 6.722,
+      "step": 310
+    },
+    {
+      "epoch": 0.0661878159084863,
+      "grad_norm": 2.0982377459362658,
+      "learning_rate": 2.2056737588652485e-05,
+      "loss": 6.6721,
+      "step": 311
+    },
+    {
+      "epoch": 0.06640063846767758,
+      "grad_norm": 3.3637152255348384,
+      "learning_rate": 2.2127659574468085e-05,
+      "loss": 6.7391,
+      "step": 312
+    },
+    {
+      "epoch": 0.06661346102686885,
+      "grad_norm": 3.2185452915352535,
+      "learning_rate": 2.2198581560283688e-05,
+      "loss": 6.7229,
+      "step": 313
+    },
+    {
+      "epoch": 0.06682628358606012,
+      "grad_norm": 1.290602262772931,
+      "learning_rate": 2.226950354609929e-05,
+      "loss": 6.7174,
+      "step": 314
+    },
+    {
+      "epoch": 0.0670391061452514,
+      "grad_norm": 1.7805797841285063,
+      "learning_rate": 2.2340425531914894e-05,
+      "loss": 6.6948,
+      "step": 315
+    },
+    {
+      "epoch": 0.06725192870444267,
+      "grad_norm": 1.7862072898094157,
+      "learning_rate": 2.2411347517730497e-05,
+      "loss": 6.685,
+      "step": 316
+    },
+    {
+      "epoch": 0.06746475126363395,
+      "grad_norm": 1.6234904893139719,
+      "learning_rate": 2.24822695035461e-05,
+      "loss": 6.6384,
+      "step": 317
+    },
+    {
+      "epoch": 0.06767757382282522,
+      "grad_norm": 2.4832236154876806,
+      "learning_rate": 2.2553191489361703e-05,
+      "loss": 6.6634,
+      "step": 318
+    },
+    {
+      "epoch": 0.0678903963820165,
+      "grad_norm": 1.8122618161262225,
+      "learning_rate": 2.2624113475177306e-05,
+      "loss": 6.6292,
+      "step": 319
+    },
+    {
+      "epoch": 0.06810321894120777,
+      "grad_norm": 1.1740982548650396,
+      "learning_rate": 2.269503546099291e-05,
+      "loss": 6.6673,
+      "step": 320
+    },
+    {
+      "epoch": 0.06831604150039904,
+      "grad_norm": 1.5341482539584008,
+      "learning_rate": 2.2765957446808512e-05,
+      "loss": 6.5993,
+      "step": 321
+    },
+    {
+      "epoch": 0.06852886405959031,
+      "grad_norm": 3.4469020437520963,
+      "learning_rate": 2.2836879432624115e-05,
+      "loss": 6.6891,
+      "step": 322
+    },
+    {
+      "epoch": 0.06874168661878159,
+      "grad_norm": 1.2090017422010528,
+      "learning_rate": 2.2907801418439718e-05,
+      "loss": 6.6598,
+      "step": 323
+    },
+    {
+      "epoch": 0.06895450917797287,
+      "grad_norm": 2.869623950867746,
+      "learning_rate": 2.297872340425532e-05,
+      "loss": 6.6282,
+      "step": 324
+    },
+    {
+      "epoch": 0.06916733173716413,
+      "grad_norm": 2.5791335292588102,
+      "learning_rate": 2.3049645390070924e-05,
+      "loss": 6.6338,
+      "step": 325
+    },
+    {
+      "epoch": 0.06938015429635541,
+      "grad_norm": 1.18186889143188,
+      "learning_rate": 2.3120567375886527e-05,
+      "loss": 6.6786,
+      "step": 326
+    },
+    {
+      "epoch": 0.06959297685554669,
+      "grad_norm": 4.11451022316282,
+      "learning_rate": 2.319148936170213e-05,
+      "loss": 6.595,
+      "step": 327
+    },
+    {
+      "epoch": 0.06980579941473797,
+      "grad_norm": 2.5234905671772134,
+      "learning_rate": 2.3262411347517733e-05,
+      "loss": 6.6361,
+      "step": 328
+    },
+    {
+      "epoch": 0.07001862197392923,
+      "grad_norm": 3.988239222295756,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 6.6072,
+      "step": 329
+    },
+    {
+      "epoch": 0.07023144453312051,
+      "grad_norm": 2.0061811249050447,
+      "learning_rate": 2.340425531914894e-05,
+      "loss": 6.6014,
+      "step": 330
+    },
+    {
+      "epoch": 0.07044426709231179,
+      "grad_norm": 4.440456940651688,
+      "learning_rate": 2.3475177304964542e-05,
+      "loss": 6.6146,
+      "step": 331
+    },
+    {
+      "epoch": 0.07065708965150307,
+      "grad_norm": 3.1281132432271233,
+      "learning_rate": 2.3546099290780145e-05,
+      "loss": 6.6351,
+      "step": 332
+    },
+    {
+      "epoch": 0.07086991221069433,
+      "grad_norm": 4.16938554989175,
+      "learning_rate": 2.3617021276595748e-05,
+      "loss": 6.6133,
+      "step": 333
+    },
+    {
+      "epoch": 0.07108273476988561,
+      "grad_norm": 3.9734728370189045,
+      "learning_rate": 2.368794326241135e-05,
+      "loss": 6.6332,
+      "step": 334
+    },
+    {
+      "epoch": 0.07129555732907689,
+      "grad_norm": 2.803999620825429,
+      "learning_rate": 2.3758865248226954e-05,
+      "loss": 6.5616,
+      "step": 335
+    },
+    {
+      "epoch": 0.07150837988826815,
+      "grad_norm": 2.7881936239543244,
+      "learning_rate": 2.3829787234042557e-05,
+      "loss": 6.6243,
+      "step": 336
+    },
+    {
+      "epoch": 0.07172120244745943,
+      "grad_norm": 3.342944622622095,
+      "learning_rate": 2.390070921985816e-05,
+      "loss": 6.5955,
+      "step": 337
+    },
+    {
+      "epoch": 0.0719340250066507,
+      "grad_norm": 2.556014051453452,
+      "learning_rate": 2.3971631205673763e-05,
+      "loss": 6.5845,
+      "step": 338
+    },
+    {
+      "epoch": 0.07214684756584198,
+      "grad_norm": 3.375379586609862,
+      "learning_rate": 2.4042553191489362e-05,
+      "loss": 6.5778,
+      "step": 339
+    },
+    {
+      "epoch": 0.07235967012503325,
+      "grad_norm": 2.7494986421384833,
+      "learning_rate": 2.4113475177304965e-05,
+      "loss": 6.5438,
+      "step": 340
+    },
+    {
+      "epoch": 0.07257249268422453,
+      "grad_norm": 3.4765964085973207,
+      "learning_rate": 2.418439716312057e-05,
+      "loss": 6.5721,
+      "step": 341
+    },
+    {
+      "epoch": 0.0727853152434158,
+      "grad_norm": 3.0864298069105005,
+      "learning_rate": 2.425531914893617e-05,
+      "loss": 6.5676,
+      "step": 342
+    },
+    {
+      "epoch": 0.07299813780260708,
+      "grad_norm": 2.602696673307974,
+      "learning_rate": 2.4326241134751774e-05,
+      "loss": 6.5923,
+      "step": 343
+    },
+    {
+      "epoch": 0.07321096036179835,
+      "grad_norm": 2.339897469228908,
+      "learning_rate": 2.4397163120567377e-05,
+      "loss": 6.5482,
+      "step": 344
+    },
+    {
+      "epoch": 0.07342378292098962,
+      "grad_norm": 2.855871775559828,
+      "learning_rate": 2.446808510638298e-05,
+      "loss": 6.5617,
+      "step": 345
+    },
+    {
+      "epoch": 0.0736366054801809,
+      "grad_norm": 2.469210546270912,
+      "learning_rate": 2.4539007092198583e-05,
+      "loss": 6.5549,
+      "step": 346
+    },
+    {
+      "epoch": 0.07384942803937217,
+      "grad_norm": 2.863386536357406,
+      "learning_rate": 2.4609929078014186e-05,
+      "loss": 6.5176,
+      "step": 347
+    },
+    {
+      "epoch": 0.07406225059856344,
+      "grad_norm": 2.680835849983496,
+      "learning_rate": 2.468085106382979e-05,
+      "loss": 6.5709,
+      "step": 348
+    },
+    {
+      "epoch": 0.07427507315775472,
+      "grad_norm": 2.5769594466555557,
+      "learning_rate": 2.4751773049645392e-05,
+      "loss": 6.5422,
+      "step": 349
+    },
+    {
+      "epoch": 0.074487895716946,
+      "grad_norm": 2.226980907199167,
+      "learning_rate": 2.4822695035460995e-05,
+      "loss": 6.5596,
+      "step": 350
+    },
+    {
+      "epoch": 0.07470071827613727,
+      "grad_norm": 2.575982000522183,
+      "learning_rate": 2.48936170212766e-05,
+      "loss": 6.4921,
+      "step": 351
+    },
+    {
+      "epoch": 0.07491354083532854,
+      "grad_norm": 2.2202109625372395,
+      "learning_rate": 2.49645390070922e-05,
+      "loss": 6.5567,
+      "step": 352
+    },
+    {
+      "epoch": 0.07512636339451982,
+      "grad_norm": 2.8129609398116657,
+      "learning_rate": 2.50354609929078e-05,
+      "loss": 6.5302,
+      "step": 353
+    },
+    {
+      "epoch": 0.0753391859537111,
+      "grad_norm": 2.4956271703499335,
+      "learning_rate": 2.5106382978723404e-05,
+      "loss": 6.5127,
+      "step": 354
+    },
+    {
+      "epoch": 0.07555200851290236,
+      "grad_norm": 2.367819045301763,
+      "learning_rate": 2.5177304964539007e-05,
+      "loss": 6.503,
+      "step": 355
+    },
+    {
+      "epoch": 0.07576483107209364,
+      "grad_norm": 2.082646649718499,
+      "learning_rate": 2.524822695035461e-05,
+      "loss": 6.4951,
+      "step": 356
+    },
+    {
+      "epoch": 0.07597765363128492,
+      "grad_norm": 2.5015401016324765,
+      "learning_rate": 2.5319148936170213e-05,
+      "loss": 6.5018,
+      "step": 357
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 2.19284269004095,
+      "learning_rate": 2.5390070921985816e-05,
+      "loss": 6.5164,
+      "step": 358
+    },
+    {
+      "epoch": 0.07640329874966746,
+      "grad_norm": 2.4276180958493683,
+      "learning_rate": 2.546099290780142e-05,
+      "loss": 6.504,
+      "step": 359
+    },
+    {
+      "epoch": 0.07661612130885874,
+      "grad_norm": 2.1548695110865466,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 6.5049,
+      "step": 360
+    },
+    {
+      "epoch": 0.07682894386805002,
+      "grad_norm": 2.3750763710331726,
+      "learning_rate": 2.5602836879432625e-05,
+      "loss": 6.4651,
+      "step": 361
+    },
+    {
+      "epoch": 0.07704176642724128,
+      "grad_norm": 2.0082971517406536,
+      "learning_rate": 2.5673758865248228e-05,
+      "loss": 6.4451,
+      "step": 362
+    },
+    {
+      "epoch": 0.07725458898643256,
+      "grad_norm": 2.2955142236084933,
+      "learning_rate": 2.574468085106383e-05,
+      "loss": 6.4792,
+      "step": 363
+    },
+    {
+      "epoch": 0.07746741154562384,
+      "grad_norm": 1.950345108910332,
+      "learning_rate": 2.581560283687943e-05,
+      "loss": 6.4843,
+      "step": 364
+    },
+    {
+      "epoch": 0.07768023410481512,
+      "grad_norm": 2.644325923544677,
+      "learning_rate": 2.5886524822695034e-05,
+      "loss": 6.4959,
+      "step": 365
+    },
+    {
+      "epoch": 0.07789305666400638,
+      "grad_norm": 2.2580546859817145,
+      "learning_rate": 2.5957446808510637e-05,
+      "loss": 6.4885,
+      "step": 366
+    },
+    {
+      "epoch": 0.07810587922319766,
+      "grad_norm": 2.0712160562092077,
+      "learning_rate": 2.602836879432624e-05,
+      "loss": 6.4854,
+      "step": 367
+    },
+    {
+      "epoch": 0.07831870178238894,
+      "grad_norm": 1.9677538925072182,
+      "learning_rate": 2.6099290780141843e-05,
+      "loss": 6.4697,
+      "step": 368
+    },
+    {
+      "epoch": 0.07853152434158021,
+      "grad_norm": 2.0731165495387134,
+      "learning_rate": 2.6170212765957446e-05,
+      "loss": 6.4513,
+      "step": 369
+    },
+    {
+      "epoch": 0.07874434690077148,
+      "grad_norm": 1.7513474906227278,
+      "learning_rate": 2.624113475177305e-05,
+      "loss": 6.4728,
+      "step": 370
+    },
+    {
+      "epoch": 0.07895716945996276,
+      "grad_norm": 2.5742350795356295,
+      "learning_rate": 2.631205673758865e-05,
+      "loss": 6.4323,
+      "step": 371
+    },
+    {
+      "epoch": 0.07916999201915403,
+      "grad_norm": 2.213926454600747,
+      "learning_rate": 2.6382978723404255e-05,
+      "loss": 6.4569,
+      "step": 372
+    },
+    {
+      "epoch": 0.0793828145783453,
+      "grad_norm": 2.0741097472238197,
+      "learning_rate": 2.6453900709219858e-05,
+      "loss": 6.423,
+      "step": 373
+    },
+    {
+      "epoch": 0.07959563713753658,
+      "grad_norm": 2.0078484022724163,
+      "learning_rate": 2.652482269503546e-05,
+      "loss": 6.4391,
+      "step": 374
+    },
+    {
+      "epoch": 0.07980845969672785,
+      "grad_norm": 1.9921032118052973,
+      "learning_rate": 2.6595744680851064e-05,
+      "loss": 6.4516,
+      "step": 375
+    },
+    {
+      "epoch": 0.08002128225591913,
+      "grad_norm": 1.7416569207341475,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 6.382,
+      "step": 376
+    },
+    {
+      "epoch": 0.0802341048151104,
+      "grad_norm": 2.5178625540964044,
+      "learning_rate": 2.673758865248227e-05,
+      "loss": 6.4122,
+      "step": 377
+    },
+    {
+      "epoch": 0.08044692737430167,
+      "grad_norm": 2.25380457161636,
+      "learning_rate": 2.6808510638297873e-05,
+      "loss": 6.4173,
+      "step": 378
+    },
+    {
+      "epoch": 0.08065974993349295,
+      "grad_norm": 1.7160244530047115,
+      "learning_rate": 2.6879432624113476e-05,
+      "loss": 6.4027,
+      "step": 379
+    },
+    {
+      "epoch": 0.08087257249268423,
+      "grad_norm": 1.7678584172245826,
+      "learning_rate": 2.695035460992908e-05,
+      "loss": 6.4282,
+      "step": 380
+    },
+    {
+      "epoch": 0.0810853950518755,
+      "grad_norm": 2.0680773311719953,
+      "learning_rate": 2.702127659574468e-05,
+      "loss": 6.4034,
+      "step": 381
+    },
+    {
+      "epoch": 0.08129821761106677,
+      "grad_norm": 1.7675550414933272,
+      "learning_rate": 2.7092198581560285e-05,
+      "loss": 6.3966,
+      "step": 382
+    },
+    {
+      "epoch": 0.08151104017025805,
+      "grad_norm": 2.0842314340125188,
+      "learning_rate": 2.7163120567375888e-05,
+      "loss": 6.3487,
+      "step": 383
+    },
+    {
+      "epoch": 0.08172386272944931,
+      "grad_norm": 2.0263626865914723,
+      "learning_rate": 2.723404255319149e-05,
+      "loss": 6.4111,
+      "step": 384
+    },
+    {
+      "epoch": 0.08193668528864059,
+      "grad_norm": 2.0715215782683023,
+      "learning_rate": 2.7304964539007094e-05,
+      "loss": 6.3723,
+      "step": 385
+    },
+    {
+      "epoch": 0.08214950784783187,
+      "grad_norm": 1.753090130747976,
+      "learning_rate": 2.7375886524822697e-05,
+      "loss": 6.3673,
+      "step": 386
+    },
+    {
+      "epoch": 0.08236233040702315,
+      "grad_norm": 2.073534328655711,
+      "learning_rate": 2.74468085106383e-05,
+      "loss": 6.3659,
+      "step": 387
+    },
+    {
+      "epoch": 0.08257515296621441,
+      "grad_norm": 2.113696062245005,
+      "learning_rate": 2.7517730496453903e-05,
+      "loss": 6.356,
+      "step": 388
+    },
+    {
+      "epoch": 0.08278797552540569,
+      "grad_norm": 1.6363372610601075,
+      "learning_rate": 2.7588652482269506e-05,
+      "loss": 6.3988,
+      "step": 389
+    },
+    {
+      "epoch": 0.08300079808459697,
+      "grad_norm": 2.0532849287150325,
+      "learning_rate": 2.765957446808511e-05,
+      "loss": 6.3286,
+      "step": 390
+    },
+    {
+      "epoch": 0.08321362064378825,
+      "grad_norm": 1.4080516878557767,
+      "learning_rate": 2.773049645390071e-05,
+      "loss": 6.3368,
+      "step": 391
+    },
+    {
+      "epoch": 0.08342644320297951,
+      "grad_norm": 1.4180982025807933,
+      "learning_rate": 2.7801418439716315e-05,
+      "loss": 6.3415,
+      "step": 392
+    },
+    {
+      "epoch": 0.08363926576217079,
+      "grad_norm": 2.2527158602349244,
+      "learning_rate": 2.7872340425531918e-05,
+      "loss": 6.339,
+      "step": 393
+    },
+    {
+      "epoch": 0.08385208832136207,
+      "grad_norm": 1.294896496090028,
+      "learning_rate": 2.794326241134752e-05,
+      "loss": 6.3389,
+      "step": 394
+    },
+    {
+      "epoch": 0.08406491088055335,
+      "grad_norm": 2.246276835451167,
+      "learning_rate": 2.8014184397163124e-05,
+      "loss": 6.2916,
+      "step": 395
+    },
+    {
+      "epoch": 0.08427773343974461,
+      "grad_norm": 1.7008474230464479,
+      "learning_rate": 2.8085106382978727e-05,
+      "loss": 6.3101,
+      "step": 396
+    },
+    {
+      "epoch": 0.08449055599893589,
+      "grad_norm": 1.7743478483643234,
+      "learning_rate": 2.815602836879433e-05,
+      "loss": 6.3318,
+      "step": 397
+    },
+    {
+      "epoch": 0.08470337855812717,
+      "grad_norm": 1.7121447676619959,
+      "learning_rate": 2.8226950354609933e-05,
+      "loss": 6.3081,
+      "step": 398
+    },
+    {
+      "epoch": 0.08491620111731843,
+      "grad_norm": 1.5929164677228564,
+      "learning_rate": 2.8297872340425536e-05,
+      "loss": 6.3288,
+      "step": 399
+    },
+    {
+      "epoch": 0.08512902367650971,
+      "grad_norm": 2.350664205338845,
+      "learning_rate": 2.836879432624114e-05,
+      "loss": 6.3552,
+      "step": 400
+    },
+    {
+      "epoch": 0.08534184623570099,
+      "grad_norm": 1.059715236523428,
+      "learning_rate": 2.843971631205674e-05,
+      "loss": 6.2619,
+      "step": 401
+    },
+    {
+      "epoch": 0.08555466879489226,
+      "grad_norm": 2.4744515788472388,
+      "learning_rate": 2.8510638297872345e-05,
+      "loss": 6.3235,
+      "step": 402
+    },
+    {
+      "epoch": 0.08576749135408353,
+      "grad_norm": 1.2108736680274637,
+      "learning_rate": 2.8581560283687948e-05,
+      "loss": 6.3064,
+      "step": 403
+    },
+    {
+      "epoch": 0.0859803139132748,
+      "grad_norm": 2.4778985397781472,
+      "learning_rate": 2.865248226950355e-05,
+      "loss": 6.3333,
+      "step": 404
+    },
+    {
+      "epoch": 0.08619313647246608,
+      "grad_norm": 2.1667840525199824,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 6.29,
+      "step": 405
+    },
+    {
+      "epoch": 0.08640595903165736,
+      "grad_norm": 1.608801567559026,
+      "learning_rate": 2.8794326241134757e-05,
+      "loss": 6.2939,
+      "step": 406
+    },
+    {
+      "epoch": 0.08661878159084863,
+      "grad_norm": 2.6577133792275034,
+      "learning_rate": 2.8865248226950353e-05,
+      "loss": 6.2611,
+      "step": 407
+    },
+    {
+      "epoch": 0.0868316041500399,
+      "grad_norm": 1.947756835176513,
+      "learning_rate": 2.8936170212765956e-05,
+      "loss": 6.3126,
+      "step": 408
+    },
+    {
+      "epoch": 0.08704442670923118,
+      "grad_norm": 1.47352965618298,
+      "learning_rate": 2.900709219858156e-05,
+      "loss": 6.3032,
+      "step": 409
+    },
+    {
+      "epoch": 0.08725724926842245,
+      "grad_norm": 3.087589370998148,
+      "learning_rate": 2.9078014184397162e-05,
+      "loss": 6.3108,
+      "step": 410
+    },
+    {
+      "epoch": 0.08747007182761372,
+      "grad_norm": 1.6955029867016813,
+      "learning_rate": 2.9148936170212765e-05,
+      "loss": 6.23,
+      "step": 411
+    },
+    {
+      "epoch": 0.087682894386805,
+      "grad_norm": 2.2415345979208503,
+      "learning_rate": 2.9219858156028368e-05,
+      "loss": 6.2003,
+      "step": 412
+    },
+    {
+      "epoch": 0.08789571694599628,
+      "grad_norm": 1.5257904595978213,
+      "learning_rate": 2.929078014184397e-05,
+      "loss": 6.2908,
+      "step": 413
+    },
+    {
+      "epoch": 0.08810853950518754,
+      "grad_norm": 2.5856897389382274,
+      "learning_rate": 2.9361702127659574e-05,
+      "loss": 6.2914,
+      "step": 414
+    },
+    {
+      "epoch": 0.08832136206437882,
+      "grad_norm": 1.682540946672848,
+      "learning_rate": 2.9432624113475177e-05,
+      "loss": 6.2071,
+      "step": 415
+    },
+    {
+      "epoch": 0.0885341846235701,
+      "grad_norm": 2.2575786178741386,
+      "learning_rate": 2.950354609929078e-05,
+      "loss": 6.222,
+      "step": 416
+    },
+    {
+      "epoch": 0.08874700718276138,
+      "grad_norm": 1.692167110994162,
+      "learning_rate": 2.9574468085106383e-05,
+      "loss": 6.1931,
+      "step": 417
+    },
+    {
+      "epoch": 0.08895982974195264,
+      "grad_norm": 1.6431292900628451,
+      "learning_rate": 2.9645390070921986e-05,
+      "loss": 6.1578,
+      "step": 418
+    },
+    {
+      "epoch": 0.08917265230114392,
+      "grad_norm": 2.3993730214603897,
+      "learning_rate": 2.971631205673759e-05,
+      "loss": 6.2268,
+      "step": 419
+    },
+    {
+      "epoch": 0.0893854748603352,
+      "grad_norm": 1.8886121387561818,
+      "learning_rate": 2.9787234042553192e-05,
+      "loss": 6.1925,
+      "step": 420
+    },
+    {
+      "epoch": 0.08959829741952648,
+      "grad_norm": 1.9209464904084865,
+      "learning_rate": 2.9858156028368795e-05,
+      "loss": 6.2214,
+      "step": 421
+    },
+    {
+      "epoch": 0.08981111997871774,
+      "grad_norm": 1.726395061550026,
+      "learning_rate": 2.9929078014184398e-05,
+      "loss": 6.2305,
+      "step": 422
+    },
+    {
+      "epoch": 0.09002394253790902,
+      "grad_norm": 1.5256333452946533,
+      "learning_rate": 3e-05,
+      "loss": 6.1796,
+      "step": 423
+    },
+    {
+      "epoch": 0.0902367650971003,
+      "grad_norm": 1.2841560278000954,
+      "learning_rate": 3.0070921985815604e-05,
+      "loss": 6.2163,
+      "step": 424
+    },
+    {
+      "epoch": 0.09044958765629156,
+      "grad_norm": 1.7058146635809994,
+      "learning_rate": 3.0141843971631207e-05,
+      "loss": 6.1786,
+      "step": 425
+    },
+    {
+      "epoch": 0.09066241021548284,
+      "grad_norm": 1.4477519776677827,
+      "learning_rate": 3.021276595744681e-05,
+      "loss": 6.2005,
+      "step": 426
+    },
+    {
+      "epoch": 0.09087523277467412,
+      "grad_norm": 1.4857057197653432,
+      "learning_rate": 3.0283687943262413e-05,
+      "loss": 6.1205,
+      "step": 427
+    },
+    {
+      "epoch": 0.0910880553338654,
+      "grad_norm": 1.9273313491782105,
+      "learning_rate": 3.0354609929078016e-05,
+      "loss": 6.2149,
+      "step": 428
+    },
+    {
+      "epoch": 0.09130087789305666,
+      "grad_norm": 1.9821030620566553,
+      "learning_rate": 3.042553191489362e-05,
+      "loss": 6.153,
+      "step": 429
+    },
+    {
+      "epoch": 0.09151370045224794,
+      "grad_norm": 1.9155632716695188,
+      "learning_rate": 3.0496453900709222e-05,
+      "loss": 6.1582,
+      "step": 430
+    },
+    {
+      "epoch": 0.09172652301143921,
+      "grad_norm": 1.688206642360842,
+      "learning_rate": 3.056737588652482e-05,
+      "loss": 6.1064,
+      "step": 431
+    },
+    {
+      "epoch": 0.09193934557063049,
+      "grad_norm": 1.658693460139449,
+      "learning_rate": 3.063829787234043e-05,
+      "loss": 6.1696,
+      "step": 432
+    },
+    {
+      "epoch": 0.09215216812982176,
+      "grad_norm": 2.157667262990166,
+      "learning_rate": 3.070921985815603e-05,
+      "loss": 6.1505,
+      "step": 433
+    },
+    {
+      "epoch": 0.09236499068901304,
+      "grad_norm": 1.6007398190418252,
+      "learning_rate": 3.0780141843971634e-05,
+      "loss": 6.1046,
+      "step": 434
+    },
+    {
+      "epoch": 0.09257781324820431,
+      "grad_norm": 1.5538027254583349,
+      "learning_rate": 3.085106382978723e-05,
+      "loss": 6.1373,
+      "step": 435
+    },
+    {
+      "epoch": 0.09279063580739558,
+      "grad_norm": 1.3863727277017157,
+      "learning_rate": 3.092198581560284e-05,
+      "loss": 6.1475,
+      "step": 436
+    },
+    {
+      "epoch": 0.09300345836658686,
+      "grad_norm": 2.200966760199646,
+      "learning_rate": 3.099290780141844e-05,
+      "loss": 6.1669,
+      "step": 437
+    },
+    {
+      "epoch": 0.09321628092577813,
+      "grad_norm": 1.2365259998771125,
+      "learning_rate": 3.1063829787234046e-05,
+      "loss": 6.1666,
+      "step": 438
+    },
+    {
+      "epoch": 0.09342910348496941,
+      "grad_norm": 2.42874055344113,
+      "learning_rate": 3.1134751773049645e-05,
+      "loss": 6.1026,
+      "step": 439
+    },
+    {
+      "epoch": 0.09364192604416068,
+      "grad_norm": 2.235912132272135,
+      "learning_rate": 3.120567375886525e-05,
+      "loss": 6.0964,
+      "step": 440
+    },
+    {
+      "epoch": 0.09385474860335195,
+      "grad_norm": 1.535202232951564,
+      "learning_rate": 3.127659574468085e-05,
+      "loss": 6.075,
+      "step": 441
+    },
+    {
+      "epoch": 0.09406757116254323,
+      "grad_norm": 2.051848130867705,
+      "learning_rate": 3.134751773049646e-05,
+      "loss": 6.1196,
+      "step": 442
+    },
+    {
+      "epoch": 0.09428039372173451,
+      "grad_norm": 1.2558125540707046,
+      "learning_rate": 3.141843971631206e-05,
+      "loss": 6.1203,
+      "step": 443
+    },
+    {
+      "epoch": 0.09449321628092577,
+      "grad_norm": 2.2378433356958336,
+      "learning_rate": 3.1489361702127664e-05,
+      "loss": 6.142,
+      "step": 444
+    },
+    {
+      "epoch": 0.09470603884011705,
+      "grad_norm": 1.4455975391357103,
+      "learning_rate": 3.156028368794326e-05,
+      "loss": 6.0614,
+      "step": 445
+    },
+    {
+      "epoch": 0.09491886139930833,
+      "grad_norm": 2.2392992133596064,
+      "learning_rate": 3.163120567375887e-05,
+      "loss": 6.0535,
+      "step": 446
+    },
+    {
+      "epoch": 0.09513168395849961,
+      "grad_norm": 1.6590113029856695,
+      "learning_rate": 3.170212765957447e-05,
+      "loss": 6.0691,
+      "step": 447
+    },
+    {
+      "epoch": 0.09534450651769087,
+      "grad_norm": 1.7111901414336221,
+      "learning_rate": 3.1773049645390076e-05,
+      "loss": 6.1223,
+      "step": 448
+    },
+    {
+      "epoch": 0.09555732907688215,
+      "grad_norm": 2.074761409238104,
+      "learning_rate": 3.1843971631205675e-05,
+      "loss": 6.1085,
+      "step": 449
+    },
+    {
+      "epoch": 0.09577015163607343,
+      "grad_norm": 1.6523657927768975,
+      "learning_rate": 3.191489361702128e-05,
+      "loss": 6.0814,
+      "step": 450
+    },
+    {
+      "epoch": 0.09598297419526469,
+      "grad_norm": 1.8830933261350262,
+      "learning_rate": 3.198581560283688e-05,
+      "loss": 6.0842,
+      "step": 451
+    },
+    {
+      "epoch": 0.09619579675445597,
+      "grad_norm": 2.0554827121551034,
+      "learning_rate": 3.205673758865249e-05,
+      "loss": 6.0911,
+      "step": 452
+    },
+    {
+      "epoch": 0.09640861931364725,
+      "grad_norm": 2.240358957928748,
+      "learning_rate": 3.212765957446809e-05,
+      "loss": 6.1013,
+      "step": 453
+    },
+    {
+      "epoch": 0.09662144187283853,
+      "grad_norm": 2.1091134052066396,
+      "learning_rate": 3.2198581560283694e-05,
+      "loss": 6.0271,
+      "step": 454
+    },
+    {
+      "epoch": 0.09683426443202979,
+      "grad_norm": 1.836285103059971,
+      "learning_rate": 3.226950354609929e-05,
+      "loss": 6.0427,
+      "step": 455
+    },
+    {
+      "epoch": 0.09704708699122107,
+      "grad_norm": 1.4219509268263444,
+      "learning_rate": 3.23404255319149e-05,
+      "loss": 6.0648,
+      "step": 456
+    },
+    {
+      "epoch": 0.09725990955041235,
+      "grad_norm": 3.1264766518437597,
+      "learning_rate": 3.24113475177305e-05,
+      "loss": 6.0128,
+      "step": 457
+    },
+    {
+      "epoch": 0.09747273210960362,
+      "grad_norm": 1.9025341842631223,
+      "learning_rate": 3.2482269503546106e-05,
+      "loss": 6.0237,
+      "step": 458
+    },
+    {
+      "epoch": 0.09768555466879489,
+      "grad_norm": 2.5350008996458957,
+      "learning_rate": 3.2553191489361705e-05,
+      "loss": 6.0583,
+      "step": 459
+    },
+    {
+      "epoch": 0.09789837722798617,
+      "grad_norm": 2.3641765829073424,
+      "learning_rate": 3.262411347517731e-05,
+      "loss": 6.0,
+      "step": 460
+    },
+    {
+      "epoch": 0.09811119978717744,
+      "grad_norm": 2.339488329970607,
+      "learning_rate": 3.2695035460992905e-05,
+      "loss": 6.0317,
+      "step": 461
+    },
+    {
+      "epoch": 0.09832402234636871,
+      "grad_norm": 1.7178984730442886,
+      "learning_rate": 3.276595744680851e-05,
+      "loss": 5.9947,
+      "step": 462
+    },
+    {
+      "epoch": 0.09853684490555999,
+      "grad_norm": 1.8612321372245293,
+      "learning_rate": 3.283687943262411e-05,
+      "loss": 6.0109,
+      "step": 463
+    },
+    {
+      "epoch": 0.09874966746475126,
+      "grad_norm": 2.6349543180778356,
+      "learning_rate": 3.290780141843972e-05,
+      "loss": 6.0094,
+      "step": 464
+    },
+    {
+      "epoch": 0.09896249002394254,
+      "grad_norm": 1.7189232682409057,
+      "learning_rate": 3.2978723404255317e-05,
+      "loss": 5.9458,
+      "step": 465
+    },
+    {
+      "epoch": 0.0991753125831338,
+      "grad_norm": 2.5037137114835835,
+      "learning_rate": 3.304964539007092e-05,
+      "loss": 5.9897,
+      "step": 466
+    },
+    {
+      "epoch": 0.09938813514232508,
+      "grad_norm": 1.559923293471545,
+      "learning_rate": 3.312056737588652e-05,
+      "loss": 5.9823,
+      "step": 467
+    },
+    {
+      "epoch": 0.09960095770151636,
+      "grad_norm": 2.3853974777392275,
+      "learning_rate": 3.319148936170213e-05,
+      "loss": 6.0084,
+      "step": 468
+    },
+    {
+      "epoch": 0.09981378026070764,
+      "grad_norm": 2.139248581408161,
+      "learning_rate": 3.326241134751773e-05,
+      "loss": 5.9898,
+      "step": 469
+    },
+    {
+      "epoch": 0.1000266028198989,
+      "grad_norm": 2.1927655364659078,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 5.9389,
+      "step": 470
+    },
+    {
+      "epoch": 0.10023942537909018,
+      "grad_norm": 1.4178295250719388,
+      "learning_rate": 3.3404255319148935e-05,
+      "loss": 5.9701,
+      "step": 471
+    },
+    {
+      "epoch": 0.10045224793828146,
+      "grad_norm": 1.7704926133847732,
+      "learning_rate": 3.347517730496454e-05,
+      "loss": 5.9667,
+      "step": 472
+    },
+    {
+      "epoch": 0.10066507049747274,
+      "grad_norm": 1.9269431313405934,
+      "learning_rate": 3.354609929078014e-05,
+      "loss": 5.9878,
+      "step": 473
+    },
+    {
+      "epoch": 0.100877893056664,
+      "grad_norm": 2.446440704942073,
+      "learning_rate": 3.361702127659575e-05,
+      "loss": 5.9589,
+      "step": 474
+    },
+    {
+      "epoch": 0.10109071561585528,
+      "grad_norm": 1.5978418040072218,
+      "learning_rate": 3.3687943262411347e-05,
+      "loss": 5.9189,
+      "step": 475
+    },
+    {
+      "epoch": 0.10130353817504656,
+      "grad_norm": 2.1362437116922246,
+      "learning_rate": 3.375886524822695e-05,
+      "loss": 5.9853,
+      "step": 476
+    },
+    {
+      "epoch": 0.10151636073423782,
+      "grad_norm": 2.2960948090577205,
+      "learning_rate": 3.382978723404255e-05,
+      "loss": 5.9596,
+      "step": 477
+    },
+    {
+      "epoch": 0.1017291832934291,
+      "grad_norm": 1.8251593680498284,
+      "learning_rate": 3.390070921985816e-05,
+      "loss": 5.9587,
+      "step": 478
+    },
+    {
+      "epoch": 0.10194200585262038,
+      "grad_norm": 2.555195351590785,
+      "learning_rate": 3.397163120567376e-05,
+      "loss": 5.9434,
+      "step": 479
+    },
+    {
+      "epoch": 0.10215482841181166,
+      "grad_norm": 1.882469015414435,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 5.8897,
+      "step": 480
+    },
+    {
+      "epoch": 0.10236765097100292,
+      "grad_norm": 2.255563208550438,
+      "learning_rate": 3.4113475177304965e-05,
+      "loss": 5.9775,
+      "step": 481
+    },
+    {
+      "epoch": 0.1025804735301942,
+      "grad_norm": 1.7706430710825514,
+      "learning_rate": 3.418439716312057e-05,
+      "loss": 5.9011,
+      "step": 482
+    },
+    {
+      "epoch": 0.10279329608938548,
+      "grad_norm": 2.482709746390058,
+      "learning_rate": 3.425531914893617e-05,
+      "loss": 5.926,
+      "step": 483
+    },
+    {
+      "epoch": 0.10300611864857676,
+      "grad_norm": 2.1784817561327463,
+      "learning_rate": 3.432624113475178e-05,
+      "loss": 5.9572,
+      "step": 484
+    },
+    {
+      "epoch": 0.10321894120776802,
+      "grad_norm": 1.4656959510699579,
+      "learning_rate": 3.4397163120567377e-05,
+      "loss": 5.8741,
+      "step": 485
+    },
+    {
+      "epoch": 0.1034317637669593,
+      "grad_norm": 2.480539156666848,
+      "learning_rate": 3.446808510638298e-05,
+      "loss": 5.9373,
+      "step": 486
+    },
+    {
+      "epoch": 0.10364458632615058,
+      "grad_norm": 2.063674824325811,
+      "learning_rate": 3.453900709219858e-05,
+      "loss": 5.906,
+      "step": 487
+    },
+    {
+      "epoch": 0.10385740888534184,
+      "grad_norm": 1.6472404526360889,
+      "learning_rate": 3.460992907801419e-05,
+      "loss": 5.9013,
+      "step": 488
+    },
+    {
+      "epoch": 0.10407023144453312,
+      "grad_norm": 2.9058582960355275,
+      "learning_rate": 3.468085106382979e-05,
+      "loss": 5.8677,
+      "step": 489
+    },
+    {
+      "epoch": 0.1042830540037244,
+      "grad_norm": 1.4393640875632463,
+      "learning_rate": 3.4751773049645395e-05,
+      "loss": 5.8847,
+      "step": 490
+    },
+    {
+      "epoch": 0.10449587656291567,
+      "grad_norm": 3.0318201709742056,
+      "learning_rate": 3.4822695035460995e-05,
+      "loss": 5.85,
+      "step": 491
+    },
+    {
+      "epoch": 0.10470869912210694,
+      "grad_norm": 1.7462058639100058,
+      "learning_rate": 3.48936170212766e-05,
+      "loss": 5.9069,
+      "step": 492
+    },
+    {
+      "epoch": 0.10492152168129822,
+      "grad_norm": 2.297207573020371,
+      "learning_rate": 3.49645390070922e-05,
+      "loss": 5.8906,
+      "step": 493
+    },
+    {
+      "epoch": 0.1051343442404895,
+      "grad_norm": 1.5266915613844079,
+      "learning_rate": 3.503546099290781e-05,
+      "loss": 5.9006,
+      "step": 494
+    },
+    {
+      "epoch": 0.10534716679968077,
+      "grad_norm": 2.2754923478843487,
+      "learning_rate": 3.5106382978723407e-05,
+      "loss": 5.849,
+      "step": 495
+    },
+    {
+      "epoch": 0.10555998935887204,
+      "grad_norm": 1.6692737811946954,
+      "learning_rate": 3.517730496453901e-05,
+      "loss": 5.8818,
+      "step": 496
+    },
+    {
+      "epoch": 0.10577281191806331,
+      "grad_norm": 2.2348599485595178,
+      "learning_rate": 3.524822695035461e-05,
+      "loss": 5.8253,
+      "step": 497
+    },
+    {
+      "epoch": 0.10598563447725459,
+      "grad_norm": 2.166768628675344,
+      "learning_rate": 3.531914893617021e-05,
+      "loss": 5.8727,
+      "step": 498
+    },
+    {
+      "epoch": 0.10619845703644586,
+      "grad_norm": 2.2754604583260214,
+      "learning_rate": 3.539007092198582e-05,
+      "loss": 5.8763,
+      "step": 499
+    },
+    {
+      "epoch": 0.10641127959563713,
+      "grad_norm": 2.1282536333984714,
+      "learning_rate": 3.546099290780142e-05,
+      "loss": 5.8526,
+      "step": 500
+    },
+    {
+      "epoch": 0.10662410215482841,
+      "grad_norm": 2.109930002807769,
+      "learning_rate": 3.5531914893617025e-05,
+      "loss": 5.8756,
+      "step": 501
+    },
+    {
+      "epoch": 0.10683692471401969,
+      "grad_norm": 2.733726324208394,
+      "learning_rate": 3.5602836879432624e-05,
+      "loss": 5.8987,
+      "step": 502
+    },
+    {
+      "epoch": 0.10704974727321095,
+      "grad_norm": 1.8721234015665826,
+      "learning_rate": 3.567375886524823e-05,
+      "loss": 5.8436,
+      "step": 503
+    },
+    {
+      "epoch": 0.10726256983240223,
+      "grad_norm": 2.215145018823617,
+      "learning_rate": 3.574468085106383e-05,
+      "loss": 5.843,
+      "step": 504
+    },
+    {
+      "epoch": 0.10747539239159351,
+      "grad_norm": 2.2907764359981075,
+      "learning_rate": 3.5815602836879437e-05,
+      "loss": 5.8799,
+      "step": 505
+    },
+    {
+      "epoch": 0.10768821495078479,
+      "grad_norm": 1.7874831884786886,
+      "learning_rate": 3.5886524822695036e-05,
+      "loss": 5.7926,
+      "step": 506
+    },
+    {
+      "epoch": 0.10790103750997605,
+      "grad_norm": 2.515093734898674,
+      "learning_rate": 3.595744680851064e-05,
+      "loss": 5.8134,
+      "step": 507
+    },
+    {
+      "epoch": 0.10811386006916733,
+      "grad_norm": 1.610740883276429,
+      "learning_rate": 3.602836879432624e-05,
+      "loss": 5.8381,
+      "step": 508
+    },
+    {
+      "epoch": 0.10832668262835861,
+      "grad_norm": 2.585512720686067,
+      "learning_rate": 3.609929078014185e-05,
+      "loss": 5.834,
+      "step": 509
+    },
+    {
+      "epoch": 0.10853950518754989,
+      "grad_norm": 1.860885872107565,
+      "learning_rate": 3.617021276595745e-05,
+      "loss": 5.8266,
+      "step": 510
+    },
+    {
+      "epoch": 0.10875232774674115,
+      "grad_norm": 2.403455662143659,
+      "learning_rate": 3.6241134751773054e-05,
+      "loss": 5.7979,
+      "step": 511
+    },
+    {
+      "epoch": 0.10896515030593243,
+      "grad_norm": 2.517788743114764,
+      "learning_rate": 3.6312056737588654e-05,
+      "loss": 5.8497,
+      "step": 512
+    },
+    {
+      "epoch": 0.1091779728651237,
+      "grad_norm": 2.180958662995085,
+      "learning_rate": 3.638297872340426e-05,
+      "loss": 5.8112,
+      "step": 513
+    },
+    {
+      "epoch": 0.10939079542431497,
+      "grad_norm": 2.0072234854820517,
+      "learning_rate": 3.645390070921986e-05,
+      "loss": 5.81,
+      "step": 514
+    },
+    {
+      "epoch": 0.10960361798350625,
+      "grad_norm": 1.9781595882798173,
+      "learning_rate": 3.6524822695035466e-05,
+      "loss": 5.8253,
+      "step": 515
+    },
+    {
+      "epoch": 0.10981644054269753,
+      "grad_norm": 2.3040977268829748,
+      "learning_rate": 3.6595744680851066e-05,
+      "loss": 5.8205,
+      "step": 516
+    },
+    {
+      "epoch": 0.1100292631018888,
+      "grad_norm": 2.350840882385147,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 5.7817,
+      "step": 517
+    },
+    {
+      "epoch": 0.11024208566108007,
+      "grad_norm": 1.8496042014483394,
+      "learning_rate": 3.673758865248227e-05,
+      "loss": 5.7871,
+      "step": 518
+    },
+    {
+      "epoch": 0.11045490822027135,
+      "grad_norm": 2.1213708267658427,
+      "learning_rate": 3.680851063829787e-05,
+      "loss": 5.7663,
+      "step": 519
+    },
+    {
+      "epoch": 0.11066773077946263,
+      "grad_norm": 1.9614978382328487,
+      "learning_rate": 3.687943262411347e-05,
+      "loss": 5.7878,
+      "step": 520
+    },
+    {
+      "epoch": 0.1108805533386539,
+      "grad_norm": 2.4585467056591472,
+      "learning_rate": 3.695035460992908e-05,
+      "loss": 5.7948,
+      "step": 521
+    },
+    {
+      "epoch": 0.11109337589784517,
+      "grad_norm": 2.0391605436378724,
+      "learning_rate": 3.702127659574468e-05,
+      "loss": 5.8055,
+      "step": 522
+    },
+    {
+      "epoch": 0.11130619845703645,
+      "grad_norm": 1.8928014560302537,
+      "learning_rate": 3.7092198581560284e-05,
+      "loss": 5.726,
+      "step": 523
+    },
+    {
+      "epoch": 0.11151902101622772,
+      "grad_norm": 2.655781278055372,
+      "learning_rate": 3.716312056737588e-05,
+      "loss": 5.7656,
+      "step": 524
+    },
+    {
+      "epoch": 0.11173184357541899,
+      "grad_norm": 1.7624698033031465,
+      "learning_rate": 3.723404255319149e-05,
+      "loss": 5.7587,
+      "step": 525
+    },
+    {
+      "epoch": 0.11194466613461027,
+      "grad_norm": 3.411338400221761,
+      "learning_rate": 3.730496453900709e-05,
+      "loss": 5.8143,
+      "step": 526
+    },
+    {
+      "epoch": 0.11215748869380154,
+      "grad_norm": 1.4344535008145765,
+      "learning_rate": 3.7375886524822696e-05,
+      "loss": 5.7584,
+      "step": 527
+    },
+    {
+      "epoch": 0.11237031125299282,
+      "grad_norm": 3.616094985736142,
+      "learning_rate": 3.7446808510638295e-05,
+      "loss": 5.8138,
+      "step": 528
+    },
+    {
+      "epoch": 0.11258313381218409,
+      "grad_norm": 2.6588368574920724,
+      "learning_rate": 3.75177304964539e-05,
+      "loss": 5.7576,
+      "step": 529
+    },
+    {
+      "epoch": 0.11279595637137536,
+      "grad_norm": 2.5145512401122554,
+      "learning_rate": 3.75886524822695e-05,
+      "loss": 5.7766,
+      "step": 530
+    },
+    {
+      "epoch": 0.11300877893056664,
+      "grad_norm": 3.1213460314825765,
+      "learning_rate": 3.765957446808511e-05,
+      "loss": 5.7978,
+      "step": 531
+    },
+    {
+      "epoch": 0.11322160148975792,
+      "grad_norm": 2.099501842604347,
+      "learning_rate": 3.773049645390071e-05,
+      "loss": 5.7783,
+      "step": 532
+    },
+    {
+      "epoch": 0.11343442404894918,
+      "grad_norm": 3.1152787741824066,
+      "learning_rate": 3.7801418439716314e-05,
+      "loss": 5.7668,
+      "step": 533
+    },
+    {
+      "epoch": 0.11364724660814046,
+      "grad_norm": 2.43648562459557,
+      "learning_rate": 3.787234042553191e-05,
+      "loss": 5.7719,
+      "step": 534
+    },
+    {
+      "epoch": 0.11386006916733174,
+      "grad_norm": 2.5061705915543353,
+      "learning_rate": 3.794326241134752e-05,
+      "loss": 5.727,
+      "step": 535
+    },
+    {
+      "epoch": 0.11407289172652302,
+      "grad_norm": 1.8588581828859954,
+      "learning_rate": 3.801418439716312e-05,
+      "loss": 5.766,
+      "step": 536
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 2.5080373568318173,
+      "learning_rate": 3.8085106382978726e-05,
+      "loss": 5.7274,
+      "step": 537
+    },
+    {
+      "epoch": 0.11449853684490556,
+      "grad_norm": 1.5221193463823146,
+      "learning_rate": 3.8156028368794325e-05,
+      "loss": 5.7037,
+      "step": 538
+    },
+    {
+      "epoch": 0.11471135940409684,
+      "grad_norm": 3.220476377977812,
+      "learning_rate": 3.822695035460993e-05,
+      "loss": 5.7637,
+      "step": 539
+    },
+    {
+      "epoch": 0.1149241819632881,
+      "grad_norm": 2.226966762946274,
+      "learning_rate": 3.829787234042553e-05,
+      "loss": 5.7461,
+      "step": 540
+    },
+    {
+      "epoch": 0.11513700452247938,
+      "grad_norm": 2.401257854215674,
+      "learning_rate": 3.836879432624114e-05,
+      "loss": 5.7352,
+      "step": 541
+    },
+    {
+      "epoch": 0.11534982708167066,
+      "grad_norm": 2.0658393915163207,
+      "learning_rate": 3.843971631205674e-05,
+      "loss": 5.6811,
+      "step": 542
+    },
+    {
+      "epoch": 0.11556264964086194,
+      "grad_norm": 1.6314619581171923,
+      "learning_rate": 3.8510638297872344e-05,
+      "loss": 5.6561,
+      "step": 543
+    },
+    {
+      "epoch": 0.1157754722000532,
+      "grad_norm": 2.494557987517589,
+      "learning_rate": 3.858156028368794e-05,
+      "loss": 5.684,
+      "step": 544
+    },
+    {
+      "epoch": 0.11598829475924448,
+      "grad_norm": 1.8197034946821167,
+      "learning_rate": 3.865248226950355e-05,
+      "loss": 5.7063,
+      "step": 545
+    },
+    {
+      "epoch": 0.11620111731843576,
+      "grad_norm": 2.656891827776325,
+      "learning_rate": 3.872340425531915e-05,
+      "loss": 5.7459,
+      "step": 546
+    },
+    {
+      "epoch": 0.11641393987762703,
+      "grad_norm": 2.444522907614219,
+      "learning_rate": 3.8794326241134756e-05,
+      "loss": 5.7058,
+      "step": 547
+    },
+    {
+      "epoch": 0.1166267624368183,
+      "grad_norm": 2.0322687856204964,
+      "learning_rate": 3.8865248226950355e-05,
+      "loss": 5.7112,
+      "step": 548
+    },
+    {
+      "epoch": 0.11683958499600958,
+      "grad_norm": 2.533337307466386,
+      "learning_rate": 3.893617021276596e-05,
+      "loss": 5.7088,
+      "step": 549
+    },
+    {
+      "epoch": 0.11705240755520085,
+      "grad_norm": 2.126147755517944,
+      "learning_rate": 3.900709219858156e-05,
+      "loss": 5.731,
+      "step": 550
+    },
+    {
+      "epoch": 0.11726523011439212,
+      "grad_norm": 2.7842067758264504,
+      "learning_rate": 3.907801418439717e-05,
+      "loss": 5.6734,
+      "step": 551
+    },
+    {
+      "epoch": 0.1174780526735834,
+      "grad_norm": 1.8773703988662098,
+      "learning_rate": 3.914893617021277e-05,
+      "loss": 5.7047,
+      "step": 552
+    },
+    {
+      "epoch": 0.11769087523277467,
+      "grad_norm": 2.6233950482705453,
+      "learning_rate": 3.9219858156028374e-05,
+      "loss": 5.6778,
+      "step": 553
+    },
+    {
+      "epoch": 0.11790369779196595,
+      "grad_norm": 1.8337198696854609,
+      "learning_rate": 3.929078014184397e-05,
+      "loss": 5.7274,
+      "step": 554
+    },
+    {
+      "epoch": 0.11811652035115722,
+      "grad_norm": 3.1214083112214106,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 5.703,
+      "step": 555
+    },
+    {
+      "epoch": 0.1183293429103485,
+      "grad_norm": 1.9470339524162625,
+      "learning_rate": 3.943262411347518e-05,
+      "loss": 5.6699,
+      "step": 556
+    },
+    {
+      "epoch": 0.11854216546953977,
+      "grad_norm": 2.5454366623764884,
+      "learning_rate": 3.9503546099290786e-05,
+      "loss": 5.7193,
+      "step": 557
+    },
+    {
+      "epoch": 0.11875498802873105,
+      "grad_norm": 2.7658018397810977,
+      "learning_rate": 3.9574468085106385e-05,
+      "loss": 5.7116,
+      "step": 558
+    },
+    {
+      "epoch": 0.11896781058792232,
+      "grad_norm": 2.047246898434044,
+      "learning_rate": 3.964539007092199e-05,
+      "loss": 5.6446,
+      "step": 559
+    },
+    {
+      "epoch": 0.1191806331471136,
+      "grad_norm": 2.4855864004264663,
+      "learning_rate": 3.971631205673759e-05,
+      "loss": 5.6801,
+      "step": 560
+    },
+    {
+      "epoch": 0.11939345570630487,
+      "grad_norm": 2.7910793129933005,
+      "learning_rate": 3.97872340425532e-05,
+      "loss": 5.6995,
+      "step": 561
+    },
+    {
+      "epoch": 0.11960627826549615,
+      "grad_norm": 2.3289460771203756,
+      "learning_rate": 3.98581560283688e-05,
+      "loss": 5.6534,
+      "step": 562
+    },
+    {
+      "epoch": 0.11981910082468741,
+      "grad_norm": 2.0400069351309247,
+      "learning_rate": 3.9929078014184404e-05,
+      "loss": 5.6963,
+      "step": 563
+    },
+    {
+      "epoch": 0.12003192338387869,
+      "grad_norm": 1.701554477270477,
+      "learning_rate": 4e-05,
+      "loss": 5.6458,
+      "step": 564
+    },
+    {
+      "epoch": 0.12024474594306997,
+      "grad_norm": 1.9851362275488094,
+      "learning_rate": 4.007092198581561e-05,
+      "loss": 5.656,
+      "step": 565
+    },
+    {
+      "epoch": 0.12045756850226123,
+      "grad_norm": 2.6155610392739455,
+      "learning_rate": 4.014184397163121e-05,
+      "loss": 5.6735,
+      "step": 566
+    },
+    {
+      "epoch": 0.12067039106145251,
+      "grad_norm": 2.1422588457831564,
+      "learning_rate": 4.0212765957446816e-05,
+      "loss": 5.6913,
+      "step": 567
+    },
+    {
+      "epoch": 0.12088321362064379,
+      "grad_norm": 2.2143064867347015,
+      "learning_rate": 4.0283687943262415e-05,
+      "loss": 5.6718,
+      "step": 568
+    },
+    {
+      "epoch": 0.12109603617983507,
+      "grad_norm": 2.629373734407472,
+      "learning_rate": 4.035460992907802e-05,
+      "loss": 5.661,
+      "step": 569
+    },
+    {
+      "epoch": 0.12130885873902633,
+      "grad_norm": 2.6209653390775522,
+      "learning_rate": 4.0425531914893614e-05,
+      "loss": 5.6432,
+      "step": 570
+    },
+    {
+      "epoch": 0.12152168129821761,
+      "grad_norm": 1.8541362632382783,
+      "learning_rate": 4.049645390070922e-05,
+      "loss": 5.6417,
+      "step": 571
+    },
+    {
+      "epoch": 0.12173450385740889,
+      "grad_norm": 2.251240210016008,
+      "learning_rate": 4.056737588652482e-05,
+      "loss": 5.6237,
+      "step": 572
+    },
+    {
+      "epoch": 0.12194732641660017,
+      "grad_norm": 2.124900958923251,
+      "learning_rate": 4.063829787234043e-05,
+      "loss": 5.6337,
+      "step": 573
+    },
+    {
+      "epoch": 0.12216014897579143,
+      "grad_norm": 2.66071511052287,
+      "learning_rate": 4.0709219858156026e-05,
+      "loss": 5.6732,
+      "step": 574
+    },
+    {
+      "epoch": 0.12237297153498271,
+      "grad_norm": 1.6682993381999567,
+      "learning_rate": 4.078014184397163e-05,
+      "loss": 5.6412,
+      "step": 575
+    },
+    {
+      "epoch": 0.12258579409417399,
+      "grad_norm": 2.002815533274544,
+      "learning_rate": 4.085106382978723e-05,
+      "loss": 5.6101,
+      "step": 576
+    },
+    {
+      "epoch": 0.12279861665336525,
+      "grad_norm": 2.0918574964051064,
+      "learning_rate": 4.092198581560284e-05,
+      "loss": 5.6255,
+      "step": 577
+    },
+    {
+      "epoch": 0.12301143921255653,
+      "grad_norm": 3.0772366215539706,
+      "learning_rate": 4.099290780141844e-05,
+      "loss": 5.6291,
+      "step": 578
+    },
+    {
+      "epoch": 0.1232242617717478,
+      "grad_norm": 1.828602603126976,
+      "learning_rate": 4.1063829787234045e-05,
+      "loss": 5.6067,
+      "step": 579
+    },
+    {
+      "epoch": 0.12343708433093908,
+      "grad_norm": 2.1975648067405675,
+      "learning_rate": 4.1134751773049644e-05,
+      "loss": 5.6829,
+      "step": 580
+    },
+    {
+      "epoch": 0.12364990689013035,
+      "grad_norm": 2.548228113959805,
+      "learning_rate": 4.120567375886525e-05,
+      "loss": 5.6215,
+      "step": 581
+    },
+    {
+      "epoch": 0.12386272944932163,
+      "grad_norm": 1.5577461239355042,
+      "learning_rate": 4.127659574468085e-05,
+      "loss": 5.5777,
+      "step": 582
+    },
+    {
+      "epoch": 0.1240755520085129,
+      "grad_norm": 2.3462360953419754,
+      "learning_rate": 4.134751773049646e-05,
+      "loss": 5.6577,
+      "step": 583
+    },
+    {
+      "epoch": 0.12428837456770418,
+      "grad_norm": 2.4830492199803715,
+      "learning_rate": 4.1418439716312056e-05,
+      "loss": 5.5994,
+      "step": 584
+    },
+    {
+      "epoch": 0.12450119712689545,
+      "grad_norm": 1.7829567619451072,
+      "learning_rate": 4.148936170212766e-05,
+      "loss": 5.5297,
+      "step": 585
+    },
+    {
+      "epoch": 0.12471401968608672,
+      "grad_norm": 2.901711735434022,
+      "learning_rate": 4.156028368794326e-05,
+      "loss": 5.6007,
+      "step": 586
+    },
+    {
+      "epoch": 0.124926842245278,
+      "grad_norm": 1.6266557722629365,
+      "learning_rate": 4.163120567375887e-05,
+      "loss": 5.61,
+      "step": 587
+    },
+    {
+      "epoch": 0.12513966480446928,
+      "grad_norm": 2.3214839537673573,
+      "learning_rate": 4.170212765957447e-05,
+      "loss": 5.5897,
+      "step": 588
+    },
+    {
+      "epoch": 0.12535248736366056,
+      "grad_norm": 1.983229432326336,
+      "learning_rate": 4.1773049645390075e-05,
+      "loss": 5.5739,
+      "step": 589
+    },
+    {
+      "epoch": 0.1255653099228518,
+      "grad_norm": 2.6575657228793603,
+      "learning_rate": 4.1843971631205674e-05,
+      "loss": 5.5818,
+      "step": 590
+    },
+    {
+      "epoch": 0.1257781324820431,
+      "grad_norm": 1.711320362952776,
+      "learning_rate": 4.191489361702128e-05,
+      "loss": 5.5876,
+      "step": 591
+    },
+    {
+      "epoch": 0.12599095504123436,
+      "grad_norm": 2.4033562589117445,
+      "learning_rate": 4.198581560283688e-05,
+      "loss": 5.5665,
+      "step": 592
+    },
+    {
+      "epoch": 0.12620377760042564,
+      "grad_norm": 2.61206137455194,
+      "learning_rate": 4.205673758865249e-05,
+      "loss": 5.6356,
+      "step": 593
+    },
+    {
+      "epoch": 0.12641660015961692,
+      "grad_norm": 2.6638780909106945,
+      "learning_rate": 4.2127659574468086e-05,
+      "loss": 5.6017,
+      "step": 594
+    },
+    {
+      "epoch": 0.1266294227188082,
+      "grad_norm": 2.3863398120773596,
+      "learning_rate": 4.219858156028369e-05,
+      "loss": 5.5838,
+      "step": 595
+    },
+    {
+      "epoch": 0.12684224527799948,
+      "grad_norm": 2.0241883469354054,
+      "learning_rate": 4.226950354609929e-05,
+      "loss": 5.5731,
+      "step": 596
+    },
+    {
+      "epoch": 0.12705506783719075,
+      "grad_norm": 2.717343129208566,
+      "learning_rate": 4.23404255319149e-05,
+      "loss": 5.591,
+      "step": 597
+    },
+    {
+      "epoch": 0.127267890396382,
+      "grad_norm": 1.5997417709182689,
+      "learning_rate": 4.24113475177305e-05,
+      "loss": 5.5563,
+      "step": 598
+    },
+    {
+      "epoch": 0.12748071295557328,
+      "grad_norm": 2.4226050612132903,
+      "learning_rate": 4.2482269503546105e-05,
+      "loss": 5.567,
+      "step": 599
+    },
+    {
+      "epoch": 0.12769353551476456,
+      "grad_norm": 2.3903443950161223,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 5.5449,
+      "step": 600
+    },
+    {
+      "epoch": 0.12790635807395584,
+      "grad_norm": 1.7409833855895869,
+      "learning_rate": 4.262411347517731e-05,
+      "loss": 5.5275,
+      "step": 601
+    },
+    {
+      "epoch": 0.12811918063314712,
+      "grad_norm": 3.346109109061518,
+      "learning_rate": 4.269503546099291e-05,
+      "loss": 5.5858,
+      "step": 602
+    },
+    {
+      "epoch": 0.1283320031923384,
+      "grad_norm": 2.6279486018296585,
+      "learning_rate": 4.276595744680851e-05,
+      "loss": 5.5542,
+      "step": 603
+    },
+    {
+      "epoch": 0.12854482575152967,
+      "grad_norm": 2.280875760556807,
+      "learning_rate": 4.2836879432624116e-05,
+      "loss": 5.5863,
+      "step": 604
+    },
+    {
+      "epoch": 0.12875764831072092,
+      "grad_norm": 2.6936553816117024,
+      "learning_rate": 4.2907801418439716e-05,
+      "loss": 5.5245,
+      "step": 605
+    },
+    {
+      "epoch": 0.1289704708699122,
+      "grad_norm": 1.8628870209298674,
+      "learning_rate": 4.297872340425532e-05,
+      "loss": 5.5671,
+      "step": 606
+    },
+    {
+      "epoch": 0.12918329342910348,
+      "grad_norm": 2.913276685833457,
+      "learning_rate": 4.304964539007092e-05,
+      "loss": 5.5491,
+      "step": 607
+    },
+    {
+      "epoch": 0.12939611598829476,
+      "grad_norm": 2.1587703241608986,
+      "learning_rate": 4.312056737588653e-05,
+      "loss": 5.4848,
+      "step": 608
+    },
+    {
+      "epoch": 0.12960893854748604,
+      "grad_norm": 2.381136467901557,
+      "learning_rate": 4.319148936170213e-05,
+      "loss": 5.5386,
+      "step": 609
+    },
+    {
+      "epoch": 0.1298217611066773,
+      "grad_norm": 2.462198915955899,
+      "learning_rate": 4.3262411347517734e-05,
+      "loss": 5.5109,
+      "step": 610
+    },
+    {
+      "epoch": 0.1300345836658686,
+      "grad_norm": 2.2700190502399398,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 5.5063,
+      "step": 611
+    },
+    {
+      "epoch": 0.13024740622505987,
+      "grad_norm": 2.0952075507295236,
+      "learning_rate": 4.340425531914894e-05,
+      "loss": 5.5135,
+      "step": 612
+    },
+    {
+      "epoch": 0.13046022878425112,
+      "grad_norm": 1.9868921835003832,
+      "learning_rate": 4.347517730496454e-05,
+      "loss": 5.5404,
+      "step": 613
+    },
+    {
+      "epoch": 0.1306730513434424,
+      "grad_norm": 3.036507304887253,
+      "learning_rate": 4.3546099290780146e-05,
+      "loss": 5.5671,
+      "step": 614
+    },
+    {
+      "epoch": 0.13088587390263368,
+      "grad_norm": 1.6361286079198474,
+      "learning_rate": 4.3617021276595746e-05,
+      "loss": 5.5421,
+      "step": 615
+    },
+    {
+      "epoch": 0.13109869646182495,
+      "grad_norm": 2.4829633750318965,
+      "learning_rate": 4.368794326241135e-05,
+      "loss": 5.4807,
+      "step": 616
+    },
+    {
+      "epoch": 0.13131151902101623,
+      "grad_norm": 1.8320391674820375,
+      "learning_rate": 4.375886524822695e-05,
+      "loss": 5.5018,
+      "step": 617
+    },
+    {
+      "epoch": 0.1315243415802075,
+      "grad_norm": 2.8948463973420817,
+      "learning_rate": 4.382978723404256e-05,
+      "loss": 5.5654,
+      "step": 618
+    },
+    {
+      "epoch": 0.1317371641393988,
+      "grad_norm": 1.8637301529401844,
+      "learning_rate": 4.390070921985816e-05,
+      "loss": 5.5462,
+      "step": 619
+    },
+    {
+      "epoch": 0.13194998669859004,
+      "grad_norm": 2.6095377625269,
+      "learning_rate": 4.3971631205673764e-05,
+      "loss": 5.4734,
+      "step": 620
+    },
+    {
+      "epoch": 0.13216280925778132,
+      "grad_norm": 2.487018780350672,
+      "learning_rate": 4.4042553191489364e-05,
+      "loss": 5.4905,
+      "step": 621
+    },
+    {
+      "epoch": 0.1323756318169726,
+      "grad_norm": 2.074956278456767,
+      "learning_rate": 4.411347517730497e-05,
+      "loss": 5.4825,
+      "step": 622
+    },
+    {
+      "epoch": 0.13258845437616387,
+      "grad_norm": 2.5100555728179637,
+      "learning_rate": 4.418439716312057e-05,
+      "loss": 5.4685,
+      "step": 623
+    },
+    {
+      "epoch": 0.13280127693535515,
+      "grad_norm": 2.036402786308918,
+      "learning_rate": 4.425531914893617e-05,
+      "loss": 5.4482,
+      "step": 624
+    },
+    {
+      "epoch": 0.13301409949454643,
+      "grad_norm": 1.8810058586046094,
+      "learning_rate": 4.432624113475177e-05,
+      "loss": 5.4798,
+      "step": 625
+    },
+    {
+      "epoch": 0.1332269220537377,
+      "grad_norm": 3.343997645424007,
+      "learning_rate": 4.4397163120567376e-05,
+      "loss": 5.512,
+      "step": 626
+    },
+    {
+      "epoch": 0.13343974461292896,
+      "grad_norm": 2.027183236729858,
+      "learning_rate": 4.4468085106382975e-05,
+      "loss": 5.5004,
+      "step": 627
+    },
+    {
+      "epoch": 0.13365256717212023,
+      "grad_norm": 2.338019688139561,
+      "learning_rate": 4.453900709219858e-05,
+      "loss": 5.4818,
+      "step": 628
+    },
+    {
+      "epoch": 0.1338653897313115,
+      "grad_norm": 2.3057677907611036,
+      "learning_rate": 4.460992907801418e-05,
+      "loss": 5.4943,
+      "step": 629
+    },
+    {
+      "epoch": 0.1340782122905028,
+      "grad_norm": 2.1200481500257284,
+      "learning_rate": 4.468085106382979e-05,
+      "loss": 5.4611,
+      "step": 630
+    },
+    {
+      "epoch": 0.13429103484969407,
+      "grad_norm": 2.263591962257459,
+      "learning_rate": 4.475177304964539e-05,
+      "loss": 5.4931,
+      "step": 631
+    },
+    {
+      "epoch": 0.13450385740888535,
+      "grad_norm": 2.295980632658951,
+      "learning_rate": 4.4822695035460994e-05,
+      "loss": 5.4898,
+      "step": 632
+    },
+    {
+      "epoch": 0.13471667996807662,
+      "grad_norm": 2.5683328865799346,
+      "learning_rate": 4.489361702127659e-05,
+      "loss": 5.4731,
+      "step": 633
+    },
+    {
+      "epoch": 0.1349295025272679,
+      "grad_norm": 1.7129161272405207,
+      "learning_rate": 4.49645390070922e-05,
+      "loss": 5.4869,
+      "step": 634
+    },
+    {
+      "epoch": 0.13514232508645915,
+      "grad_norm": 2.5361513168089096,
+      "learning_rate": 4.50354609929078e-05,
+      "loss": 5.4412,
+      "step": 635
+    },
+    {
+      "epoch": 0.13535514764565043,
+      "grad_norm": 2.623175319916039,
+      "learning_rate": 4.5106382978723406e-05,
+      "loss": 5.4543,
+      "step": 636
+    },
+    {
+      "epoch": 0.1355679702048417,
+      "grad_norm": 2.124318155902657,
+      "learning_rate": 4.5177304964539005e-05,
+      "loss": 5.4675,
+      "step": 637
+    },
+    {
+      "epoch": 0.135780792764033,
+      "grad_norm": 2.261822918454049,
+      "learning_rate": 4.524822695035461e-05,
+      "loss": 5.4084,
+      "step": 638
+    },
+    {
+      "epoch": 0.13599361532322427,
+      "grad_norm": 2.1507919704545273,
+      "learning_rate": 4.531914893617021e-05,
+      "loss": 5.4373,
+      "step": 639
+    },
+    {
+      "epoch": 0.13620643788241554,
+      "grad_norm": 2.447701969500142,
+      "learning_rate": 4.539007092198582e-05,
+      "loss": 5.4314,
+      "step": 640
+    },
+    {
+      "epoch": 0.13641926044160682,
+      "grad_norm": 1.8026407656921348,
+      "learning_rate": 4.546099290780142e-05,
+      "loss": 5.4017,
+      "step": 641
+    },
+    {
+      "epoch": 0.13663208300079807,
+      "grad_norm": 2.5250091199744444,
+      "learning_rate": 4.5531914893617024e-05,
+      "loss": 5.421,
+      "step": 642
+    },
+    {
+      "epoch": 0.13684490555998935,
+      "grad_norm": 1.9573242232255295,
+      "learning_rate": 4.560283687943262e-05,
+      "loss": 5.4629,
+      "step": 643
+    },
+    {
+      "epoch": 0.13705772811918063,
+      "grad_norm": 1.8831275027415533,
+      "learning_rate": 4.567375886524823e-05,
+      "loss": 5.4308,
+      "step": 644
+    },
+    {
+      "epoch": 0.1372705506783719,
+      "grad_norm": 2.679496254862922,
+      "learning_rate": 4.574468085106383e-05,
+      "loss": 5.4532,
+      "step": 645
+    },
+    {
+      "epoch": 0.13748337323756318,
+      "grad_norm": 1.8226532300404443,
+      "learning_rate": 4.5815602836879436e-05,
+      "loss": 5.4509,
+      "step": 646
+    },
+    {
+      "epoch": 0.13769619579675446,
+      "grad_norm": 2.1186135337300347,
+      "learning_rate": 4.5886524822695035e-05,
+      "loss": 5.4741,
+      "step": 647
+    },
+    {
+      "epoch": 0.13790901835594574,
+      "grad_norm": 1.8562668192603644,
+      "learning_rate": 4.595744680851064e-05,
+      "loss": 5.4357,
+      "step": 648
+    },
+    {
+      "epoch": 0.13812184091513702,
+      "grad_norm": 2.8038145289551246,
+      "learning_rate": 4.602836879432624e-05,
+      "loss": 5.4195,
+      "step": 649
+    },
+    {
+      "epoch": 0.13833466347432827,
+      "grad_norm": 1.7010096312299654,
+      "learning_rate": 4.609929078014185e-05,
+      "loss": 5.4267,
+      "step": 650
+    },
+    {
+      "epoch": 0.13854748603351955,
+      "grad_norm": 2.720225487506592,
+      "learning_rate": 4.617021276595745e-05,
+      "loss": 5.4064,
+      "step": 651
+    },
+    {
+      "epoch": 0.13876030859271082,
+      "grad_norm": 2.0383040147934373,
+      "learning_rate": 4.6241134751773054e-05,
+      "loss": 5.4174,
+      "step": 652
+    },
+    {
+      "epoch": 0.1389731311519021,
+      "grad_norm": 3.009364937830362,
+      "learning_rate": 4.631205673758865e-05,
+      "loss": 5.4231,
+      "step": 653
+    },
+    {
+      "epoch": 0.13918595371109338,
+      "grad_norm": 2.3058175896747963,
+      "learning_rate": 4.638297872340426e-05,
+      "loss": 5.4192,
+      "step": 654
+    },
+    {
+      "epoch": 0.13939877627028466,
+      "grad_norm": 1.8459248951030456,
+      "learning_rate": 4.645390070921986e-05,
+      "loss": 5.4311,
+      "step": 655
+    },
+    {
+      "epoch": 0.13961159882947594,
+      "grad_norm": 2.638701686044538,
+      "learning_rate": 4.6524822695035466e-05,
+      "loss": 5.4122,
+      "step": 656
+    },
+    {
+      "epoch": 0.13982442138866719,
+      "grad_norm": 2.0510386336192803,
+      "learning_rate": 4.6595744680851065e-05,
+      "loss": 5.3838,
+      "step": 657
+    },
+    {
+      "epoch": 0.14003724394785846,
+      "grad_norm": 2.5030880414713015,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 5.4407,
+      "step": 658
+    },
+    {
+      "epoch": 0.14025006650704974,
+      "grad_norm": 2.3018692428946577,
+      "learning_rate": 4.673758865248227e-05,
+      "loss": 5.3822,
+      "step": 659
+    },
+    {
+      "epoch": 0.14046288906624102,
+      "grad_norm": 2.657245679575528,
+      "learning_rate": 4.680851063829788e-05,
+      "loss": 5.3892,
+      "step": 660
+    },
+    {
+      "epoch": 0.1406757116254323,
+      "grad_norm": 2.448766274722233,
+      "learning_rate": 4.687943262411348e-05,
+      "loss": 5.4089,
+      "step": 661
+    },
+    {
+      "epoch": 0.14088853418462358,
+      "grad_norm": 2.409030557092896,
+      "learning_rate": 4.6950354609929084e-05,
+      "loss": 5.4118,
+      "step": 662
+    },
+    {
+      "epoch": 0.14110135674381485,
+      "grad_norm": 1.8855221463300356,
+      "learning_rate": 4.702127659574468e-05,
+      "loss": 5.3826,
+      "step": 663
+    },
+    {
+      "epoch": 0.14131417930300613,
+      "grad_norm": 2.5510875259415724,
+      "learning_rate": 4.709219858156029e-05,
+      "loss": 5.3881,
+      "step": 664
+    },
+    {
+      "epoch": 0.14152700186219738,
+      "grad_norm": 2.2226075392395903,
+      "learning_rate": 4.716312056737589e-05,
+      "loss": 5.366,
+      "step": 665
+    },
+    {
+      "epoch": 0.14173982442138866,
+      "grad_norm": 2.3406166652132456,
+      "learning_rate": 4.7234042553191496e-05,
+      "loss": 5.3794,
+      "step": 666
+    },
+    {
+      "epoch": 0.14195264698057994,
+      "grad_norm": 1.9869693405796383,
+      "learning_rate": 4.7304964539007095e-05,
+      "loss": 5.3974,
+      "step": 667
+    },
+    {
+      "epoch": 0.14216546953977122,
+      "grad_norm": 2.6343219754613854,
+      "learning_rate": 4.73758865248227e-05,
+      "loss": 5.3356,
+      "step": 668
+    },
+    {
+      "epoch": 0.1423782920989625,
+      "grad_norm": 2.0922394465694856,
+      "learning_rate": 4.74468085106383e-05,
+      "loss": 5.3818,
+      "step": 669
+    },
+    {
+      "epoch": 0.14259111465815377,
+      "grad_norm": 2.0643498433347363,
+      "learning_rate": 4.751773049645391e-05,
+      "loss": 5.3462,
+      "step": 670
+    },
+    {
+      "epoch": 0.14280393721734505,
+      "grad_norm": 3.083833963887123,
+      "learning_rate": 4.758865248226951e-05,
+      "loss": 5.4269,
+      "step": 671
+    },
+    {
+      "epoch": 0.1430167597765363,
+      "grad_norm": 2.1123205222601618,
+      "learning_rate": 4.7659574468085114e-05,
+      "loss": 5.3794,
+      "step": 672
+    },
+    {
+      "epoch": 0.14322958233572758,
+      "grad_norm": 3.0249387981175793,
+      "learning_rate": 4.773049645390071e-05,
+      "loss": 5.4028,
+      "step": 673
+    },
+    {
+      "epoch": 0.14344240489491886,
+      "grad_norm": 1.697668683971371,
+      "learning_rate": 4.780141843971632e-05,
+      "loss": 5.3364,
+      "step": 674
+    },
+    {
+      "epoch": 0.14365522745411013,
+      "grad_norm": 2.5512005676892255,
+      "learning_rate": 4.787234042553192e-05,
+      "loss": 5.3753,
+      "step": 675
+    },
+    {
+      "epoch": 0.1438680500133014,
+      "grad_norm": 2.287898571822724,
+      "learning_rate": 4.7943262411347526e-05,
+      "loss": 5.3807,
+      "step": 676
+    },
+    {
+      "epoch": 0.1440808725724927,
+      "grad_norm": 2.8703667654143517,
+      "learning_rate": 4.8014184397163125e-05,
+      "loss": 5.3893,
+      "step": 677
+    },
+    {
+      "epoch": 0.14429369513168397,
+      "grad_norm": 2.2006603626162327,
+      "learning_rate": 4.8085106382978725e-05,
+      "loss": 5.3356,
+      "step": 678
+    },
+    {
+      "epoch": 0.14450651769087522,
+      "grad_norm": 2.4548287720536734,
+      "learning_rate": 4.8156028368794324e-05,
+      "loss": 5.3365,
+      "step": 679
+    },
+    {
+      "epoch": 0.1447193402500665,
+      "grad_norm": 2.266650338676386,
+      "learning_rate": 4.822695035460993e-05,
+      "loss": 5.3097,
+      "step": 680
+    },
+    {
+      "epoch": 0.14493216280925778,
+      "grad_norm": 1.982935222519939,
+      "learning_rate": 4.829787234042553e-05,
+      "loss": 5.3582,
+      "step": 681
+    },
+    {
+      "epoch": 0.14514498536844905,
+      "grad_norm": 2.292675688959624,
+      "learning_rate": 4.836879432624114e-05,
+      "loss": 5.3299,
+      "step": 682
+    },
+    {
+      "epoch": 0.14535780792764033,
+      "grad_norm": 2.492353698740901,
+      "learning_rate": 4.8439716312056736e-05,
+      "loss": 5.3223,
+      "step": 683
+    },
+    {
+      "epoch": 0.1455706304868316,
+      "grad_norm": 1.9209522691836325,
+      "learning_rate": 4.851063829787234e-05,
+      "loss": 5.3516,
+      "step": 684
+    },
+    {
+      "epoch": 0.1457834530460229,
+      "grad_norm": 2.840332179549403,
+      "learning_rate": 4.858156028368794e-05,
+      "loss": 5.3195,
+      "step": 685
+    },
+    {
+      "epoch": 0.14599627560521417,
+      "grad_norm": 1.6461372577425581,
+      "learning_rate": 4.865248226950355e-05,
+      "loss": 5.3735,
+      "step": 686
+    },
+    {
+      "epoch": 0.14620909816440542,
+      "grad_norm": 2.5578441931432803,
+      "learning_rate": 4.872340425531915e-05,
+      "loss": 5.3078,
+      "step": 687
+    },
+    {
+      "epoch": 0.1464219207235967,
+      "grad_norm": 2.170961498247686,
+      "learning_rate": 4.8794326241134755e-05,
+      "loss": 5.3922,
+      "step": 688
+    },
+    {
+      "epoch": 0.14663474328278797,
+      "grad_norm": 2.0339324252860043,
+      "learning_rate": 4.8865248226950354e-05,
+      "loss": 5.3042,
+      "step": 689
+    },
+    {
+      "epoch": 0.14684756584197925,
+      "grad_norm": 2.1772065180691356,
+      "learning_rate": 4.893617021276596e-05,
+      "loss": 5.3291,
+      "step": 690
+    },
+    {
+      "epoch": 0.14706038840117053,
+      "grad_norm": 1.8173165036104597,
+      "learning_rate": 4.900709219858156e-05,
+      "loss": 5.3092,
+      "step": 691
+    },
+    {
+      "epoch": 0.1472732109603618,
+      "grad_norm": 2.824103950037195,
+      "learning_rate": 4.907801418439717e-05,
+      "loss": 5.3054,
+      "step": 692
+    },
+    {
+      "epoch": 0.14748603351955308,
+      "grad_norm": 1.6872265805822193,
+      "learning_rate": 4.9148936170212766e-05,
+      "loss": 5.3185,
+      "step": 693
+    },
+    {
+      "epoch": 0.14769885607874433,
+      "grad_norm": 2.584234382031475,
+      "learning_rate": 4.921985815602837e-05,
+      "loss": 5.3067,
+      "step": 694
+    },
+    {
+      "epoch": 0.1479116786379356,
+      "grad_norm": 2.4319226342032016,
+      "learning_rate": 4.929078014184397e-05,
+      "loss": 5.2994,
+      "step": 695
+    },
+    {
+      "epoch": 0.1481245011971269,
+      "grad_norm": 2.090091636152284,
+      "learning_rate": 4.936170212765958e-05,
+      "loss": 5.3326,
+      "step": 696
+    },
+    {
+      "epoch": 0.14833732375631817,
+      "grad_norm": 1.783226067690749,
+      "learning_rate": 4.943262411347518e-05,
+      "loss": 5.276,
+      "step": 697
+    },
+    {
+      "epoch": 0.14855014631550945,
+      "grad_norm": 3.255843401718798,
+      "learning_rate": 4.9503546099290785e-05,
+      "loss": 5.3251,
+      "step": 698
+    },
+    {
+      "epoch": 0.14876296887470072,
+      "grad_norm": 1.7969178459988682,
+      "learning_rate": 4.9574468085106384e-05,
+      "loss": 5.3618,
+      "step": 699
+    },
+    {
+      "epoch": 0.148975791433892,
+      "grad_norm": 2.633384515971593,
+      "learning_rate": 4.964539007092199e-05,
+      "loss": 5.3072,
+      "step": 700
+    },
+    {
+      "epoch": 0.14918861399308328,
+      "grad_norm": 2.2634843659642736,
+      "learning_rate": 4.971631205673759e-05,
+      "loss": 5.3501,
+      "step": 701
+    },
+    {
+      "epoch": 0.14940143655227453,
+      "grad_norm": 2.651874486822907,
+      "learning_rate": 4.97872340425532e-05,
+      "loss": 5.3447,
+      "step": 702
+    },
+    {
+      "epoch": 0.1496142591114658,
+      "grad_norm": 2.6489422399194025,
+      "learning_rate": 4.9858156028368796e-05,
+      "loss": 5.3407,
+      "step": 703
+    },
+    {
+      "epoch": 0.1498270816706571,
+      "grad_norm": 2.4012908057835656,
+      "learning_rate": 4.99290780141844e-05,
+      "loss": 5.3165,
+      "step": 704
+    },
+    {
+      "epoch": 0.15003990422984836,
+      "grad_norm": 1.7820402956012007,
+      "learning_rate": 5e-05,
+      "loss": 5.3012,
+      "step": 705
+    },
+    {
+      "epoch": 0.15025272678903964,
+      "grad_norm": 2.3955763579667293,
+      "learning_rate": 5.00709219858156e-05,
+      "loss": 5.2932,
+      "step": 706
+    },
+    {
+      "epoch": 0.15046554934823092,
+      "grad_norm": 2.300503048501749,
+      "learning_rate": 5.014184397163121e-05,
+      "loss": 5.3212,
+      "step": 707
+    },
+    {
+      "epoch": 0.1506783719074222,
+      "grad_norm": 2.0660442755635855,
+      "learning_rate": 5.021276595744681e-05,
+      "loss": 5.2687,
+      "step": 708
+    },
+    {
+      "epoch": 0.15089119446661345,
+      "grad_norm": 2.1059808244639076,
+      "learning_rate": 5.0283687943262414e-05,
+      "loss": 5.298,
+      "step": 709
+    },
+    {
+      "epoch": 0.15110401702580473,
+      "grad_norm": 2.160072039011158,
+      "learning_rate": 5.0354609929078014e-05,
+      "loss": 5.242,
+      "step": 710
+    },
+    {
+      "epoch": 0.151316839584996,
+      "grad_norm": 2.387736347898628,
+      "learning_rate": 5.042553191489362e-05,
+      "loss": 5.3047,
+      "step": 711
+    },
+    {
+      "epoch": 0.15152966214418728,
+      "grad_norm": 2.01055786980767,
+      "learning_rate": 5.049645390070922e-05,
+      "loss": 5.2597,
+      "step": 712
+    },
+    {
+      "epoch": 0.15174248470337856,
+      "grad_norm": 3.106134213810948,
+      "learning_rate": 5.0567375886524826e-05,
+      "loss": 5.2527,
+      "step": 713
+    },
+    {
+      "epoch": 0.15195530726256984,
+      "grad_norm": 1.8476084666092316,
+      "learning_rate": 5.0638297872340426e-05,
+      "loss": 5.2572,
+      "step": 714
+    },
+    {
+      "epoch": 0.15216812982176112,
+      "grad_norm": 2.992799874803428,
+      "learning_rate": 5.070921985815603e-05,
+      "loss": 5.2939,
+      "step": 715
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 2.2361746620117895,
+      "learning_rate": 5.078014184397163e-05,
+      "loss": 5.2743,
+      "step": 716
+    },
+    {
+      "epoch": 0.15259377494014364,
+      "grad_norm": 2.638386965479752,
+      "learning_rate": 5.085106382978724e-05,
+      "loss": 5.2669,
+      "step": 717
+    },
+    {
+      "epoch": 0.15280659749933492,
+      "grad_norm": 2.5921323943843224,
+      "learning_rate": 5.092198581560284e-05,
+      "loss": 5.2175,
+      "step": 718
+    },
+    {
+      "epoch": 0.1530194200585262,
+      "grad_norm": 1.8919617077439208,
+      "learning_rate": 5.0992907801418444e-05,
+      "loss": 5.2515,
+      "step": 719
+    },
+    {
+      "epoch": 0.15323224261771748,
+      "grad_norm": 3.0291009315207247,
+      "learning_rate": 5.1063829787234044e-05,
+      "loss": 5.2536,
+      "step": 720
+    },
+    {
+      "epoch": 0.15344506517690876,
+      "grad_norm": 1.9007242581658932,
+      "learning_rate": 5.113475177304965e-05,
+      "loss": 5.3008,
+      "step": 721
+    },
+    {
+      "epoch": 0.15365788773610004,
+      "grad_norm": 3.036011436673675,
+      "learning_rate": 5.120567375886525e-05,
+      "loss": 5.3043,
+      "step": 722
+    },
+    {
+      "epoch": 0.1538707102952913,
+      "grad_norm": 2.081392043998818,
+      "learning_rate": 5.1276595744680856e-05,
+      "loss": 5.2558,
+      "step": 723
+    },
+    {
+      "epoch": 0.15408353285448256,
+      "grad_norm": 2.35123897656844,
+      "learning_rate": 5.1347517730496456e-05,
+      "loss": 5.2774,
+      "step": 724
+    },
+    {
+      "epoch": 0.15429635541367384,
+      "grad_norm": 2.204038242248021,
+      "learning_rate": 5.141843971631206e-05,
+      "loss": 5.2554,
+      "step": 725
+    },
+    {
+      "epoch": 0.15450917797286512,
+      "grad_norm": 2.7104101760128043,
+      "learning_rate": 5.148936170212766e-05,
+      "loss": 5.2342,
+      "step": 726
+    },
+    {
+      "epoch": 0.1547220005320564,
+      "grad_norm": 1.8664468222328827,
+      "learning_rate": 5.156028368794327e-05,
+      "loss": 5.2462,
+      "step": 727
+    },
+    {
+      "epoch": 0.15493482309124768,
+      "grad_norm": 2.6141970824437974,
+      "learning_rate": 5.163120567375886e-05,
+      "loss": 5.2439,
+      "step": 728
+    },
+    {
+      "epoch": 0.15514764565043895,
+      "grad_norm": 2.5228593168805244,
+      "learning_rate": 5.1702127659574474e-05,
+      "loss": 5.2822,
+      "step": 729
+    },
+    {
+      "epoch": 0.15536046820963023,
+      "grad_norm": 2.0808558388718295,
+      "learning_rate": 5.177304964539007e-05,
+      "loss": 5.312,
+      "step": 730
+    },
+    {
+      "epoch": 0.15557329076882148,
+      "grad_norm": 2.4248472403975576,
+      "learning_rate": 5.184397163120568e-05,
+      "loss": 5.23,
+      "step": 731
+    },
+    {
+      "epoch": 0.15578611332801276,
+      "grad_norm": 2.309845034458759,
+      "learning_rate": 5.191489361702127e-05,
+      "loss": 5.2425,
+      "step": 732
+    },
+    {
+      "epoch": 0.15599893588720404,
+      "grad_norm": 2.4225441821084135,
+      "learning_rate": 5.1985815602836886e-05,
+      "loss": 5.2535,
+      "step": 733
+    },
+    {
+      "epoch": 0.15621175844639532,
+      "grad_norm": 2.252626487722863,
+      "learning_rate": 5.205673758865248e-05,
+      "loss": 5.2231,
+      "step": 734
+    },
+    {
+      "epoch": 0.1564245810055866,
+      "grad_norm": 2.432784648422067,
+      "learning_rate": 5.212765957446809e-05,
+      "loss": 5.2521,
+      "step": 735
+    },
+    {
+      "epoch": 0.15663740356477787,
+      "grad_norm": 2.2194572533911625,
+      "learning_rate": 5.2198581560283685e-05,
+      "loss": 5.1938,
+      "step": 736
+    },
+    {
+      "epoch": 0.15685022612396915,
+      "grad_norm": 2.5188150012732695,
+      "learning_rate": 5.22695035460993e-05,
+      "loss": 5.2504,
+      "step": 737
+    },
+    {
+      "epoch": 0.15706304868316043,
+      "grad_norm": 1.8893326801932147,
+      "learning_rate": 5.234042553191489e-05,
+      "loss": 5.2302,
+      "step": 738
+    },
+    {
+      "epoch": 0.15727587124235168,
+      "grad_norm": 2.641944238215526,
+      "learning_rate": 5.2411347517730504e-05,
+      "loss": 5.2042,
+      "step": 739
+    },
+    {
+      "epoch": 0.15748869380154296,
+      "grad_norm": 1.9651748275433583,
+      "learning_rate": 5.24822695035461e-05,
+      "loss": 5.237,
+      "step": 740
+    },
+    {
+      "epoch": 0.15770151636073423,
+      "grad_norm": 2.5583942267875117,
+      "learning_rate": 5.255319148936171e-05,
+      "loss": 5.2183,
+      "step": 741
+    },
+    {
+      "epoch": 0.1579143389199255,
+      "grad_norm": 1.9522181071716098,
+      "learning_rate": 5.26241134751773e-05,
+      "loss": 5.2319,
+      "step": 742
+    },
+    {
+      "epoch": 0.1581271614791168,
+      "grad_norm": 1.9792037565308387,
+      "learning_rate": 5.2695035460992916e-05,
+      "loss": 5.1933,
+      "step": 743
+    },
+    {
+      "epoch": 0.15833998403830807,
+      "grad_norm": 2.3349902828910496,
+      "learning_rate": 5.276595744680851e-05,
+      "loss": 5.2184,
+      "step": 744
+    },
+    {
+      "epoch": 0.15855280659749935,
+      "grad_norm": 2.479862241639395,
+      "learning_rate": 5.283687943262412e-05,
+      "loss": 5.2356,
+      "step": 745
+    },
+    {
+      "epoch": 0.1587656291566906,
+      "grad_norm": 2.3172623539722963,
+      "learning_rate": 5.2907801418439715e-05,
+      "loss": 5.2159,
+      "step": 746
+    },
+    {
+      "epoch": 0.15897845171588187,
+      "grad_norm": 2.2974013840524057,
+      "learning_rate": 5.297872340425533e-05,
+      "loss": 5.2098,
+      "step": 747
+    },
+    {
+      "epoch": 0.15919127427507315,
+      "grad_norm": 1.9941233474069722,
+      "learning_rate": 5.304964539007092e-05,
+      "loss": 5.1903,
+      "step": 748
+    },
+    {
+      "epoch": 0.15940409683426443,
+      "grad_norm": 2.893320619288781,
+      "learning_rate": 5.3120567375886534e-05,
+      "loss": 5.175,
+      "step": 749
+    },
+    {
+      "epoch": 0.1596169193934557,
+      "grad_norm": 2.340055902159033,
+      "learning_rate": 5.319148936170213e-05,
+      "loss": 5.2094,
+      "step": 750
+    },
+    {
+      "epoch": 0.159829741952647,
+      "grad_norm": 2.7153525284452913,
+      "learning_rate": 5.326241134751774e-05,
+      "loss": 5.1853,
+      "step": 751
+    },
+    {
+      "epoch": 0.16004256451183826,
+      "grad_norm": 1.5346465385894903,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 5.1675,
+      "step": 752
+    },
+    {
+      "epoch": 0.16025538707102954,
+      "grad_norm": 3.0768257062654607,
+      "learning_rate": 5.3404255319148946e-05,
+      "loss": 5.2456,
+      "step": 753
+    },
+    {
+      "epoch": 0.1604682096302208,
+      "grad_norm": 1.9462528807829407,
+      "learning_rate": 5.347517730496454e-05,
+      "loss": 5.182,
+      "step": 754
+    },
+    {
+      "epoch": 0.16068103218941207,
+      "grad_norm": 2.9017196038046635,
+      "learning_rate": 5.354609929078015e-05,
+      "loss": 5.243,
+      "step": 755
+    },
+    {
+      "epoch": 0.16089385474860335,
+      "grad_norm": 2.3127954444127656,
+      "learning_rate": 5.3617021276595745e-05,
+      "loss": 5.2508,
+      "step": 756
+    },
+    {
+      "epoch": 0.16110667730779463,
+      "grad_norm": 1.6076153949874434,
+      "learning_rate": 5.368794326241136e-05,
+      "loss": 5.2182,
+      "step": 757
+    },
+    {
+      "epoch": 0.1613194998669859,
+      "grad_norm": 3.19106979518661,
+      "learning_rate": 5.375886524822695e-05,
+      "loss": 5.1727,
+      "step": 758
+    },
+    {
+      "epoch": 0.16153232242617718,
+      "grad_norm": 2.1162648683331495,
+      "learning_rate": 5.3829787234042564e-05,
+      "loss": 5.1911,
+      "step": 759
+    },
+    {
+      "epoch": 0.16174514498536846,
+      "grad_norm": 2.1554675863359445,
+      "learning_rate": 5.390070921985816e-05,
+      "loss": 5.2016,
+      "step": 760
+    },
+    {
+      "epoch": 0.1619579675445597,
+      "grad_norm": 2.4691226746410657,
+      "learning_rate": 5.397163120567376e-05,
+      "loss": 5.2213,
+      "step": 761
+    },
+    {
+      "epoch": 0.162170790103751,
+      "grad_norm": 2.785452540051746,
+      "learning_rate": 5.404255319148936e-05,
+      "loss": 5.1786,
+      "step": 762
+    },
+    {
+      "epoch": 0.16238361266294227,
+      "grad_norm": 1.9812806251165131,
+      "learning_rate": 5.411347517730496e-05,
+      "loss": 5.1739,
+      "step": 763
+    },
+    {
+      "epoch": 0.16259643522213355,
+      "grad_norm": 2.417607804200543,
+      "learning_rate": 5.418439716312057e-05,
+      "loss": 5.1916,
+      "step": 764
+    },
+    {
+      "epoch": 0.16280925778132482,
+      "grad_norm": 2.4073449633431627,
+      "learning_rate": 5.425531914893617e-05,
+      "loss": 5.165,
+      "step": 765
+    },
+    {
+      "epoch": 0.1630220803405161,
+      "grad_norm": 2.2933334174932574,
+      "learning_rate": 5.4326241134751775e-05,
+      "loss": 5.1595,
+      "step": 766
+    },
+    {
+      "epoch": 0.16323490289970738,
+      "grad_norm": 2.166649164935627,
+      "learning_rate": 5.4397163120567375e-05,
+      "loss": 5.1814,
+      "step": 767
+    },
+    {
+      "epoch": 0.16344772545889863,
+      "grad_norm": 2.4224864829896484,
+      "learning_rate": 5.446808510638298e-05,
+      "loss": 5.2095,
+      "step": 768
+    },
+    {
+      "epoch": 0.1636605480180899,
+      "grad_norm": 2.3427171538330134,
+      "learning_rate": 5.453900709219858e-05,
+      "loss": 5.1726,
+      "step": 769
+    },
+    {
+      "epoch": 0.16387337057728119,
+      "grad_norm": 2.234270926762359,
+      "learning_rate": 5.460992907801419e-05,
+      "loss": 5.146,
+      "step": 770
+    },
+    {
+      "epoch": 0.16408619313647246,
+      "grad_norm": 2.9037598889008285,
+      "learning_rate": 5.468085106382979e-05,
+      "loss": 5.1717,
+      "step": 771
+    },
+    {
+      "epoch": 0.16429901569566374,
+      "grad_norm": 1.8094346917635893,
+      "learning_rate": 5.475177304964539e-05,
+      "loss": 5.1035,
+      "step": 772
+    },
+    {
+      "epoch": 0.16451183825485502,
+      "grad_norm": 2.6745197062133292,
+      "learning_rate": 5.482269503546099e-05,
+      "loss": 5.1201,
+      "step": 773
+    },
+    {
+      "epoch": 0.1647246608140463,
+      "grad_norm": 1.6907051222913603,
+      "learning_rate": 5.48936170212766e-05,
+      "loss": 5.1685,
+      "step": 774
+    },
+    {
+      "epoch": 0.16493748337323758,
+      "grad_norm": 3.056042004433111,
+      "learning_rate": 5.49645390070922e-05,
+      "loss": 5.1611,
+      "step": 775
+    },
+    {
+      "epoch": 0.16515030593242883,
+      "grad_norm": 2.025279959624647,
+      "learning_rate": 5.5035460992907805e-05,
+      "loss": 5.1564,
+      "step": 776
+    },
+    {
+      "epoch": 0.1653631284916201,
+      "grad_norm": 2.426658560906082,
+      "learning_rate": 5.5106382978723405e-05,
+      "loss": 5.1511,
+      "step": 777
+    },
+    {
+      "epoch": 0.16557595105081138,
+      "grad_norm": 2.681374246577058,
+      "learning_rate": 5.517730496453901e-05,
+      "loss": 5.1317,
+      "step": 778
+    },
+    {
+      "epoch": 0.16578877361000266,
+      "grad_norm": 2.20429790220224,
+      "learning_rate": 5.524822695035461e-05,
+      "loss": 5.1276,
+      "step": 779
+    },
+    {
+      "epoch": 0.16600159616919394,
+      "grad_norm": 2.6130677484959732,
+      "learning_rate": 5.531914893617022e-05,
+      "loss": 5.1246,
+      "step": 780
+    },
+    {
+      "epoch": 0.16621441872838522,
+      "grad_norm": 2.454141710421343,
+      "learning_rate": 5.539007092198582e-05,
+      "loss": 5.1782,
+      "step": 781
+    },
+    {
+      "epoch": 0.1664272412875765,
+      "grad_norm": 1.8218100893406277,
+      "learning_rate": 5.546099290780142e-05,
+      "loss": 5.1289,
+      "step": 782
+    },
+    {
+      "epoch": 0.16664006384676774,
+      "grad_norm": 2.808961109978377,
+      "learning_rate": 5.553191489361702e-05,
+      "loss": 5.1583,
+      "step": 783
+    },
+    {
+      "epoch": 0.16685288640595902,
+      "grad_norm": 2.3397601097762086,
+      "learning_rate": 5.560283687943263e-05,
+      "loss": 5.1503,
+      "step": 784
+    },
+    {
+      "epoch": 0.1670657089651503,
+      "grad_norm": 2.1250990972834605,
+      "learning_rate": 5.567375886524823e-05,
+      "loss": 5.1321,
+      "step": 785
+    },
+    {
+      "epoch": 0.16727853152434158,
+      "grad_norm": 2.0037492999194444,
+      "learning_rate": 5.5744680851063835e-05,
+      "loss": 5.1025,
+      "step": 786
+    },
+    {
+      "epoch": 0.16749135408353286,
+      "grad_norm": 2.7189434338569893,
+      "learning_rate": 5.5815602836879435e-05,
+      "loss": 5.1572,
+      "step": 787
+    },
+    {
+      "epoch": 0.16770417664272413,
+      "grad_norm": 1.66064844378753,
+      "learning_rate": 5.588652482269504e-05,
+      "loss": 5.1428,
+      "step": 788
+    },
+    {
+      "epoch": 0.1679169992019154,
+      "grad_norm": 3.3975903137432875,
+      "learning_rate": 5.595744680851064e-05,
+      "loss": 5.17,
+      "step": 789
+    },
+    {
+      "epoch": 0.1681298217611067,
+      "grad_norm": 1.8146298837168027,
+      "learning_rate": 5.602836879432625e-05,
+      "loss": 5.1263,
+      "step": 790
+    },
+    {
+      "epoch": 0.16834264432029794,
+      "grad_norm": 2.372432689216207,
+      "learning_rate": 5.609929078014185e-05,
+      "loss": 5.1236,
+      "step": 791
+    },
+    {
+      "epoch": 0.16855546687948922,
+      "grad_norm": 2.304171948906371,
+      "learning_rate": 5.617021276595745e-05,
+      "loss": 5.1481,
+      "step": 792
+    },
+    {
+      "epoch": 0.1687682894386805,
+      "grad_norm": 2.7447720265844904,
+      "learning_rate": 5.624113475177305e-05,
+      "loss": 5.0954,
+      "step": 793
+    },
+    {
+      "epoch": 0.16898111199787177,
+      "grad_norm": 2.2639464865759673,
+      "learning_rate": 5.631205673758866e-05,
+      "loss": 5.1433,
+      "step": 794
+    },
+    {
+      "epoch": 0.16919393455706305,
+      "grad_norm": 2.2669375407649603,
+      "learning_rate": 5.638297872340426e-05,
+      "loss": 5.1188,
+      "step": 795
+    },
+    {
+      "epoch": 0.16940675711625433,
+      "grad_norm": 3.0069725768879407,
+      "learning_rate": 5.6453900709219865e-05,
+      "loss": 5.1161,
+      "step": 796
+    },
+    {
+      "epoch": 0.1696195796754456,
+      "grad_norm": 1.8558089835150273,
+      "learning_rate": 5.6524822695035465e-05,
+      "loss": 5.0994,
+      "step": 797
+    },
+    {
+      "epoch": 0.16983240223463686,
+      "grad_norm": 2.529608930134193,
+      "learning_rate": 5.659574468085107e-05,
+      "loss": 5.1204,
+      "step": 798
+    },
+    {
+      "epoch": 0.17004522479382814,
+      "grad_norm": 1.9074363444295417,
+      "learning_rate": 5.666666666666667e-05,
+      "loss": 5.1663,
+      "step": 799
+    },
+    {
+      "epoch": 0.17025804735301941,
+      "grad_norm": 2.4686046951386036,
+      "learning_rate": 5.673758865248228e-05,
+      "loss": 5.1059,
+      "step": 800
+    },
+    {
+      "epoch": 0.1704708699122107,
+      "grad_norm": 2.0724936699819994,
+      "learning_rate": 5.680851063829788e-05,
+      "loss": 5.1216,
+      "step": 801
+    },
+    {
+      "epoch": 0.17068369247140197,
+      "grad_norm": 2.4463177908394704,
+      "learning_rate": 5.687943262411348e-05,
+      "loss": 5.1184,
+      "step": 802
+    },
+    {
+      "epoch": 0.17089651503059325,
+      "grad_norm": 2.1339579687428687,
+      "learning_rate": 5.695035460992908e-05,
+      "loss": 5.0975,
+      "step": 803
+    },
+    {
+      "epoch": 0.17110933758978453,
+      "grad_norm": 2.473333211920951,
+      "learning_rate": 5.702127659574469e-05,
+      "loss": 5.159,
+      "step": 804
+    },
+    {
+      "epoch": 0.1713221601489758,
+      "grad_norm": 2.2837682834763195,
+      "learning_rate": 5.709219858156029e-05,
+      "loss": 5.0917,
+      "step": 805
+    },
+    {
+      "epoch": 0.17153498270816706,
+      "grad_norm": 2.5647237175929103,
+      "learning_rate": 5.7163120567375895e-05,
+      "loss": 5.0611,
+      "step": 806
+    },
+    {
+      "epoch": 0.17174780526735833,
+      "grad_norm": 2.3890908043729815,
+      "learning_rate": 5.7234042553191495e-05,
+      "loss": 5.1042,
+      "step": 807
+    },
+    {
+      "epoch": 0.1719606278265496,
+      "grad_norm": 2.1896761316615554,
+      "learning_rate": 5.73049645390071e-05,
+      "loss": 5.104,
+      "step": 808
+    },
+    {
+      "epoch": 0.1721734503857409,
+      "grad_norm": 3.0709451427425534,
+      "learning_rate": 5.7375886524822694e-05,
+      "loss": 5.1246,
+      "step": 809
+    },
+    {
+      "epoch": 0.17238627294493217,
+      "grad_norm": 1.7498304798050435,
+      "learning_rate": 5.744680851063831e-05,
+      "loss": 5.0564,
+      "step": 810
+    },
+    {
+      "epoch": 0.17259909550412345,
+      "grad_norm": 3.047143893475971,
+      "learning_rate": 5.75177304964539e-05,
+      "loss": 5.1186,
+      "step": 811
+    },
+    {
+      "epoch": 0.17281191806331472,
+      "grad_norm": 1.9880472739846204,
+      "learning_rate": 5.758865248226951e-05,
+      "loss": 5.0666,
+      "step": 812
+    },
+    {
+      "epoch": 0.17302474062250597,
+      "grad_norm": 2.903878425707053,
+      "learning_rate": 5.7659574468085106e-05,
+      "loss": 5.1031,
+      "step": 813
+    },
+    {
+      "epoch": 0.17323756318169725,
+      "grad_norm": 2.3801588706989762,
+      "learning_rate": 5.7730496453900706e-05,
+      "loss": 5.0777,
+      "step": 814
+    },
+    {
+      "epoch": 0.17345038574088853,
+      "grad_norm": 1.8405751597042204,
+      "learning_rate": 5.780141843971631e-05,
+      "loss": 5.0539,
+      "step": 815
+    },
+    {
+      "epoch": 0.1736632083000798,
+      "grad_norm": 2.7547429904546896,
+      "learning_rate": 5.787234042553191e-05,
+      "loss": 5.1078,
+      "step": 816
+    },
+    {
+      "epoch": 0.17387603085927109,
+      "grad_norm": 1.9311920853725681,
+      "learning_rate": 5.794326241134752e-05,
+      "loss": 5.0689,
+      "step": 817
+    },
+    {
+      "epoch": 0.17408885341846236,
+      "grad_norm": 2.029937392745179,
+      "learning_rate": 5.801418439716312e-05,
+      "loss": 5.1019,
+      "step": 818
+    },
+    {
+      "epoch": 0.17430167597765364,
+      "grad_norm": 1.9100233568802407,
+      "learning_rate": 5.8085106382978724e-05,
+      "loss": 5.1098,
+      "step": 819
+    },
+    {
+      "epoch": 0.1745144985368449,
+      "grad_norm": 3.3616443038909365,
+      "learning_rate": 5.8156028368794324e-05,
+      "loss": 5.0644,
+      "step": 820
+    },
+    {
+      "epoch": 0.17472732109603617,
+      "grad_norm": 1.695520390740612,
+      "learning_rate": 5.822695035460993e-05,
+      "loss": 5.0547,
+      "step": 821
+    },
+    {
+      "epoch": 0.17494014365522745,
+      "grad_norm": 3.0855875675568254,
+      "learning_rate": 5.829787234042553e-05,
+      "loss": 5.1024,
+      "step": 822
+    },
+    {
+      "epoch": 0.17515296621441873,
+      "grad_norm": 2.6804383474590194,
+      "learning_rate": 5.8368794326241136e-05,
+      "loss": 5.1091,
+      "step": 823
+    },
+    {
+      "epoch": 0.17536578877361,
+      "grad_norm": 2.429319893472608,
+      "learning_rate": 5.8439716312056736e-05,
+      "loss": 5.0702,
+      "step": 824
+    },
+    {
+      "epoch": 0.17557861133280128,
+      "grad_norm": 2.4426801113766055,
+      "learning_rate": 5.851063829787234e-05,
+      "loss": 5.0748,
+      "step": 825
+    },
+    {
+      "epoch": 0.17579143389199256,
+      "grad_norm": 2.269280046786937,
+      "learning_rate": 5.858156028368794e-05,
+      "loss": 5.0698,
+      "step": 826
+    },
+    {
+      "epoch": 0.17600425645118384,
+      "grad_norm": 2.1353016258435122,
+      "learning_rate": 5.865248226950355e-05,
+      "loss": 5.0282,
+      "step": 827
+    },
+    {
+      "epoch": 0.1762170790103751,
+      "grad_norm": 2.3117792159676775,
+      "learning_rate": 5.872340425531915e-05,
+      "loss": 5.0919,
+      "step": 828
+    },
+    {
+      "epoch": 0.17642990156956637,
+      "grad_norm": 2.1312682131258756,
+      "learning_rate": 5.8794326241134754e-05,
+      "loss": 5.0989,
+      "step": 829
+    },
+    {
+      "epoch": 0.17664272412875764,
+      "grad_norm": 2.5065299319261802,
+      "learning_rate": 5.8865248226950354e-05,
+      "loss": 5.065,
+      "step": 830
+    },
+    {
+      "epoch": 0.17685554668794892,
+      "grad_norm": 2.0669154850457234,
+      "learning_rate": 5.893617021276596e-05,
+      "loss": 5.0746,
+      "step": 831
+    },
+    {
+      "epoch": 0.1770683692471402,
+      "grad_norm": 2.2585706173871802,
+      "learning_rate": 5.900709219858156e-05,
+      "loss": 5.1126,
+      "step": 832
+    },
+    {
+      "epoch": 0.17728119180633148,
+      "grad_norm": 2.2988776015711996,
+      "learning_rate": 5.9078014184397166e-05,
+      "loss": 5.0595,
+      "step": 833
+    },
+    {
+      "epoch": 0.17749401436552276,
+      "grad_norm": 2.7669245969827037,
+      "learning_rate": 5.9148936170212766e-05,
+      "loss": 5.1006,
+      "step": 834
+    },
+    {
+      "epoch": 0.177706836924714,
+      "grad_norm": 1.8051479562914206,
+      "learning_rate": 5.921985815602837e-05,
+      "loss": 5.0556,
+      "step": 835
+    },
+    {
+      "epoch": 0.17791965948390528,
+      "grad_norm": 2.547756121578628,
+      "learning_rate": 5.929078014184397e-05,
+      "loss": 5.0517,
+      "step": 836
+    },
+    {
+      "epoch": 0.17813248204309656,
+      "grad_norm": 2.0887432705118107,
+      "learning_rate": 5.936170212765958e-05,
+      "loss": 4.974,
+      "step": 837
+    },
+    {
+      "epoch": 0.17834530460228784,
+      "grad_norm": 1.9676283781371802,
+      "learning_rate": 5.943262411347518e-05,
+      "loss": 5.0382,
+      "step": 838
+    },
+    {
+      "epoch": 0.17855812716147912,
+      "grad_norm": 2.45648381670877,
+      "learning_rate": 5.9503546099290784e-05,
+      "loss": 5.034,
+      "step": 839
+    },
+    {
+      "epoch": 0.1787709497206704,
+      "grad_norm": 2.3394442161581765,
+      "learning_rate": 5.9574468085106384e-05,
+      "loss": 5.0157,
+      "step": 840
+    },
+    {
+      "epoch": 0.17898377227986167,
+      "grad_norm": 2.8561865462723666,
+      "learning_rate": 5.964539007092199e-05,
+      "loss": 5.056,
+      "step": 841
+    },
+    {
+      "epoch": 0.17919659483905295,
+      "grad_norm": 2.072981187949757,
+      "learning_rate": 5.971631205673759e-05,
+      "loss": 5.025,
+      "step": 842
+    },
+    {
+      "epoch": 0.1794094173982442,
+      "grad_norm": 2.779170443637206,
+      "learning_rate": 5.9787234042553196e-05,
+      "loss": 5.0559,
+      "step": 843
+    },
+    {
+      "epoch": 0.17962223995743548,
+      "grad_norm": 2.1754761653745316,
+      "learning_rate": 5.9858156028368796e-05,
+      "loss": 5.0384,
+      "step": 844
+    },
+    {
+      "epoch": 0.17983506251662676,
+      "grad_norm": 2.6838215902203277,
+      "learning_rate": 5.99290780141844e-05,
+      "loss": 5.0091,
+      "step": 845
+    },
+    {
+      "epoch": 0.18004788507581804,
+      "grad_norm": 2.1937761592057554,
+      "learning_rate": 6e-05,
+      "loss": 5.0032,
+      "step": 846
+    },
+    {
+      "epoch": 0.18026070763500932,
+      "grad_norm": 2.35659185147198,
+      "learning_rate": 6.007092198581561e-05,
+      "loss": 5.0594,
+      "step": 847
+    },
+    {
+      "epoch": 0.1804735301942006,
+      "grad_norm": 2.3062128890761597,
+      "learning_rate": 6.014184397163121e-05,
+      "loss": 5.0195,
+      "step": 848
+    },
+    {
+      "epoch": 0.18068635275339187,
+      "grad_norm": 2.263507848175629,
+      "learning_rate": 6.0212765957446814e-05,
+      "loss": 5.0499,
+      "step": 849
+    },
+    {
+      "epoch": 0.18089917531258312,
+      "grad_norm": 2.7188514327102022,
+      "learning_rate": 6.0283687943262414e-05,
+      "loss": 5.0555,
+      "step": 850
+    },
+    {
+      "epoch": 0.1811119978717744,
+      "grad_norm": 2.170378320066492,
+      "learning_rate": 6.035460992907802e-05,
+      "loss": 5.0388,
+      "step": 851
+    },
+    {
+      "epoch": 0.18132482043096568,
+      "grad_norm": 2.388471408969847,
+      "learning_rate": 6.042553191489362e-05,
+      "loss": 4.9955,
+      "step": 852
+    },
+    {
+      "epoch": 0.18153764299015696,
+      "grad_norm": 2.00384182022081,
+      "learning_rate": 6.0496453900709226e-05,
+      "loss": 4.9927,
+      "step": 853
+    },
+    {
+      "epoch": 0.18175046554934823,
+      "grad_norm": 2.552554662587966,
+      "learning_rate": 6.0567375886524826e-05,
+      "loss": 5.0111,
+      "step": 854
+    },
+    {
+      "epoch": 0.1819632881085395,
+      "grad_norm": 1.9491764709324937,
+      "learning_rate": 6.063829787234043e-05,
+      "loss": 5.0166,
+      "step": 855
+    },
+    {
+      "epoch": 0.1821761106677308,
+      "grad_norm": 2.371266700839773,
+      "learning_rate": 6.070921985815603e-05,
+      "loss": 4.9399,
+      "step": 856
+    },
+    {
+      "epoch": 0.18238893322692207,
+      "grad_norm": 2.0760361379645995,
+      "learning_rate": 6.078014184397164e-05,
+      "loss": 4.9895,
+      "step": 857
+    },
+    {
+      "epoch": 0.18260175578611332,
+      "grad_norm": 2.2125569954820894,
+      "learning_rate": 6.085106382978724e-05,
+      "loss": 4.9976,
+      "step": 858
+    },
+    {
+      "epoch": 0.1828145783453046,
+      "grad_norm": 2.4190879031910186,
+      "learning_rate": 6.0921985815602844e-05,
+      "loss": 5.019,
+      "step": 859
+    },
+    {
+      "epoch": 0.18302740090449587,
+      "grad_norm": 2.037787042098734,
+      "learning_rate": 6.0992907801418444e-05,
+      "loss": 4.9441,
+      "step": 860
+    },
+    {
+      "epoch": 0.18324022346368715,
+      "grad_norm": 2.6457568914310556,
+      "learning_rate": 6.106382978723405e-05,
+      "loss": 4.992,
+      "step": 861
+    },
+    {
+      "epoch": 0.18345304602287843,
+      "grad_norm": 1.822648122946518,
+      "learning_rate": 6.113475177304964e-05,
+      "loss": 4.9821,
+      "step": 862
+    },
+    {
+      "epoch": 0.1836658685820697,
+      "grad_norm": 2.50519965853648,
+      "learning_rate": 6.120567375886526e-05,
+      "loss": 5.0279,
+      "step": 863
+    },
+    {
+      "epoch": 0.18387869114126099,
+      "grad_norm": 2.1712461178351847,
+      "learning_rate": 6.127659574468086e-05,
+      "loss": 5.0152,
+      "step": 864
+    },
+    {
+      "epoch": 0.18409151370045224,
+      "grad_norm": 2.2696289423254914,
+      "learning_rate": 6.134751773049646e-05,
+      "loss": 4.9489,
+      "step": 865
+    },
+    {
+      "epoch": 0.18430433625964351,
+      "grad_norm": 2.174140476914681,
+      "learning_rate": 6.141843971631205e-05,
+      "loss": 5.0067,
+      "step": 866
+    },
+    {
+      "epoch": 0.1845171588188348,
+      "grad_norm": 3.000418344272543,
+      "learning_rate": 6.148936170212767e-05,
+      "loss": 5.0204,
+      "step": 867
+    },
+    {
+      "epoch": 0.18472998137802607,
+      "grad_norm": 1.9218111119219279,
+      "learning_rate": 6.156028368794327e-05,
+      "loss": 4.9548,
+      "step": 868
+    },
+    {
+      "epoch": 0.18494280393721735,
+      "grad_norm": 2.4203666857914943,
+      "learning_rate": 6.163120567375886e-05,
+      "loss": 5.0546,
+      "step": 869
+    },
+    {
+      "epoch": 0.18515562649640863,
+      "grad_norm": 2.164269907605388,
+      "learning_rate": 6.170212765957447e-05,
+      "loss": 4.9813,
+      "step": 870
+    },
+    {
+      "epoch": 0.1853684490555999,
+      "grad_norm": 2.108370962745802,
+      "learning_rate": 6.177304964539007e-05,
+      "loss": 4.9663,
+      "step": 871
+    },
+    {
+      "epoch": 0.18558127161479115,
+      "grad_norm": 2.318819009256907,
+      "learning_rate": 6.184397163120568e-05,
+      "loss": 4.9852,
+      "step": 872
+    },
+    {
+      "epoch": 0.18579409417398243,
+      "grad_norm": 1.5076731821074494,
+      "learning_rate": 6.191489361702127e-05,
+      "loss": 4.9758,
+      "step": 873
+    },
+    {
+      "epoch": 0.1860069167331737,
+      "grad_norm": 3.2359533385291774,
+      "learning_rate": 6.198581560283688e-05,
+      "loss": 4.9656,
+      "step": 874
+    },
+    {
+      "epoch": 0.186219739292365,
+      "grad_norm": 1.6999531390151823,
+      "learning_rate": 6.205673758865249e-05,
+      "loss": 4.9797,
+      "step": 875
+    },
+    {
+      "epoch": 0.18643256185155627,
+      "grad_norm": 2.716619377271891,
+      "learning_rate": 6.212765957446809e-05,
+      "loss": 4.9912,
+      "step": 876
+    },
+    {
+      "epoch": 0.18664538441074754,
+      "grad_norm": 2.1460745352933284,
+      "learning_rate": 6.219858156028368e-05,
+      "loss": 4.983,
+      "step": 877
+    },
+    {
+      "epoch": 0.18685820696993882,
+      "grad_norm": 2.9381611451066205,
+      "learning_rate": 6.226950354609929e-05,
+      "loss": 5.0019,
+      "step": 878
+    },
+    {
+      "epoch": 0.1870710295291301,
+      "grad_norm": 2.1926788413558804,
+      "learning_rate": 6.23404255319149e-05,
+      "loss": 4.9135,
+      "step": 879
+    },
+    {
+      "epoch": 0.18728385208832135,
+      "grad_norm": 2.5153354862181367,
+      "learning_rate": 6.24113475177305e-05,
+      "loss": 4.9849,
+      "step": 880
+    },
+    {
+      "epoch": 0.18749667464751263,
+      "grad_norm": 2.485700827833569,
+      "learning_rate": 6.24822695035461e-05,
+      "loss": 4.9424,
+      "step": 881
+    },
+    {
+      "epoch": 0.1877094972067039,
+      "grad_norm": 2.4192983741507272,
+      "learning_rate": 6.25531914893617e-05,
+      "loss": 4.9425,
+      "step": 882
+    },
+    {
+      "epoch": 0.18792231976589518,
+      "grad_norm": 2.231388607593025,
+      "learning_rate": 6.262411347517731e-05,
+      "loss": 4.9501,
+      "step": 883
+    },
+    {
+      "epoch": 0.18813514232508646,
+      "grad_norm": 2.3924128018991033,
+      "learning_rate": 6.269503546099292e-05,
+      "loss": 4.9699,
+      "step": 884
+    },
+    {
+      "epoch": 0.18834796488427774,
+      "grad_norm": 2.498251140169476,
+      "learning_rate": 6.276595744680851e-05,
+      "loss": 4.98,
+      "step": 885
+    },
+    {
+      "epoch": 0.18856078744346902,
+      "grad_norm": 2.1141904825291413,
+      "learning_rate": 6.283687943262411e-05,
+      "loss": 4.9263,
+      "step": 886
+    },
+    {
+      "epoch": 0.18877361000266027,
+      "grad_norm": 2.211111542331923,
+      "learning_rate": 6.290780141843972e-05,
+      "loss": 4.9662,
+      "step": 887
+    },
+    {
+      "epoch": 0.18898643256185155,
+      "grad_norm": 2.323020388877635,
+      "learning_rate": 6.297872340425533e-05,
+      "loss": 4.9516,
+      "step": 888
+    },
+    {
+      "epoch": 0.18919925512104283,
+      "grad_norm": 2.1739690809601955,
+      "learning_rate": 6.304964539007092e-05,
+      "loss": 4.9708,
+      "step": 889
+    },
+    {
+      "epoch": 0.1894120776802341,
+      "grad_norm": 2.631659444160793,
+      "learning_rate": 6.312056737588653e-05,
+      "loss": 4.9197,
+      "step": 890
+    },
+    {
+      "epoch": 0.18962490023942538,
+      "grad_norm": 2.2351036816446723,
+      "learning_rate": 6.319148936170213e-05,
+      "loss": 4.9322,
+      "step": 891
+    },
+    {
+      "epoch": 0.18983772279861666,
+      "grad_norm": 2.586965778105869,
+      "learning_rate": 6.326241134751774e-05,
+      "loss": 4.9568,
+      "step": 892
+    },
+    {
+      "epoch": 0.19005054535780794,
+      "grad_norm": 1.6242008337679124,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 4.9507,
+      "step": 893
+    },
+    {
+      "epoch": 0.19026336791699922,
+      "grad_norm": 3.0764738283429667,
+      "learning_rate": 6.340425531914894e-05,
+      "loss": 4.9616,
+      "step": 894
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 1.8432592593003045,
+      "learning_rate": 6.347517730496455e-05,
+      "loss": 4.981,
+      "step": 895
+    },
+    {
+      "epoch": 0.19068901303538174,
+      "grad_norm": 2.951909803301084,
+      "learning_rate": 6.354609929078015e-05,
+      "loss": 4.9931,
+      "step": 896
+    },
+    {
+      "epoch": 0.19090183559457302,
+      "grad_norm": 1.960322993494104,
+      "learning_rate": 6.361702127659574e-05,
+      "loss": 4.9554,
+      "step": 897
+    },
+    {
+      "epoch": 0.1911146581537643,
+      "grad_norm": 2.6261670093180642,
+      "learning_rate": 6.368794326241135e-05,
+      "loss": 4.9272,
+      "step": 898
+    },
+    {
+      "epoch": 0.19132748071295558,
+      "grad_norm": 2.4004585941662024,
+      "learning_rate": 6.375886524822696e-05,
+      "loss": 4.9063,
+      "step": 899
+    },
+    {
+      "epoch": 0.19154030327214686,
+      "grad_norm": 2.093768461660987,
+      "learning_rate": 6.382978723404256e-05,
+      "loss": 4.9588,
+      "step": 900
+    },
+    {
+      "epoch": 0.19175312583133813,
+      "grad_norm": 2.541873939390849,
+      "learning_rate": 6.390070921985816e-05,
+      "loss": 4.8925,
+      "step": 901
+    },
+    {
+      "epoch": 0.19196594839052938,
+      "grad_norm": 2.222671150848907,
+      "learning_rate": 6.397163120567376e-05,
+      "loss": 4.9531,
+      "step": 902
+    },
+    {
+      "epoch": 0.19217877094972066,
+      "grad_norm": 2.7214626648830467,
+      "learning_rate": 6.404255319148937e-05,
+      "loss": 4.9387,
+      "step": 903
+    },
+    {
+      "epoch": 0.19239159350891194,
+      "grad_norm": 2.0957800188565736,
+      "learning_rate": 6.411347517730498e-05,
+      "loss": 4.924,
+      "step": 904
+    },
+    {
+      "epoch": 0.19260441606810322,
+      "grad_norm": 2.3568532438936924,
+      "learning_rate": 6.418439716312057e-05,
+      "loss": 4.9248,
+      "step": 905
+    },
+    {
+      "epoch": 0.1928172386272945,
+      "grad_norm": 2.5693612628535356,
+      "learning_rate": 6.425531914893617e-05,
+      "loss": 4.9591,
+      "step": 906
+    },
+    {
+      "epoch": 0.19303006118648577,
+      "grad_norm": 2.3644716890188113,
+      "learning_rate": 6.432624113475178e-05,
+      "loss": 4.9556,
+      "step": 907
+    },
+    {
+      "epoch": 0.19324288374567705,
+      "grad_norm": 2.2614432339327277,
+      "learning_rate": 6.439716312056739e-05,
+      "loss": 4.8597,
+      "step": 908
+    },
+    {
+      "epoch": 0.1934557063048683,
+      "grad_norm": 1.8913363671980772,
+      "learning_rate": 6.446808510638298e-05,
+      "loss": 4.9046,
+      "step": 909
+    },
+    {
+      "epoch": 0.19366852886405958,
+      "grad_norm": 2.6625761824639267,
+      "learning_rate": 6.453900709219859e-05,
+      "loss": 4.9043,
+      "step": 910
+    },
+    {
+      "epoch": 0.19388135142325086,
+      "grad_norm": 1.9844983165926182,
+      "learning_rate": 6.460992907801419e-05,
+      "loss": 4.9292,
+      "step": 911
+    },
+    {
+      "epoch": 0.19409417398244214,
+      "grad_norm": 2.645119021955229,
+      "learning_rate": 6.46808510638298e-05,
+      "loss": 4.9064,
+      "step": 912
+    },
+    {
+      "epoch": 0.19430699654163341,
+      "grad_norm": 1.7686370188448874,
+      "learning_rate": 6.475177304964539e-05,
+      "loss": 4.9189,
+      "step": 913
+    },
+    {
+      "epoch": 0.1945198191008247,
+      "grad_norm": 2.8385453421056654,
+      "learning_rate": 6.4822695035461e-05,
+      "loss": 4.9143,
+      "step": 914
+    },
+    {
+      "epoch": 0.19473264166001597,
+      "grad_norm": 1.7723704897344263,
+      "learning_rate": 6.489361702127659e-05,
+      "loss": 4.9164,
+      "step": 915
+    },
+    {
+      "epoch": 0.19494546421920725,
+      "grad_norm": 2.739512951388155,
+      "learning_rate": 6.496453900709221e-05,
+      "loss": 4.8999,
+      "step": 916
+    },
+    {
+      "epoch": 0.1951582867783985,
+      "grad_norm": 2.1119707971067676,
+      "learning_rate": 6.50354609929078e-05,
+      "loss": 4.8679,
+      "step": 917
+    },
+    {
+      "epoch": 0.19537110933758978,
+      "grad_norm": 2.106907393168478,
+      "learning_rate": 6.510638297872341e-05,
+      "loss": 4.9177,
+      "step": 918
+    },
+    {
+      "epoch": 0.19558393189678105,
+      "grad_norm": 2.3923210152838332,
+      "learning_rate": 6.5177304964539e-05,
+      "loss": 4.8948,
+      "step": 919
+    },
+    {
+      "epoch": 0.19579675445597233,
+      "grad_norm": 2.261205027916062,
+      "learning_rate": 6.524822695035462e-05,
+      "loss": 4.9521,
+      "step": 920
+    },
+    {
+      "epoch": 0.1960095770151636,
+      "grad_norm": 1.811115764564855,
+      "learning_rate": 6.531914893617022e-05,
+      "loss": 4.8593,
+      "step": 921
+    },
+    {
+      "epoch": 0.1962223995743549,
+      "grad_norm": 2.6471895507647623,
+      "learning_rate": 6.539007092198581e-05,
+      "loss": 4.9502,
+      "step": 922
+    },
+    {
+      "epoch": 0.19643522213354617,
+      "grad_norm": 2.037700486742533,
+      "learning_rate": 6.546099290780142e-05,
+      "loss": 4.8872,
+      "step": 923
+    },
+    {
+      "epoch": 0.19664804469273742,
+      "grad_norm": 2.2405870266082206,
+      "learning_rate": 6.553191489361702e-05,
+      "loss": 4.898,
+      "step": 924
+    },
+    {
+      "epoch": 0.1968608672519287,
+      "grad_norm": 2.022963774160343,
+      "learning_rate": 6.560283687943263e-05,
+      "loss": 4.8915,
+      "step": 925
+    },
+    {
+      "epoch": 0.19707368981111997,
+      "grad_norm": 2.2786288258375413,
+      "learning_rate": 6.567375886524822e-05,
+      "loss": 4.9259,
+      "step": 926
+    },
+    {
+      "epoch": 0.19728651237031125,
+      "grad_norm": 2.8780244039201888,
+      "learning_rate": 6.574468085106383e-05,
+      "loss": 4.9169,
+      "step": 927
+    },
+    {
+      "epoch": 0.19749933492950253,
+      "grad_norm": 1.694110158423583,
+      "learning_rate": 6.581560283687943e-05,
+      "loss": 4.9334,
+      "step": 928
+    },
+    {
+      "epoch": 0.1977121574886938,
+      "grad_norm": 2.1610333523027974,
+      "learning_rate": 6.588652482269504e-05,
+      "loss": 4.8852,
+      "step": 929
+    },
+    {
+      "epoch": 0.19792498004788509,
+      "grad_norm": 2.5759791192519748,
+      "learning_rate": 6.595744680851063e-05,
+      "loss": 4.8701,
+      "step": 930
+    },
+    {
+      "epoch": 0.19813780260707636,
+      "grad_norm": 2.1304008451310397,
+      "learning_rate": 6.602836879432624e-05,
+      "loss": 4.8826,
+      "step": 931
+    },
+    {
+      "epoch": 0.1983506251662676,
+      "grad_norm": 2.353198446733796,
+      "learning_rate": 6.609929078014185e-05,
+      "loss": 4.9092,
+      "step": 932
+    },
+    {
+      "epoch": 0.1985634477254589,
+      "grad_norm": 2.1109800966987384,
+      "learning_rate": 6.617021276595745e-05,
+      "loss": 4.894,
+      "step": 933
+    },
+    {
+      "epoch": 0.19877627028465017,
+      "grad_norm": 2.276492568351092,
+      "learning_rate": 6.624113475177305e-05,
+      "loss": 4.8677,
+      "step": 934
+    },
+    {
+      "epoch": 0.19898909284384145,
+      "grad_norm": 2.657136093794072,
+      "learning_rate": 6.631205673758865e-05,
+      "loss": 4.8616,
+      "step": 935
+    },
+    {
+      "epoch": 0.19920191540303273,
+      "grad_norm": 2.2543125385702267,
+      "learning_rate": 6.638297872340426e-05,
+      "loss": 4.8812,
+      "step": 936
+    },
+    {
+      "epoch": 0.199414737962224,
+      "grad_norm": 2.4472869305256153,
+      "learning_rate": 6.645390070921986e-05,
+      "loss": 4.8792,
+      "step": 937
+    },
+    {
+      "epoch": 0.19962756052141528,
+      "grad_norm": 2.5182050360799777,
+      "learning_rate": 6.652482269503546e-05,
+      "loss": 4.8442,
+      "step": 938
+    },
+    {
+      "epoch": 0.19984038308060653,
+      "grad_norm": 1.8024595638701086,
+      "learning_rate": 6.659574468085106e-05,
+      "loss": 4.9151,
+      "step": 939
+    },
+    {
+      "epoch": 0.2000532056397978,
+      "grad_norm": 2.68220844813569,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 4.9059,
+      "step": 940
+    },
+    {
+      "epoch": 0.2002660281989891,
+      "grad_norm": 1.99133727567097,
+      "learning_rate": 6.673758865248228e-05,
+      "loss": 4.8244,
+      "step": 941
+    },
+    {
+      "epoch": 0.20047885075818037,
+      "grad_norm": 2.920651175561805,
+      "learning_rate": 6.680851063829787e-05,
+      "loss": 4.9135,
+      "step": 942
+    },
+    {
+      "epoch": 0.20069167331737164,
+      "grad_norm": 1.939452935403264,
+      "learning_rate": 6.687943262411348e-05,
+      "loss": 4.8538,
+      "step": 943
+    },
+    {
+      "epoch": 0.20090449587656292,
+      "grad_norm": 2.364148592416172,
+      "learning_rate": 6.695035460992908e-05,
+      "loss": 4.8964,
+      "step": 944
+    },
+    {
+      "epoch": 0.2011173184357542,
+      "grad_norm": 2.4221844021971455,
+      "learning_rate": 6.702127659574469e-05,
+      "loss": 4.8713,
+      "step": 945
+    },
+    {
+      "epoch": 0.20133014099494548,
+      "grad_norm": 2.2137313308980313,
+      "learning_rate": 6.709219858156028e-05,
+      "loss": 4.8459,
+      "step": 946
+    },
+    {
+      "epoch": 0.20154296355413673,
+      "grad_norm": 2.8287137416385995,
+      "learning_rate": 6.716312056737589e-05,
+      "loss": 4.9117,
+      "step": 947
+    },
+    {
+      "epoch": 0.201755786113328,
+      "grad_norm": 1.9939178440296723,
+      "learning_rate": 6.72340425531915e-05,
+      "loss": 4.8697,
+      "step": 948
+    },
+    {
+      "epoch": 0.20196860867251928,
+      "grad_norm": 2.5471164989446184,
+      "learning_rate": 6.73049645390071e-05,
+      "loss": 4.9177,
+      "step": 949
+    },
+    {
+      "epoch": 0.20218143123171056,
+      "grad_norm": 1.7357290483931247,
+      "learning_rate": 6.737588652482269e-05,
+      "loss": 4.8936,
+      "step": 950
+    },
+    {
+      "epoch": 0.20239425379090184,
+      "grad_norm": 2.323358025420465,
+      "learning_rate": 6.74468085106383e-05,
+      "loss": 4.8148,
+      "step": 951
+    },
+    {
+      "epoch": 0.20260707635009312,
+      "grad_norm": 2.1856975829962755,
+      "learning_rate": 6.75177304964539e-05,
+      "loss": 4.9061,
+      "step": 952
+    },
+    {
+      "epoch": 0.2028198989092844,
+      "grad_norm": 2.4529449168929087,
+      "learning_rate": 6.758865248226951e-05,
+      "loss": 4.8443,
+      "step": 953
+    },
+    {
+      "epoch": 0.20303272146847565,
+      "grad_norm": 2.209680163202125,
+      "learning_rate": 6.76595744680851e-05,
+      "loss": 4.8931,
+      "step": 954
+    },
+    {
+      "epoch": 0.20324554402766692,
+      "grad_norm": 2.2303707981411387,
+      "learning_rate": 6.773049645390071e-05,
+      "loss": 4.84,
+      "step": 955
+    },
+    {
+      "epoch": 0.2034583665868582,
+      "grad_norm": 2.3971096816178363,
+      "learning_rate": 6.780141843971632e-05,
+      "loss": 4.9054,
+      "step": 956
+    },
+    {
+      "epoch": 0.20367118914604948,
+      "grad_norm": 2.0714679625205514,
+      "learning_rate": 6.787234042553192e-05,
+      "loss": 4.8568,
+      "step": 957
+    },
+    {
+      "epoch": 0.20388401170524076,
+      "grad_norm": 2.1162002090995955,
+      "learning_rate": 6.794326241134752e-05,
+      "loss": 4.8257,
+      "step": 958
+    },
+    {
+      "epoch": 0.20409683426443204,
+      "grad_norm": 2.4127651844072493,
+      "learning_rate": 6.801418439716312e-05,
+      "loss": 4.8547,
+      "step": 959
+    },
+    {
+      "epoch": 0.20430965682362331,
+      "grad_norm": 2.034350722531473,
+      "learning_rate": 6.808510638297873e-05,
+      "loss": 4.8585,
+      "step": 960
+    },
+    {
+      "epoch": 0.20452247938281456,
+      "grad_norm": 2.1846114739358122,
+      "learning_rate": 6.815602836879434e-05,
+      "loss": 4.8433,
+      "step": 961
+    },
+    {
+      "epoch": 0.20473530194200584,
+      "grad_norm": 2.0121668678313807,
+      "learning_rate": 6.822695035460993e-05,
+      "loss": 4.8345,
+      "step": 962
+    },
+    {
+      "epoch": 0.20494812450119712,
+      "grad_norm": 3.037051411725841,
+      "learning_rate": 6.829787234042554e-05,
+      "loss": 4.8739,
+      "step": 963
+    },
+    {
+      "epoch": 0.2051609470603884,
+      "grad_norm": 1.8753054331688839,
+      "learning_rate": 6.836879432624114e-05,
+      "loss": 4.7929,
+      "step": 964
+    },
+    {
+      "epoch": 0.20537376961957968,
+      "grad_norm": 2.732821950485442,
+      "learning_rate": 6.843971631205675e-05,
+      "loss": 4.8017,
+      "step": 965
+    },
+    {
+      "epoch": 0.20558659217877095,
+      "grad_norm": 1.6899150757321946,
+      "learning_rate": 6.851063829787234e-05,
+      "loss": 4.8308,
+      "step": 966
+    },
+    {
+      "epoch": 0.20579941473796223,
+      "grad_norm": 3.1052727200846455,
+      "learning_rate": 6.858156028368795e-05,
+      "loss": 4.8742,
+      "step": 967
+    },
+    {
+      "epoch": 0.2060122372971535,
+      "grad_norm": 2.065548629803988,
+      "learning_rate": 6.865248226950355e-05,
+      "loss": 4.86,
+      "step": 968
+    },
+    {
+      "epoch": 0.20622505985634476,
+      "grad_norm": 2.4979848125523985,
+      "learning_rate": 6.872340425531916e-05,
+      "loss": 4.8331,
+      "step": 969
+    },
+    {
+      "epoch": 0.20643788241553604,
+      "grad_norm": 2.1867596237980123,
+      "learning_rate": 6.879432624113475e-05,
+      "loss": 4.8706,
+      "step": 970
+    },
+    {
+      "epoch": 0.20665070497472732,
+      "grad_norm": 2.4570934617885483,
+      "learning_rate": 6.886524822695036e-05,
+      "loss": 4.8342,
+      "step": 971
+    },
+    {
+      "epoch": 0.2068635275339186,
+      "grad_norm": 2.3006005902112574,
+      "learning_rate": 6.893617021276597e-05,
+      "loss": 4.8602,
+      "step": 972
+    },
+    {
+      "epoch": 0.20707635009310987,
+      "grad_norm": 2.0272972490088415,
+      "learning_rate": 6.900709219858157e-05,
+      "loss": 4.8876,
+      "step": 973
+    },
+    {
+      "epoch": 0.20728917265230115,
+      "grad_norm": 2.770186061334574,
+      "learning_rate": 6.907801418439717e-05,
+      "loss": 4.8511,
+      "step": 974
+    },
+    {
+      "epoch": 0.20750199521149243,
+      "grad_norm": 2.2380646214061395,
+      "learning_rate": 6.914893617021277e-05,
+      "loss": 4.8561,
+      "step": 975
+    },
+    {
+      "epoch": 0.20771481777068368,
+      "grad_norm": 1.9737294937305119,
+      "learning_rate": 6.921985815602838e-05,
+      "loss": 4.7881,
+      "step": 976
+    },
+    {
+      "epoch": 0.20792764032987496,
+      "grad_norm": 2.403559763772364,
+      "learning_rate": 6.929078014184397e-05,
+      "loss": 4.8283,
+      "step": 977
+    },
+    {
+      "epoch": 0.20814046288906624,
+      "grad_norm": 2.2632324063100673,
+      "learning_rate": 6.936170212765958e-05,
+      "loss": 4.8257,
+      "step": 978
+    },
+    {
+      "epoch": 0.2083532854482575,
+      "grad_norm": 2.332972473365571,
+      "learning_rate": 6.943262411347517e-05,
+      "loss": 4.8558,
+      "step": 979
+    },
+    {
+      "epoch": 0.2085661080074488,
+      "grad_norm": 2.1145052820267622,
+      "learning_rate": 6.950354609929079e-05,
+      "loss": 4.8143,
+      "step": 980
+    },
+    {
+      "epoch": 0.20877893056664007,
+      "grad_norm": 2.376740342818793,
+      "learning_rate": 6.957446808510638e-05,
+      "loss": 4.8196,
+      "step": 981
+    },
+    {
+      "epoch": 0.20899175312583135,
+      "grad_norm": 1.9149169298265858,
+      "learning_rate": 6.964539007092199e-05,
+      "loss": 4.84,
+      "step": 982
+    },
+    {
+      "epoch": 0.20920457568502263,
+      "grad_norm": 2.3547366708982733,
+      "learning_rate": 6.971631205673758e-05,
+      "loss": 4.8446,
+      "step": 983
+    },
+    {
+      "epoch": 0.20941739824421388,
+      "grad_norm": 1.8171645167683712,
+      "learning_rate": 6.97872340425532e-05,
+      "loss": 4.8469,
+      "step": 984
+    },
+    {
+      "epoch": 0.20963022080340515,
+      "grad_norm": 2.3687352645732798,
+      "learning_rate": 6.98581560283688e-05,
+      "loss": 4.7923,
+      "step": 985
+    },
+    {
+      "epoch": 0.20984304336259643,
+      "grad_norm": 1.9896810841022377,
+      "learning_rate": 6.99290780141844e-05,
+      "loss": 4.851,
+      "step": 986
+    },
+    {
+      "epoch": 0.2100558659217877,
+      "grad_norm": 2.423226973258002,
+      "learning_rate": 7e-05,
+      "loss": 4.8138,
+      "step": 987
+    },
+    {
+      "epoch": 0.210268688480979,
+      "grad_norm": 1.690367766739755,
+      "learning_rate": 7.007092198581561e-05,
+      "loss": 4.7523,
+      "step": 988
+    },
+    {
+      "epoch": 0.21048151104017027,
+      "grad_norm": 3.0171654394365337,
+      "learning_rate": 7.01418439716312e-05,
+      "loss": 4.8024,
+      "step": 989
+    },
+    {
+      "epoch": 0.21069433359936154,
+      "grad_norm": 1.7944053473712451,
+      "learning_rate": 7.021276595744681e-05,
+      "loss": 4.8062,
+      "step": 990
+    },
+    {
+      "epoch": 0.2109071561585528,
+      "grad_norm": 2.0865587927519553,
+      "learning_rate": 7.02836879432624e-05,
+      "loss": 4.7818,
+      "step": 991
+    },
+    {
+      "epoch": 0.21111997871774407,
+      "grad_norm": 2.268644468212445,
+      "learning_rate": 7.035460992907803e-05,
+      "loss": 4.8427,
+      "step": 992
+    },
+    {
+      "epoch": 0.21133280127693535,
+      "grad_norm": 2.220803019116365,
+      "learning_rate": 7.042553191489362e-05,
+      "loss": 4.8001,
+      "step": 993
+    },
+    {
+      "epoch": 0.21154562383612663,
+      "grad_norm": 2.6658009970799683,
+      "learning_rate": 7.049645390070923e-05,
+      "loss": 4.8323,
+      "step": 994
+    },
+    {
+      "epoch": 0.2117584463953179,
+      "grad_norm": 1.971456001098186,
+      "learning_rate": 7.056737588652482e-05,
+      "loss": 4.8426,
+      "step": 995
+    },
+    {
+      "epoch": 0.21197126895450918,
+      "grad_norm": 1.9951119207079262,
+      "learning_rate": 7.063829787234042e-05,
+      "loss": 4.8029,
+      "step": 996
+    },
+    {
+      "epoch": 0.21218409151370046,
+      "grad_norm": 2.5879511940908406,
+      "learning_rate": 7.070921985815603e-05,
+      "loss": 4.8074,
+      "step": 997
+    },
+    {
+      "epoch": 0.2123969140728917,
+      "grad_norm": 2.1332346863334832,
+      "learning_rate": 7.078014184397164e-05,
+      "loss": 4.7543,
+      "step": 998
+    },
+    {
+      "epoch": 0.212609736632083,
+      "grad_norm": 2.697748619639635,
+      "learning_rate": 7.085106382978723e-05,
+      "loss": 4.7735,
+      "step": 999
+    },
+    {
+      "epoch": 0.21282255919127427,
+      "grad_norm": 1.8111399494673548,
+      "learning_rate": 7.092198581560284e-05,
+      "loss": 4.8012,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21303538175046555,
+      "grad_norm": 2.558400615145693,
+      "learning_rate": 7.099290780141844e-05,
+      "loss": 4.7799,
+      "step": 1001
+    },
+    {
+      "epoch": 0.21324820430965682,
+      "grad_norm": 2.0847838333559676,
+      "learning_rate": 7.106382978723405e-05,
+      "loss": 4.7627,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2134610268688481,
+      "grad_norm": 2.4386645656029775,
+      "learning_rate": 7.113475177304964e-05,
+      "loss": 4.7761,
+      "step": 1003
+    },
+    {
+      "epoch": 0.21367384942803938,
+      "grad_norm": 1.9330313249781312,
+      "learning_rate": 7.120567375886525e-05,
+      "loss": 4.7757,
+      "step": 1004
+    },
+    {
+      "epoch": 0.21388667198723066,
+      "grad_norm": 2.569613210456521,
+      "learning_rate": 7.127659574468085e-05,
+      "loss": 4.7381,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2140994945464219,
+      "grad_norm": 1.9910653315978328,
+      "learning_rate": 7.134751773049646e-05,
+      "loss": 4.7914,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2143123171056132,
+      "grad_norm": 1.9866384251033626,
+      "learning_rate": 7.141843971631205e-05,
+      "loss": 4.7786,
+      "step": 1007
+    },
+    {
+      "epoch": 0.21452513966480447,
+      "grad_norm": 2.372430223927559,
+      "learning_rate": 7.148936170212766e-05,
+      "loss": 4.7353,
+      "step": 1008
+    },
+    {
+      "epoch": 0.21473796222399574,
+      "grad_norm": 1.700771410136623,
+      "learning_rate": 7.156028368794327e-05,
+      "loss": 4.7904,
+      "step": 1009
+    },
+    {
+      "epoch": 0.21495078478318702,
+      "grad_norm": 2.3221394029825,
+      "learning_rate": 7.163120567375887e-05,
+      "loss": 4.7715,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2151636073423783,
+      "grad_norm": 1.9250791363442843,
+      "learning_rate": 7.170212765957447e-05,
+      "loss": 4.7721,
+      "step": 1011
+    },
+    {
+      "epoch": 0.21537642990156958,
+      "grad_norm": 2.5751883074370405,
+      "learning_rate": 7.177304964539007e-05,
+      "loss": 4.7723,
+      "step": 1012
+    },
+    {
+      "epoch": 0.21558925246076083,
+      "grad_norm": 2.3688856629309103,
+      "learning_rate": 7.184397163120568e-05,
+      "loss": 4.7741,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2158020750199521,
+      "grad_norm": 2.5695014470888813,
+      "learning_rate": 7.191489361702129e-05,
+      "loss": 4.7944,
+      "step": 1014
+    },
+    {
+      "epoch": 0.21601489757914338,
+      "grad_norm": 2.1325794787850962,
+      "learning_rate": 7.198581560283688e-05,
+      "loss": 4.7859,
+      "step": 1015
+    },
+    {
+      "epoch": 0.21622772013833466,
+      "grad_norm": 2.541248489661995,
+      "learning_rate": 7.205673758865248e-05,
+      "loss": 4.7923,
+      "step": 1016
+    },
+    {
+      "epoch": 0.21644054269752594,
+      "grad_norm": 2.2740510588831007,
+      "learning_rate": 7.212765957446809e-05,
+      "loss": 4.7451,
+      "step": 1017
+    },
+    {
+      "epoch": 0.21665336525671722,
+      "grad_norm": 2.4638666904229187,
+      "learning_rate": 7.21985815602837e-05,
+      "loss": 4.7725,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2168661878159085,
+      "grad_norm": 1.7400178646895241,
+      "learning_rate": 7.226950354609929e-05,
+      "loss": 4.7473,
+      "step": 1019
+    },
+    {
+      "epoch": 0.21707901037509977,
+      "grad_norm": 2.725398354282957,
+      "learning_rate": 7.23404255319149e-05,
+      "loss": 4.7328,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21729183293429102,
+      "grad_norm": 1.982400092225215,
+      "learning_rate": 7.24113475177305e-05,
+      "loss": 4.7265,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2175046554934823,
+      "grad_norm": 2.614966111288543,
+      "learning_rate": 7.248226950354611e-05,
+      "loss": 4.7276,
+      "step": 1022
+    },
+    {
+      "epoch": 0.21771747805267358,
+      "grad_norm": 2.160296381157991,
+      "learning_rate": 7.25531914893617e-05,
+      "loss": 4.7771,
+      "step": 1023
+    },
+    {
+      "epoch": 0.21793030061186486,
+      "grad_norm": 2.237118549931467,
+      "learning_rate": 7.262411347517731e-05,
+      "loss": 4.7784,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21814312317105614,
+      "grad_norm": 1.9973579357045235,
+      "learning_rate": 7.269503546099291e-05,
+      "loss": 4.7431,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2183559457302474,
+      "grad_norm": 2.5569212940856185,
+      "learning_rate": 7.276595744680852e-05,
+      "loss": 4.7587,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2185687682894387,
+      "grad_norm": 1.9270342661313047,
+      "learning_rate": 7.283687943262411e-05,
+      "loss": 4.727,
+      "step": 1027
+    },
+    {
+      "epoch": 0.21878159084862994,
+      "grad_norm": 2.655753102682302,
+      "learning_rate": 7.290780141843972e-05,
+      "loss": 4.7331,
+      "step": 1028
+    },
+    {
+      "epoch": 0.21899441340782122,
+      "grad_norm": 1.8506534724205232,
+      "learning_rate": 7.297872340425533e-05,
+      "loss": 4.7146,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2192072359670125,
+      "grad_norm": 2.553997009674833,
+      "learning_rate": 7.304964539007093e-05,
+      "loss": 4.755,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21942005852620378,
+      "grad_norm": 2.1125489196914766,
+      "learning_rate": 7.312056737588653e-05,
+      "loss": 4.7756,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21963288108539505,
+      "grad_norm": 2.3716561550979107,
+      "learning_rate": 7.319148936170213e-05,
+      "loss": 4.7401,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21984570364458633,
+      "grad_norm": 2.119049115425343,
+      "learning_rate": 7.326241134751774e-05,
+      "loss": 4.7277,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2200585262037776,
+      "grad_norm": 2.092932638080152,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 4.7244,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2202713487629689,
+      "grad_norm": 2.2023690193379792,
+      "learning_rate": 7.340425531914894e-05,
+      "loss": 4.7084,
+      "step": 1035
+    },
+    {
+      "epoch": 0.22048417132216014,
+      "grad_norm": 2.0688579695368685,
+      "learning_rate": 7.347517730496454e-05,
+      "loss": 4.7167,
+      "step": 1036
+    },
+    {
+      "epoch": 0.22069699388135142,
+      "grad_norm": 2.196373477454418,
+      "learning_rate": 7.354609929078015e-05,
+      "loss": 4.7518,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2209098164405427,
+      "grad_norm": 2.052440112474495,
+      "learning_rate": 7.361702127659574e-05,
+      "loss": 4.736,
+      "step": 1038
+    },
+    {
+      "epoch": 0.22112263899973397,
+      "grad_norm": 2.6604672364104758,
+      "learning_rate": 7.368794326241135e-05,
+      "loss": 4.7229,
+      "step": 1039
+    },
+    {
+      "epoch": 0.22133546155892525,
+      "grad_norm": 2.2749724971866763,
+      "learning_rate": 7.375886524822694e-05,
+      "loss": 4.7141,
+      "step": 1040
+    },
+    {
+      "epoch": 0.22154828411811653,
+      "grad_norm": 2.0883372018566804,
+      "learning_rate": 7.382978723404256e-05,
+      "loss": 4.6884,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2217611066773078,
+      "grad_norm": 2.2430832530389773,
+      "learning_rate": 7.390070921985816e-05,
+      "loss": 4.7123,
+      "step": 1042
+    },
+    {
+      "epoch": 0.22197392923649906,
+      "grad_norm": 2.0481875796013944,
+      "learning_rate": 7.397163120567376e-05,
+      "loss": 4.6892,
+      "step": 1043
+    },
+    {
+      "epoch": 0.22218675179569033,
+      "grad_norm": 2.77965046609862,
+      "learning_rate": 7.404255319148935e-05,
+      "loss": 4.7509,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2223995743548816,
+      "grad_norm": 1.6852008187052705,
+      "learning_rate": 7.411347517730497e-05,
+      "loss": 4.7687,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2226123969140729,
+      "grad_norm": 2.2757244037762416,
+      "learning_rate": 7.418439716312057e-05,
+      "loss": 4.7583,
+      "step": 1046
+    },
+    {
+      "epoch": 0.22282521947326417,
+      "grad_norm": 1.7981338980588462,
+      "learning_rate": 7.425531914893617e-05,
+      "loss": 4.6917,
+      "step": 1047
+    },
+    {
+      "epoch": 0.22303804203245545,
+      "grad_norm": 2.3376484957424895,
+      "learning_rate": 7.432624113475177e-05,
+      "loss": 4.6915,
+      "step": 1048
+    },
+    {
+      "epoch": 0.22325086459164672,
+      "grad_norm": 2.2894176584214017,
+      "learning_rate": 7.439716312056739e-05,
+      "loss": 4.7149,
+      "step": 1049
+    },
+    {
+      "epoch": 0.22346368715083798,
+      "grad_norm": 1.7521096079341707,
+      "learning_rate": 7.446808510638298e-05,
+      "loss": 4.7262,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22367650971002925,
+      "grad_norm": 2.6418891413638295,
+      "learning_rate": 7.453900709219859e-05,
+      "loss": 4.7319,
+      "step": 1051
+    },
+    {
+      "epoch": 0.22388933226922053,
+      "grad_norm": 1.673508849576304,
+      "learning_rate": 7.460992907801418e-05,
+      "loss": 4.6968,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2241021548284118,
+      "grad_norm": 2.4061519347182827,
+      "learning_rate": 7.46808510638298e-05,
+      "loss": 4.7199,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2243149773876031,
+      "grad_norm": 2.611800772341198,
+      "learning_rate": 7.475177304964539e-05,
+      "loss": 4.7127,
+      "step": 1054
+    },
+    {
+      "epoch": 0.22452779994679437,
+      "grad_norm": 1.9550811917312076,
+      "learning_rate": 7.4822695035461e-05,
+      "loss": 4.7117,
+      "step": 1055
+    },
+    {
+      "epoch": 0.22474062250598564,
+      "grad_norm": 2.296103313058607,
+      "learning_rate": 7.489361702127659e-05,
+      "loss": 4.7052,
+      "step": 1056
+    },
+    {
+      "epoch": 0.22495344506517692,
+      "grad_norm": 2.463121387048863,
+      "learning_rate": 7.496453900709221e-05,
+      "loss": 4.6636,
+      "step": 1057
+    },
+    {
+      "epoch": 0.22516626762436817,
+      "grad_norm": 1.815583150995484,
+      "learning_rate": 7.50354609929078e-05,
+      "loss": 4.7483,
+      "step": 1058
+    },
+    {
+      "epoch": 0.22537909018355945,
+      "grad_norm": 2.400434552109923,
+      "learning_rate": 7.510638297872341e-05,
+      "loss": 4.725,
+      "step": 1059
+    },
+    {
+      "epoch": 0.22559191274275073,
+      "grad_norm": 2.328674244530798,
+      "learning_rate": 7.5177304964539e-05,
+      "loss": 4.6904,
+      "step": 1060
+    },
+    {
+      "epoch": 0.225804735301942,
+      "grad_norm": 2.210552221323414,
+      "learning_rate": 7.524822695035462e-05,
+      "loss": 4.7118,
+      "step": 1061
+    },
+    {
+      "epoch": 0.22601755786113328,
+      "grad_norm": 2.4776221908573914,
+      "learning_rate": 7.531914893617022e-05,
+      "loss": 4.6758,
+      "step": 1062
+    },
+    {
+      "epoch": 0.22623038042032456,
+      "grad_norm": 1.8970446518918476,
+      "learning_rate": 7.539007092198582e-05,
+      "loss": 4.7302,
+      "step": 1063
+    },
+    {
+      "epoch": 0.22644320297951584,
+      "grad_norm": 2.5419741804545146,
+      "learning_rate": 7.546099290780141e-05,
+      "loss": 4.7072,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2266560255387071,
+      "grad_norm": 2.2439067998690607,
+      "learning_rate": 7.553191489361703e-05,
+      "loss": 4.6951,
+      "step": 1065
+    },
+    {
+      "epoch": 0.22686884809789837,
+      "grad_norm": 2.3453745638753647,
+      "learning_rate": 7.560283687943263e-05,
+      "loss": 4.6176,
+      "step": 1066
+    },
+    {
+      "epoch": 0.22708167065708965,
+      "grad_norm": 1.8200976961628224,
+      "learning_rate": 7.567375886524823e-05,
+      "loss": 4.693,
+      "step": 1067
+    },
+    {
+      "epoch": 0.22729449321628092,
+      "grad_norm": 2.713878752175517,
+      "learning_rate": 7.574468085106383e-05,
+      "loss": 4.6875,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2275073157754722,
+      "grad_norm": 1.8747129168622003,
+      "learning_rate": 7.581560283687945e-05,
+      "loss": 4.6842,
+      "step": 1069
+    },
+    {
+      "epoch": 0.22772013833466348,
+      "grad_norm": 2.5756172695116786,
+      "learning_rate": 7.588652482269504e-05,
+      "loss": 4.6961,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22793296089385476,
+      "grad_norm": 2.0034681210936056,
+      "learning_rate": 7.595744680851065e-05,
+      "loss": 4.6669,
+      "step": 1071
+    },
+    {
+      "epoch": 0.22814578345304604,
+      "grad_norm": 3.070477670347369,
+      "learning_rate": 7.602836879432624e-05,
+      "loss": 4.7177,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2283586060122373,
+      "grad_norm": 1.8063071822257097,
+      "learning_rate": 7.609929078014186e-05,
+      "loss": 4.6808,
+      "step": 1073
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 2.627429278269945,
+      "learning_rate": 7.617021276595745e-05,
+      "loss": 4.6973,
+      "step": 1074
+    },
+    {
+      "epoch": 0.22878425113061984,
+      "grad_norm": 2.131508577335705,
+      "learning_rate": 7.624113475177306e-05,
+      "loss": 4.673,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22899707368981112,
+      "grad_norm": 2.02149956570449,
+      "learning_rate": 7.631205673758865e-05,
+      "loss": 4.6722,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2292098962490024,
+      "grad_norm": 2.0228839576912314,
+      "learning_rate": 7.638297872340426e-05,
+      "loss": 4.6737,
+      "step": 1077
+    },
+    {
+      "epoch": 0.22942271880819368,
+      "grad_norm": 2.6127337528859864,
+      "learning_rate": 7.645390070921986e-05,
+      "loss": 4.7095,
+      "step": 1078
+    },
+    {
+      "epoch": 0.22963554136738495,
+      "grad_norm": 1.8219331591472332,
+      "learning_rate": 7.652482269503547e-05,
+      "loss": 4.6733,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2298483639265762,
+      "grad_norm": 2.1056627983176366,
+      "learning_rate": 7.659574468085106e-05,
+      "loss": 4.6383,
+      "step": 1080
+    },
+    {
+      "epoch": 0.23006118648576748,
+      "grad_norm": 2.1508770703383058,
+      "learning_rate": 7.666666666666667e-05,
+      "loss": 4.7061,
+      "step": 1081
+    },
+    {
+      "epoch": 0.23027400904495876,
+      "grad_norm": 2.224625549530659,
+      "learning_rate": 7.673758865248228e-05,
+      "loss": 4.7048,
+      "step": 1082
+    },
+    {
+      "epoch": 0.23048683160415004,
+      "grad_norm": 2.507635717885509,
+      "learning_rate": 7.680851063829788e-05,
+      "loss": 4.6847,
+      "step": 1083
+    },
+    {
+      "epoch": 0.23069965416334132,
+      "grad_norm": 2.267300881717267,
+      "learning_rate": 7.687943262411347e-05,
+      "loss": 4.6758,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2309124767225326,
+      "grad_norm": 2.252846892131252,
+      "learning_rate": 7.695035460992908e-05,
+      "loss": 4.6875,
+      "step": 1085
+    },
+    {
+      "epoch": 0.23112529928172387,
+      "grad_norm": 2.0062158136213593,
+      "learning_rate": 7.702127659574469e-05,
+      "loss": 4.6685,
+      "step": 1086
+    },
+    {
+      "epoch": 0.23133812184091515,
+      "grad_norm": 2.519769626286259,
+      "learning_rate": 7.709219858156028e-05,
+      "loss": 4.7334,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2315509444001064,
+      "grad_norm": 1.9520291652511763,
+      "learning_rate": 7.716312056737589e-05,
+      "loss": 4.6757,
+      "step": 1088
+    },
+    {
+      "epoch": 0.23176376695929768,
+      "grad_norm": 2.565319293960427,
+      "learning_rate": 7.723404255319149e-05,
+      "loss": 4.681,
+      "step": 1089
+    },
+    {
+      "epoch": 0.23197658951848896,
+      "grad_norm": 2.0403782802225847,
+      "learning_rate": 7.73049645390071e-05,
+      "loss": 4.6948,
+      "step": 1090
+    },
+    {
+      "epoch": 0.23218941207768024,
+      "grad_norm": 2.426667885924509,
+      "learning_rate": 7.737588652482269e-05,
+      "loss": 4.6654,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2324022346368715,
+      "grad_norm": 2.245652686340649,
+      "learning_rate": 7.74468085106383e-05,
+      "loss": 4.6431,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2326150571960628,
+      "grad_norm": 1.8875049785884286,
+      "learning_rate": 7.75177304964539e-05,
+      "loss": 4.6459,
+      "step": 1093
+    },
+    {
+      "epoch": 0.23282787975525407,
+      "grad_norm": 2.177433432518013,
+      "learning_rate": 7.758865248226951e-05,
+      "loss": 4.6745,
+      "step": 1094
+    },
+    {
+      "epoch": 0.23304070231444532,
+      "grad_norm": 1.7784640005168257,
+      "learning_rate": 7.76595744680851e-05,
+      "loss": 4.6229,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2332535248736366,
+      "grad_norm": 2.3045607155676175,
+      "learning_rate": 7.773049645390071e-05,
+      "loss": 4.679,
+      "step": 1096
+    },
+    {
+      "epoch": 0.23346634743282788,
+      "grad_norm": 2.035871193209823,
+      "learning_rate": 7.780141843971632e-05,
+      "loss": 4.6297,
+      "step": 1097
+    },
+    {
+      "epoch": 0.23367916999201915,
+      "grad_norm": 2.177858330847088,
+      "learning_rate": 7.787234042553192e-05,
+      "loss": 4.6208,
+      "step": 1098
+    },
+    {
+      "epoch": 0.23389199255121043,
+      "grad_norm": 2.609942479709796,
+      "learning_rate": 7.794326241134752e-05,
+      "loss": 4.6254,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2341048151104017,
+      "grad_norm": 2.5546915955886034,
+      "learning_rate": 7.801418439716312e-05,
+      "loss": 4.6666,
+      "step": 1100
+    },
+    {
+      "epoch": 0.234317637669593,
+      "grad_norm": 2.129240777043863,
+      "learning_rate": 7.808510638297873e-05,
+      "loss": 4.5621,
+      "step": 1101
+    },
+    {
+      "epoch": 0.23453046022878424,
+      "grad_norm": 1.8264730187673297,
+      "learning_rate": 7.815602836879434e-05,
+      "loss": 4.6889,
+      "step": 1102
+    },
+    {
+      "epoch": 0.23474328278797552,
+      "grad_norm": 2.4527799883336656,
+      "learning_rate": 7.822695035460993e-05,
+      "loss": 4.6632,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2349561053471668,
+      "grad_norm": 1.8525185702596993,
+      "learning_rate": 7.829787234042553e-05,
+      "loss": 4.6249,
+      "step": 1104
+    },
+    {
+      "epoch": 0.23516892790635807,
+      "grad_norm": 2.3317946249638792,
+      "learning_rate": 7.836879432624114e-05,
+      "loss": 4.6159,
+      "step": 1105
+    },
+    {
+      "epoch": 0.23538175046554935,
+      "grad_norm": 2.023129419067022,
+      "learning_rate": 7.843971631205675e-05,
+      "loss": 4.6494,
+      "step": 1106
+    },
+    {
+      "epoch": 0.23559457302474063,
+      "grad_norm": 2.052815393213699,
+      "learning_rate": 7.851063829787234e-05,
+      "loss": 4.6016,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2358073955839319,
+      "grad_norm": 2.256224668112829,
+      "learning_rate": 7.858156028368795e-05,
+      "loss": 4.6783,
+      "step": 1108
+    },
+    {
+      "epoch": 0.23602021814312318,
+      "grad_norm": 2.1841383282018842,
+      "learning_rate": 7.865248226950355e-05,
+      "loss": 4.6309,
+      "step": 1109
+    },
+    {
+      "epoch": 0.23623304070231443,
+      "grad_norm": 2.1606005949759997,
+      "learning_rate": 7.872340425531916e-05,
+      "loss": 4.6195,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2364458632615057,
+      "grad_norm": 2.1962539218357433,
+      "learning_rate": 7.879432624113475e-05,
+      "loss": 4.6164,
+      "step": 1111
+    },
+    {
+      "epoch": 0.236658685820697,
+      "grad_norm": 2.3445349913299594,
+      "learning_rate": 7.886524822695036e-05,
+      "loss": 4.6662,
+      "step": 1112
+    },
+    {
+      "epoch": 0.23687150837988827,
+      "grad_norm": 2.0606210432871364,
+      "learning_rate": 7.893617021276596e-05,
+      "loss": 4.6867,
+      "step": 1113
+    },
+    {
+      "epoch": 0.23708433093907955,
+      "grad_norm": 2.1699657291839958,
+      "learning_rate": 7.900709219858157e-05,
+      "loss": 4.5914,
+      "step": 1114
+    },
+    {
+      "epoch": 0.23729715349827082,
+      "grad_norm": 2.526386165100392,
+      "learning_rate": 7.907801418439716e-05,
+      "loss": 4.6057,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2375099760574621,
+      "grad_norm": 1.9687626325461836,
+      "learning_rate": 7.914893617021277e-05,
+      "loss": 4.6431,
+      "step": 1116
+    },
+    {
+      "epoch": 0.23772279861665335,
+      "grad_norm": 2.302847799792408,
+      "learning_rate": 7.921985815602838e-05,
+      "loss": 4.6511,
+      "step": 1117
+    },
+    {
+      "epoch": 0.23793562117584463,
+      "grad_norm": 2.0529016478812725,
+      "learning_rate": 7.929078014184398e-05,
+      "loss": 4.641,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2381484437350359,
+      "grad_norm": 2.9869117019870335,
+      "learning_rate": 7.936170212765958e-05,
+      "loss": 4.6298,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2383612662942272,
+      "grad_norm": 1.6678243957171466,
+      "learning_rate": 7.943262411347518e-05,
+      "loss": 4.6132,
+      "step": 1120
+    },
+    {
+      "epoch": 0.23857408885341846,
+      "grad_norm": 2.9733821935043125,
+      "learning_rate": 7.950354609929078e-05,
+      "loss": 4.6068,
+      "step": 1121
+    },
+    {
+      "epoch": 0.23878691141260974,
+      "grad_norm": 1.6532472596433818,
+      "learning_rate": 7.95744680851064e-05,
+      "loss": 4.6534,
+      "step": 1122
+    },
+    {
+      "epoch": 0.23899973397180102,
+      "grad_norm": 2.5703892938359725,
+      "learning_rate": 7.964539007092199e-05,
+      "loss": 4.6293,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2392125565309923,
+      "grad_norm": 1.9391533364650138,
+      "learning_rate": 7.97163120567376e-05,
+      "loss": 4.5916,
+      "step": 1124
+    },
+    {
+      "epoch": 0.23942537909018355,
+      "grad_norm": 2.292718443329716,
+      "learning_rate": 7.978723404255319e-05,
+      "loss": 4.662,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23963820164937483,
+      "grad_norm": 2.846624484801799,
+      "learning_rate": 7.985815602836881e-05,
+      "loss": 4.6252,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2398510242085661,
+      "grad_norm": 1.8424715537618859,
+      "learning_rate": 7.99290780141844e-05,
+      "loss": 4.5834,
+      "step": 1127
+    },
+    {
+      "epoch": 0.24006384676775738,
+      "grad_norm": 1.9419244827891768,
+      "learning_rate": 8e-05,
+      "loss": 4.6302,
+      "step": 1128
+    },
+    {
+      "epoch": 0.24027666932694866,
+      "grad_norm": 2.238153366708233,
+      "learning_rate": 8.00709219858156e-05,
+      "loss": 4.6215,
+      "step": 1129
+    },
+    {
+      "epoch": 0.24048949188613994,
+      "grad_norm": 2.096304054955189,
+      "learning_rate": 8.014184397163122e-05,
+      "loss": 4.5911,
+      "step": 1130
+    },
+    {
+      "epoch": 0.24070231444533122,
+      "grad_norm": 2.2898842813198934,
+      "learning_rate": 8.021276595744681e-05,
+      "loss": 4.6076,
+      "step": 1131
+    },
+    {
+      "epoch": 0.24091513700452247,
+      "grad_norm": 2.58091080254549,
+      "learning_rate": 8.028368794326242e-05,
+      "loss": 4.6103,
+      "step": 1132
+    },
+    {
+      "epoch": 0.24112795956371375,
+      "grad_norm": 2.1075839385526587,
+      "learning_rate": 8.035460992907801e-05,
+      "loss": 4.6238,
+      "step": 1133
+    },
+    {
+      "epoch": 0.24134078212290502,
+      "grad_norm": 2.776815007954266,
+      "learning_rate": 8.042553191489363e-05,
+      "loss": 4.5948,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2415536046820963,
+      "grad_norm": 1.9776369782882979,
+      "learning_rate": 8.049645390070922e-05,
+      "loss": 4.6738,
+      "step": 1135
+    },
+    {
+      "epoch": 0.24176642724128758,
+      "grad_norm": 2.151477668316568,
+      "learning_rate": 8.056737588652483e-05,
+      "loss": 4.5776,
+      "step": 1136
+    },
+    {
+      "epoch": 0.24197924980047886,
+      "grad_norm": 1.9266178004015626,
+      "learning_rate": 8.063829787234042e-05,
+      "loss": 4.6053,
+      "step": 1137
+    },
+    {
+      "epoch": 0.24219207235967014,
+      "grad_norm": 2.148079427810053,
+      "learning_rate": 8.070921985815604e-05,
+      "loss": 4.624,
+      "step": 1138
+    },
+    {
+      "epoch": 0.24240489491886139,
+      "grad_norm": 1.9710762457629334,
+      "learning_rate": 8.078014184397164e-05,
+      "loss": 4.6049,
+      "step": 1139
+    },
+    {
+      "epoch": 0.24261771747805266,
+      "grad_norm": 2.1231323213289226,
+      "learning_rate": 8.085106382978723e-05,
+      "loss": 4.5629,
+      "step": 1140
+    },
+    {
+      "epoch": 0.24283054003724394,
+      "grad_norm": 2.1690762396564387,
+      "learning_rate": 8.092198581560284e-05,
+      "loss": 4.5935,
+      "step": 1141
+    },
+    {
+      "epoch": 0.24304336259643522,
+      "grad_norm": 2.2909234444666944,
+      "learning_rate": 8.099290780141844e-05,
+      "loss": 4.5739,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2432561851556265,
+      "grad_norm": 2.151651792819353,
+      "learning_rate": 8.106382978723405e-05,
+      "loss": 4.558,
+      "step": 1143
+    },
+    {
+      "epoch": 0.24346900771481778,
+      "grad_norm": 2.2440400759776997,
+      "learning_rate": 8.113475177304964e-05,
+      "loss": 4.5978,
+      "step": 1144
+    },
+    {
+      "epoch": 0.24368183027400905,
+      "grad_norm": 1.8424789370399344,
+      "learning_rate": 8.120567375886525e-05,
+      "loss": 4.5753,
+      "step": 1145
+    },
+    {
+      "epoch": 0.24389465283320033,
+      "grad_norm": 2.585810613722092,
+      "learning_rate": 8.127659574468085e-05,
+      "loss": 4.5515,
+      "step": 1146
+    },
+    {
+      "epoch": 0.24410747539239158,
+      "grad_norm": 1.7151903884104605,
+      "learning_rate": 8.134751773049646e-05,
+      "loss": 4.6172,
+      "step": 1147
+    },
+    {
+      "epoch": 0.24432029795158286,
+      "grad_norm": 2.6863542354201018,
+      "learning_rate": 8.141843971631205e-05,
+      "loss": 4.6336,
+      "step": 1148
+    },
+    {
+      "epoch": 0.24453312051077414,
+      "grad_norm": 1.5990285237285065,
+      "learning_rate": 8.148936170212766e-05,
+      "loss": 4.5585,
+      "step": 1149
+    },
+    {
+      "epoch": 0.24474594306996542,
+      "grad_norm": 2.6441643018612258,
+      "learning_rate": 8.156028368794327e-05,
+      "loss": 4.6171,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2449587656291567,
+      "grad_norm": 2.0648405184926997,
+      "learning_rate": 8.163120567375887e-05,
+      "loss": 4.5387,
+      "step": 1151
+    },
+    {
+      "epoch": 0.24517158818834797,
+      "grad_norm": 2.066531582524445,
+      "learning_rate": 8.170212765957446e-05,
+      "loss": 4.5887,
+      "step": 1152
+    },
+    {
+      "epoch": 0.24538441074753925,
+      "grad_norm": 2.1761248746920496,
+      "learning_rate": 8.177304964539007e-05,
+      "loss": 4.6027,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2455972333067305,
+      "grad_norm": 1.8933279222030766,
+      "learning_rate": 8.184397163120568e-05,
+      "loss": 4.5566,
+      "step": 1154
+    },
+    {
+      "epoch": 0.24581005586592178,
+      "grad_norm": 2.0277911819393886,
+      "learning_rate": 8.191489361702128e-05,
+      "loss": 4.5869,
+      "step": 1155
+    },
+    {
+      "epoch": 0.24602287842511306,
+      "grad_norm": 2.077451900424791,
+      "learning_rate": 8.198581560283688e-05,
+      "loss": 4.5964,
+      "step": 1156
+    },
+    {
+      "epoch": 0.24623570098430433,
+      "grad_norm": 2.410944957993582,
+      "learning_rate": 8.205673758865248e-05,
+      "loss": 4.5742,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2464485235434956,
+      "grad_norm": 2.0563977003929574,
+      "learning_rate": 8.212765957446809e-05,
+      "loss": 4.5579,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2466613461026869,
+      "grad_norm": 1.6825394703909005,
+      "learning_rate": 8.21985815602837e-05,
+      "loss": 4.5036,
+      "step": 1159
+    },
+    {
+      "epoch": 0.24687416866187817,
+      "grad_norm": 2.2297273202804124,
+      "learning_rate": 8.226950354609929e-05,
+      "loss": 4.5624,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24708699122106945,
+      "grad_norm": 2.03982615600575,
+      "learning_rate": 8.23404255319149e-05,
+      "loss": 4.5597,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2472998137802607,
+      "grad_norm": 2.2656176185953485,
+      "learning_rate": 8.24113475177305e-05,
+      "loss": 4.5713,
+      "step": 1162
+    },
+    {
+      "epoch": 0.24751263633945197,
+      "grad_norm": 2.3807675696815025,
+      "learning_rate": 8.248226950354611e-05,
+      "loss": 4.5563,
+      "step": 1163
+    },
+    {
+      "epoch": 0.24772545889864325,
+      "grad_norm": 2.0823799790364395,
+      "learning_rate": 8.25531914893617e-05,
+      "loss": 4.5518,
+      "step": 1164
+    },
+    {
+      "epoch": 0.24793828145783453,
+      "grad_norm": 2.1229382007769058,
+      "learning_rate": 8.262411347517731e-05,
+      "loss": 4.5515,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2481511040170258,
+      "grad_norm": 2.5922931042241215,
+      "learning_rate": 8.269503546099291e-05,
+      "loss": 4.5853,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2483639265762171,
+      "grad_norm": 2.2498962026930567,
+      "learning_rate": 8.276595744680852e-05,
+      "loss": 4.5445,
+      "step": 1167
+    },
+    {
+      "epoch": 0.24857674913540836,
+      "grad_norm": 1.9593985436467805,
+      "learning_rate": 8.283687943262411e-05,
+      "loss": 4.5279,
+      "step": 1168
+    },
+    {
+      "epoch": 0.24878957169459961,
+      "grad_norm": 2.504471562820281,
+      "learning_rate": 8.290780141843972e-05,
+      "loss": 4.54,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2490023942537909,
+      "grad_norm": 1.8200015518518415,
+      "learning_rate": 8.297872340425533e-05,
+      "loss": 4.5482,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24921521681298217,
+      "grad_norm": 2.259754801703336,
+      "learning_rate": 8.304964539007093e-05,
+      "loss": 4.5786,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24942803937217345,
+      "grad_norm": 2.0808829021053006,
+      "learning_rate": 8.312056737588652e-05,
+      "loss": 4.562,
+      "step": 1172
+    },
+    {
+      "epoch": 0.24964086193136473,
+      "grad_norm": 2.9630394983059483,
+      "learning_rate": 8.319148936170213e-05,
+      "loss": 4.5415,
+      "step": 1173
+    },
+    {
+      "epoch": 0.249853684490556,
+      "grad_norm": 1.6982777234508344,
+      "learning_rate": 8.326241134751774e-05,
+      "loss": 4.5594,
+      "step": 1174
+    },
+    {
+      "epoch": 0.25006650704974726,
+      "grad_norm": 2.6549928179121602,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 4.5603,
+      "step": 1175
+    },
+    {
+      "epoch": 0.25027932960893856,
+      "grad_norm": 1.8497569401556213,
+      "learning_rate": 8.340425531914894e-05,
+      "loss": 4.5579,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2504921521681298,
+      "grad_norm": 2.479441994277811,
+      "learning_rate": 8.347517730496454e-05,
+      "loss": 4.5457,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2507049747273211,
+      "grad_norm": 2.587567938062552,
+      "learning_rate": 8.354609929078015e-05,
+      "loss": 4.5482,
+      "step": 1178
+    },
+    {
+      "epoch": 0.25091779728651237,
+      "grad_norm": 1.8612065145542658,
+      "learning_rate": 8.361702127659576e-05,
+      "loss": 4.5325,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2511306198457036,
+      "grad_norm": 2.2279023296088334,
+      "learning_rate": 8.368794326241135e-05,
+      "loss": 4.5571,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2513434424048949,
+      "grad_norm": 1.8059657780413936,
+      "learning_rate": 8.375886524822696e-05,
+      "loss": 4.5578,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2515562649640862,
+      "grad_norm": 2.0567974706194376,
+      "learning_rate": 8.382978723404256e-05,
+      "loss": 4.5335,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2517690875232775,
+      "grad_norm": 1.925732722114557,
+      "learning_rate": 8.390070921985817e-05,
+      "loss": 4.5459,
+      "step": 1183
+    },
+    {
+      "epoch": 0.25198191008246873,
+      "grad_norm": 2.2549919717929807,
+      "learning_rate": 8.397163120567376e-05,
+      "loss": 4.5298,
+      "step": 1184
+    },
+    {
+      "epoch": 0.25219473264166004,
+      "grad_norm": 2.2976117582979247,
+      "learning_rate": 8.404255319148937e-05,
+      "loss": 4.528,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2524075552008513,
+      "grad_norm": 2.4719715673936427,
+      "learning_rate": 8.411347517730497e-05,
+      "loss": 4.577,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2526203777600426,
+      "grad_norm": 2.1977918348997987,
+      "learning_rate": 8.418439716312058e-05,
+      "loss": 4.5436,
+      "step": 1187
+    },
+    {
+      "epoch": 0.25283320031923384,
+      "grad_norm": 2.0369065787627165,
+      "learning_rate": 8.425531914893617e-05,
+      "loss": 4.5408,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2530460228784251,
+      "grad_norm": 2.227242600092366,
+      "learning_rate": 8.432624113475178e-05,
+      "loss": 4.5354,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2532588454376164,
+      "grad_norm": 1.7991850380658947,
+      "learning_rate": 8.439716312056739e-05,
+      "loss": 4.5333,
+      "step": 1190
+    },
+    {
+      "epoch": 0.25347166799680765,
+      "grad_norm": 2.290640013275384,
+      "learning_rate": 8.446808510638299e-05,
+      "loss": 4.5218,
+      "step": 1191
+    },
+    {
+      "epoch": 0.25368449055599895,
+      "grad_norm": 2.1708142358338587,
+      "learning_rate": 8.453900709219858e-05,
+      "loss": 4.5434,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2538973131151902,
+      "grad_norm": 2.221740649234827,
+      "learning_rate": 8.460992907801419e-05,
+      "loss": 4.5486,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2541101356743815,
+      "grad_norm": 2.017289862162329,
+      "learning_rate": 8.46808510638298e-05,
+      "loss": 4.518,
+      "step": 1194
+    },
+    {
+      "epoch": 0.25432295823357276,
+      "grad_norm": 2.3938812948344412,
+      "learning_rate": 8.475177304964539e-05,
+      "loss": 4.561,
+      "step": 1195
+    },
+    {
+      "epoch": 0.254535780792764,
+      "grad_norm": 2.030254051010367,
+      "learning_rate": 8.4822695035461e-05,
+      "loss": 4.5774,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2547486033519553,
+      "grad_norm": 1.8942950084814238,
+      "learning_rate": 8.489361702127659e-05,
+      "loss": 4.5884,
+      "step": 1197
+    },
+    {
+      "epoch": 0.25496142591114657,
+      "grad_norm": 1.7506886283301966,
+      "learning_rate": 8.496453900709221e-05,
+      "loss": 4.5066,
+      "step": 1198
+    },
+    {
+      "epoch": 0.25517424847033787,
+      "grad_norm": 2.3960722323443537,
+      "learning_rate": 8.50354609929078e-05,
+      "loss": 4.528,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2553870710295291,
+      "grad_norm": 1.9497570330046727,
+      "learning_rate": 8.510638297872341e-05,
+      "loss": 4.547,
+      "step": 1200
+    },
+    {
+      "epoch": 0.25559989358872043,
+      "grad_norm": 2.1888175272091437,
+      "learning_rate": 8.5177304964539e-05,
+      "loss": 4.4669,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2558127161479117,
+      "grad_norm": 1.8943928759207451,
+      "learning_rate": 8.524822695035462e-05,
+      "loss": 4.4776,
+      "step": 1202
+    },
+    {
+      "epoch": 0.25602553870710293,
+      "grad_norm": 2.5939391713362103,
+      "learning_rate": 8.531914893617021e-05,
+      "loss": 4.5406,
+      "step": 1203
+    },
+    {
+      "epoch": 0.25623836126629423,
+      "grad_norm": 1.6570314864140978,
+      "learning_rate": 8.539007092198582e-05,
+      "loss": 4.5336,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2564511838254855,
+      "grad_norm": 2.5556720106520117,
+      "learning_rate": 8.546099290780141e-05,
+      "loss": 4.5064,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2566640063846768,
+      "grad_norm": 2.065893049296645,
+      "learning_rate": 8.553191489361702e-05,
+      "loss": 4.5387,
+      "step": 1206
+    },
+    {
+      "epoch": 0.25687682894386804,
+      "grad_norm": 2.4073877742150476,
+      "learning_rate": 8.560283687943263e-05,
+      "loss": 4.4848,
+      "step": 1207
+    },
+    {
+      "epoch": 0.25708965150305935,
+      "grad_norm": 2.0720192852690285,
+      "learning_rate": 8.567375886524823e-05,
+      "loss": 4.5155,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2573024740622506,
+      "grad_norm": 2.3659628367642602,
+      "learning_rate": 8.574468085106383e-05,
+      "loss": 4.5214,
+      "step": 1209
+    },
+    {
+      "epoch": 0.25751529662144185,
+      "grad_norm": 2.021071429628008,
+      "learning_rate": 8.581560283687943e-05,
+      "loss": 4.4946,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25772811918063315,
+      "grad_norm": 1.913874265082252,
+      "learning_rate": 8.588652482269504e-05,
+      "loss": 4.5179,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2579409417398244,
+      "grad_norm": 2.0432492659516637,
+      "learning_rate": 8.595744680851064e-05,
+      "loss": 4.4772,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2581537642990157,
+      "grad_norm": 2.2070943542299664,
+      "learning_rate": 8.602836879432624e-05,
+      "loss": 4.4707,
+      "step": 1213
+    },
+    {
+      "epoch": 0.25836658685820696,
+      "grad_norm": 1.7479620854372533,
+      "learning_rate": 8.609929078014184e-05,
+      "loss": 4.4997,
+      "step": 1214
+    },
+    {
+      "epoch": 0.25857940941739826,
+      "grad_norm": 2.2937319418323967,
+      "learning_rate": 8.617021276595745e-05,
+      "loss": 4.4894,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2587922319765895,
+      "grad_norm": 1.7020648291421276,
+      "learning_rate": 8.624113475177306e-05,
+      "loss": 4.5187,
+      "step": 1216
+    },
+    {
+      "epoch": 0.25900505453578077,
+      "grad_norm": 2.7680018315436343,
+      "learning_rate": 8.631205673758865e-05,
+      "loss": 4.5003,
+      "step": 1217
+    },
+    {
+      "epoch": 0.25921787709497207,
+      "grad_norm": 1.6399976394522822,
+      "learning_rate": 8.638297872340426e-05,
+      "loss": 4.5138,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2594306996541633,
+      "grad_norm": 2.556121225493609,
+      "learning_rate": 8.645390070921986e-05,
+      "loss": 4.5094,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2596435222133546,
+      "grad_norm": 2.3445465652279616,
+      "learning_rate": 8.652482269503547e-05,
+      "loss": 4.5094,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2598563447725459,
+      "grad_norm": 1.840739991707624,
+      "learning_rate": 8.659574468085106e-05,
+      "loss": 4.5036,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2600691673317372,
+      "grad_norm": 2.301209417005153,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 4.5079,
+      "step": 1222
+    },
+    {
+      "epoch": 0.26028198989092843,
+      "grad_norm": 1.9467060974233428,
+      "learning_rate": 8.673758865248227e-05,
+      "loss": 4.5212,
+      "step": 1223
+    },
+    {
+      "epoch": 0.26049481245011974,
+      "grad_norm": 2.4416264420813927,
+      "learning_rate": 8.680851063829788e-05,
+      "loss": 4.4426,
+      "step": 1224
+    },
+    {
+      "epoch": 0.260707635009311,
+      "grad_norm": 1.982022914884236,
+      "learning_rate": 8.687943262411347e-05,
+      "loss": 4.4809,
+      "step": 1225
+    },
+    {
+      "epoch": 0.26092045756850224,
+      "grad_norm": 2.3579201433943244,
+      "learning_rate": 8.695035460992908e-05,
+      "loss": 4.4609,
+      "step": 1226
+    },
+    {
+      "epoch": 0.26113328012769355,
+      "grad_norm": 1.78924757780999,
+      "learning_rate": 8.702127659574469e-05,
+      "loss": 4.5194,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2613461026868848,
+      "grad_norm": 2.3526441202978816,
+      "learning_rate": 8.709219858156029e-05,
+      "loss": 4.4679,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2615589252460761,
+      "grad_norm": 1.9505391267540415,
+      "learning_rate": 8.716312056737589e-05,
+      "loss": 4.4993,
+      "step": 1229
+    },
+    {
+      "epoch": 0.26177174780526735,
+      "grad_norm": 2.0508483135038404,
+      "learning_rate": 8.723404255319149e-05,
+      "loss": 4.4781,
+      "step": 1230
+    },
+    {
+      "epoch": 0.26198457036445866,
+      "grad_norm": 2.0468338306608382,
+      "learning_rate": 8.73049645390071e-05,
+      "loss": 4.4437,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2621973929236499,
+      "grad_norm": 1.8148128657889993,
+      "learning_rate": 8.73758865248227e-05,
+      "loss": 4.4768,
+      "step": 1232
+    },
+    {
+      "epoch": 0.26241021548284116,
+      "grad_norm": 2.021373579082034,
+      "learning_rate": 8.74468085106383e-05,
+      "loss": 4.4512,
+      "step": 1233
+    },
+    {
+      "epoch": 0.26262303804203246,
+      "grad_norm": 1.938196225830006,
+      "learning_rate": 8.75177304964539e-05,
+      "loss": 4.4453,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2628358606012237,
+      "grad_norm": 2.3345435208069705,
+      "learning_rate": 8.758865248226951e-05,
+      "loss": 4.5113,
+      "step": 1235
+    },
+    {
+      "epoch": 0.263048683160415,
+      "grad_norm": 2.0571073123839763,
+      "learning_rate": 8.765957446808512e-05,
+      "loss": 4.4881,
+      "step": 1236
+    },
+    {
+      "epoch": 0.26326150571960627,
+      "grad_norm": 1.9829340838466438,
+      "learning_rate": 8.773049645390071e-05,
+      "loss": 4.4813,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2634743282787976,
+      "grad_norm": 2.370690129128863,
+      "learning_rate": 8.780141843971632e-05,
+      "loss": 4.471,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2636871508379888,
+      "grad_norm": 2.1358415471803713,
+      "learning_rate": 8.787234042553192e-05,
+      "loss": 4.4305,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2638999733971801,
+      "grad_norm": 2.2034927962901762,
+      "learning_rate": 8.794326241134753e-05,
+      "loss": 4.4313,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2641127959563714,
+      "grad_norm": 1.8965177416290921,
+      "learning_rate": 8.801418439716312e-05,
+      "loss": 4.4519,
+      "step": 1241
+    },
+    {
+      "epoch": 0.26432561851556263,
+      "grad_norm": 2.582522849508449,
+      "learning_rate": 8.808510638297873e-05,
+      "loss": 4.4691,
+      "step": 1242
+    },
+    {
+      "epoch": 0.26453844107475394,
+      "grad_norm": 1.7869778087070884,
+      "learning_rate": 8.815602836879433e-05,
+      "loss": 4.4724,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2647512636339452,
+      "grad_norm": 2.0695393418391306,
+      "learning_rate": 8.822695035460994e-05,
+      "loss": 4.4304,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2649640861931365,
+      "grad_norm": 2.0436290930916035,
+      "learning_rate": 8.829787234042553e-05,
+      "loss": 4.4461,
+      "step": 1245
+    },
+    {
+      "epoch": 0.26517690875232774,
+      "grad_norm": 2.1676631300231772,
+      "learning_rate": 8.836879432624114e-05,
+      "loss": 4.4339,
+      "step": 1246
+    },
+    {
+      "epoch": 0.265389731311519,
+      "grad_norm": 2.171465263905363,
+      "learning_rate": 8.843971631205675e-05,
+      "loss": 4.4059,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2656025538707103,
+      "grad_norm": 2.2075946775817554,
+      "learning_rate": 8.851063829787234e-05,
+      "loss": 4.4244,
+      "step": 1248
+    },
+    {
+      "epoch": 0.26581537642990155,
+      "grad_norm": 2.0426565257921423,
+      "learning_rate": 8.858156028368795e-05,
+      "loss": 4.4531,
+      "step": 1249
+    },
+    {
+      "epoch": 0.26602819898909286,
+      "grad_norm": 1.7368812547331358,
+      "learning_rate": 8.865248226950354e-05,
+      "loss": 4.4394,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2662410215482841,
+      "grad_norm": 2.1875341666222266,
+      "learning_rate": 8.872340425531916e-05,
+      "loss": 4.4341,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2664538441074754,
+      "grad_norm": 2.214784892381288,
+      "learning_rate": 8.879432624113475e-05,
+      "loss": 4.4848,
+      "step": 1252
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.784341062589949,
+      "learning_rate": 8.886524822695036e-05,
+      "loss": 4.4192,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2668794892258579,
+      "grad_norm": 2.0968839158183314,
+      "learning_rate": 8.893617021276595e-05,
+      "loss": 4.4249,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2670923117850492,
+      "grad_norm": 1.9557442925083117,
+      "learning_rate": 8.900709219858157e-05,
+      "loss": 4.4244,
+      "step": 1255
+    },
+    {
+      "epoch": 0.26730513434424047,
+      "grad_norm": 2.310558572346616,
+      "learning_rate": 8.907801418439716e-05,
+      "loss": 4.4118,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2675179569034318,
+      "grad_norm": 1.8109217801718724,
+      "learning_rate": 8.914893617021277e-05,
+      "loss": 4.4266,
+      "step": 1257
+    },
+    {
+      "epoch": 0.267730779462623,
+      "grad_norm": 2.4772314126988926,
+      "learning_rate": 8.921985815602836e-05,
+      "loss": 4.4214,
+      "step": 1258
+    },
+    {
+      "epoch": 0.26794360202181433,
+      "grad_norm": 1.8981365522948694,
+      "learning_rate": 8.929078014184398e-05,
+      "loss": 4.4018,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2681564245810056,
+      "grad_norm": 2.728881117704685,
+      "learning_rate": 8.936170212765958e-05,
+      "loss": 4.4639,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2683692471401969,
+      "grad_norm": 1.7660274272719563,
+      "learning_rate": 8.943262411347518e-05,
+      "loss": 4.4143,
+      "step": 1261
+    },
+    {
+      "epoch": 0.26858206969938814,
+      "grad_norm": 2.4998608613542572,
+      "learning_rate": 8.950354609929077e-05,
+      "loss": 4.4663,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2687948922585794,
+      "grad_norm": 2.1171977149693815,
+      "learning_rate": 8.95744680851064e-05,
+      "loss": 4.4126,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2690077148177707,
+      "grad_norm": 2.2924412017295266,
+      "learning_rate": 8.964539007092199e-05,
+      "loss": 4.4525,
+      "step": 1264
+    },
+    {
+      "epoch": 0.26922053737696194,
+      "grad_norm": 2.3825118860768595,
+      "learning_rate": 8.97163120567376e-05,
+      "loss": 4.3969,
+      "step": 1265
+    },
+    {
+      "epoch": 0.26943335993615325,
+      "grad_norm": 1.9103523622836265,
+      "learning_rate": 8.978723404255319e-05,
+      "loss": 4.4196,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2696461824953445,
+      "grad_norm": 2.21799708738235,
+      "learning_rate": 8.98581560283688e-05,
+      "loss": 4.4621,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2698590050545358,
+      "grad_norm": 2.1340835471598205,
+      "learning_rate": 8.99290780141844e-05,
+      "loss": 4.3998,
+      "step": 1268
+    },
+    {
+      "epoch": 0.27007182761372706,
+      "grad_norm": 1.732449658311025,
+      "learning_rate": 9e-05,
+      "loss": 4.4208,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2702846501729183,
+      "grad_norm": 2.129833829155536,
+      "learning_rate": 9.00709219858156e-05,
+      "loss": 4.4162,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2704974727321096,
+      "grad_norm": 2.240590267586227,
+      "learning_rate": 9.014184397163122e-05,
+      "loss": 4.4224,
+      "step": 1271
+    },
+    {
+      "epoch": 0.27071029529130086,
+      "grad_norm": 1.6539727721692077,
+      "learning_rate": 9.021276595744681e-05,
+      "loss": 4.4228,
+      "step": 1272
+    },
+    {
+      "epoch": 0.27092311785049217,
+      "grad_norm": 2.3674567908635322,
+      "learning_rate": 9.028368794326242e-05,
+      "loss": 4.4561,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2711359404096834,
+      "grad_norm": 1.78633588796918,
+      "learning_rate": 9.035460992907801e-05,
+      "loss": 4.3854,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2713487629688747,
+      "grad_norm": 2.471589134576028,
+      "learning_rate": 9.042553191489363e-05,
+      "loss": 4.4563,
+      "step": 1275
+    },
+    {
+      "epoch": 0.271561585528066,
+      "grad_norm": 1.6746319112500396,
+      "learning_rate": 9.049645390070922e-05,
+      "loss": 4.4047,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2717744080872572,
+      "grad_norm": 2.8019116893107214,
+      "learning_rate": 9.056737588652483e-05,
+      "loss": 4.4405,
+      "step": 1277
+    },
+    {
+      "epoch": 0.27198723064644853,
+      "grad_norm": 1.7702295782879944,
+      "learning_rate": 9.063829787234042e-05,
+      "loss": 4.4479,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2722000532056398,
+      "grad_norm": 2.6290617095153195,
+      "learning_rate": 9.070921985815604e-05,
+      "loss": 4.4386,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2724128757648311,
+      "grad_norm": 1.9616692097534432,
+      "learning_rate": 9.078014184397164e-05,
+      "loss": 4.3873,
+      "step": 1280
+    },
+    {
+      "epoch": 0.27262569832402234,
+      "grad_norm": 1.9208683302873366,
+      "learning_rate": 9.085106382978724e-05,
+      "loss": 4.4269,
+      "step": 1281
+    },
+    {
+      "epoch": 0.27283852088321364,
+      "grad_norm": 2.21269345892241,
+      "learning_rate": 9.092198581560283e-05,
+      "loss": 4.4183,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2730513434424049,
+      "grad_norm": 1.932919630538583,
+      "learning_rate": 9.099290780141845e-05,
+      "loss": 4.4044,
+      "step": 1283
+    },
+    {
+      "epoch": 0.27326416600159614,
+      "grad_norm": 1.8472748483484176,
+      "learning_rate": 9.106382978723405e-05,
+      "loss": 4.3577,
+      "step": 1284
+    },
+    {
+      "epoch": 0.27347698856078745,
+      "grad_norm": 2.3636309689148605,
+      "learning_rate": 9.113475177304965e-05,
+      "loss": 4.433,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2736898111199787,
+      "grad_norm": 2.1815932628520507,
+      "learning_rate": 9.120567375886525e-05,
+      "loss": 4.4162,
+      "step": 1286
+    },
+    {
+      "epoch": 0.27390263367917,
+      "grad_norm": 2.5027330556017646,
+      "learning_rate": 9.127659574468085e-05,
+      "loss": 4.4242,
+      "step": 1287
+    },
+    {
+      "epoch": 0.27411545623836125,
+      "grad_norm": 2.8100939322640106,
+      "learning_rate": 9.134751773049646e-05,
+      "loss": 4.4669,
+      "step": 1288
+    },
+    {
+      "epoch": 0.27432827879755256,
+      "grad_norm": 2.0505293015276527,
+      "learning_rate": 9.141843971631207e-05,
+      "loss": 4.4381,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2745411013567438,
+      "grad_norm": 1.8638203069935435,
+      "learning_rate": 9.148936170212766e-05,
+      "loss": 4.4536,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2747539239159351,
+      "grad_norm": 2.3492884136709486,
+      "learning_rate": 9.156028368794326e-05,
+      "loss": 4.389,
+      "step": 1291
+    },
+    {
+      "epoch": 0.27496674647512637,
+      "grad_norm": 2.036902994389397,
+      "learning_rate": 9.163120567375887e-05,
+      "loss": 4.3977,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2751795690343176,
+      "grad_norm": 1.832952660394472,
+      "learning_rate": 9.170212765957448e-05,
+      "loss": 4.4088,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2753923915935089,
+      "grad_norm": 1.9665885812479744,
+      "learning_rate": 9.177304964539007e-05,
+      "loss": 4.3959,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2756052141527002,
+      "grad_norm": 1.9670445056968109,
+      "learning_rate": 9.184397163120568e-05,
+      "loss": 4.377,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2758180367118915,
+      "grad_norm": 1.9455949553433813,
+      "learning_rate": 9.191489361702128e-05,
+      "loss": 4.4075,
+      "step": 1296
+    },
+    {
+      "epoch": 0.27603085927108273,
+      "grad_norm": 2.195674690864208,
+      "learning_rate": 9.198581560283689e-05,
+      "loss": 4.3596,
+      "step": 1297
+    },
+    {
+      "epoch": 0.27624368183027403,
+      "grad_norm": 1.8916935333733407,
+      "learning_rate": 9.205673758865248e-05,
+      "loss": 4.3332,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2764565043894653,
+      "grad_norm": 2.107906928849221,
+      "learning_rate": 9.212765957446809e-05,
+      "loss": 4.3786,
+      "step": 1299
+    },
+    {
+      "epoch": 0.27666932694865654,
+      "grad_norm": 2.1605243180317886,
+      "learning_rate": 9.21985815602837e-05,
+      "loss": 4.407,
+      "step": 1300
+    },
+    {
+      "epoch": 0.27688214950784784,
+      "grad_norm": 1.9229224160612461,
+      "learning_rate": 9.22695035460993e-05,
+      "loss": 4.4011,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2770949720670391,
+      "grad_norm": 2.209852019365747,
+      "learning_rate": 9.23404255319149e-05,
+      "loss": 4.4053,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2773077946262304,
+      "grad_norm": 1.9425087760284903,
+      "learning_rate": 9.24113475177305e-05,
+      "loss": 4.3751,
+      "step": 1303
+    },
+    {
+      "epoch": 0.27752061718542165,
+      "grad_norm": 2.029492906066455,
+      "learning_rate": 9.248226950354611e-05,
+      "loss": 4.4176,
+      "step": 1304
+    },
+    {
+      "epoch": 0.27773343974461295,
+      "grad_norm": 2.323261174277387,
+      "learning_rate": 9.25531914893617e-05,
+      "loss": 4.3623,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2779462623038042,
+      "grad_norm": 1.9657271564104475,
+      "learning_rate": 9.26241134751773e-05,
+      "loss": 4.3165,
+      "step": 1306
+    },
+    {
+      "epoch": 0.27815908486299545,
+      "grad_norm": 2.016521959086735,
+      "learning_rate": 9.269503546099291e-05,
+      "loss": 4.4055,
+      "step": 1307
+    },
+    {
+      "epoch": 0.27837190742218676,
+      "grad_norm": 2.1699231403388275,
+      "learning_rate": 9.276595744680852e-05,
+      "loss": 4.3509,
+      "step": 1308
+    },
+    {
+      "epoch": 0.278584729981378,
+      "grad_norm": 2.1094280991980146,
+      "learning_rate": 9.283687943262411e-05,
+      "loss": 4.4091,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2787975525405693,
+      "grad_norm": 2.073454765383787,
+      "learning_rate": 9.290780141843972e-05,
+      "loss": 4.3656,
+      "step": 1310
+    },
+    {
+      "epoch": 0.27901037509976057,
+      "grad_norm": 1.47994376113186,
+      "learning_rate": 9.297872340425532e-05,
+      "loss": 4.4019,
+      "step": 1311
+    },
+    {
+      "epoch": 0.27922319765895187,
+      "grad_norm": 1.982983428376297,
+      "learning_rate": 9.304964539007093e-05,
+      "loss": 4.4032,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2794360202181431,
+      "grad_norm": 1.856971410623541,
+      "learning_rate": 9.312056737588652e-05,
+      "loss": 4.3495,
+      "step": 1313
+    },
+    {
+      "epoch": 0.27964884277733437,
+      "grad_norm": 2.216396182759835,
+      "learning_rate": 9.319148936170213e-05,
+      "loss": 4.3731,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2798616653365257,
+      "grad_norm": 2.3053183718663126,
+      "learning_rate": 9.326241134751774e-05,
+      "loss": 4.3648,
+      "step": 1315
+    },
+    {
+      "epoch": 0.28007448789571693,
+      "grad_norm": 1.8538950012932216,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 4.2816,
+      "step": 1316
+    },
+    {
+      "epoch": 0.28028731045490823,
+      "grad_norm": 2.4153349882838917,
+      "learning_rate": 9.340425531914894e-05,
+      "loss": 4.3482,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2805001330140995,
+      "grad_norm": 1.7039059596859067,
+      "learning_rate": 9.347517730496454e-05,
+      "loss": 4.3575,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2807129555732908,
+      "grad_norm": 2.509732614389778,
+      "learning_rate": 9.354609929078015e-05,
+      "loss": 4.371,
+      "step": 1319
+    },
+    {
+      "epoch": 0.28092577813248204,
+      "grad_norm": 1.9037183040734393,
+      "learning_rate": 9.361702127659576e-05,
+      "loss": 4.3513,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2811386006916733,
+      "grad_norm": 2.2816070807709843,
+      "learning_rate": 9.368794326241135e-05,
+      "loss": 4.3352,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2813514232508646,
+      "grad_norm": 1.5946226806416328,
+      "learning_rate": 9.375886524822695e-05,
+      "loss": 4.3815,
+      "step": 1322
+    },
+    {
+      "epoch": 0.28156424581005585,
+      "grad_norm": 2.1179313054365183,
+      "learning_rate": 9.382978723404256e-05,
+      "loss": 4.3271,
+      "step": 1323
+    },
+    {
+      "epoch": 0.28177706836924715,
+      "grad_norm": 2.174861772000027,
+      "learning_rate": 9.390070921985817e-05,
+      "loss": 4.3633,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2819898909284384,
+      "grad_norm": 2.0558540998344492,
+      "learning_rate": 9.397163120567376e-05,
+      "loss": 4.3451,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2822027134876297,
+      "grad_norm": 1.6103636952587326,
+      "learning_rate": 9.404255319148937e-05,
+      "loss": 4.335,
+      "step": 1326
+    },
+    {
+      "epoch": 0.28241553604682096,
+      "grad_norm": 2.363006345402175,
+      "learning_rate": 9.411347517730497e-05,
+      "loss": 4.3881,
+      "step": 1327
+    },
+    {
+      "epoch": 0.28262835860601226,
+      "grad_norm": 1.9705614407451602,
+      "learning_rate": 9.418439716312058e-05,
+      "loss": 4.3506,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2828411811652035,
+      "grad_norm": 2.2066235424189253,
+      "learning_rate": 9.425531914893617e-05,
+      "loss": 4.3524,
+      "step": 1329
+    },
+    {
+      "epoch": 0.28305400372439476,
+      "grad_norm": 1.790562672551193,
+      "learning_rate": 9.432624113475178e-05,
+      "loss": 4.315,
+      "step": 1330
+    },
+    {
+      "epoch": 0.28326682628358607,
+      "grad_norm": 2.2919166412442173,
+      "learning_rate": 9.439716312056737e-05,
+      "loss": 4.3498,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2834796488427773,
+      "grad_norm": 1.7862364214317787,
+      "learning_rate": 9.446808510638299e-05,
+      "loss": 4.2967,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2836924714019686,
+      "grad_norm": 2.318013664150992,
+      "learning_rate": 9.453900709219858e-05,
+      "loss": 4.3694,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2839052939611599,
+      "grad_norm": 1.7692232465526676,
+      "learning_rate": 9.460992907801419e-05,
+      "loss": 4.2859,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2841181165203512,
+      "grad_norm": 2.416848221461421,
+      "learning_rate": 9.468085106382978e-05,
+      "loss": 4.3432,
+      "step": 1335
+    },
+    {
+      "epoch": 0.28433093907954243,
+      "grad_norm": 1.914503823958653,
+      "learning_rate": 9.47517730496454e-05,
+      "loss": 4.2612,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2845437616387337,
+      "grad_norm": 1.9310014490853675,
+      "learning_rate": 9.4822695035461e-05,
+      "loss": 4.3438,
+      "step": 1337
+    },
+    {
+      "epoch": 0.284756584197925,
+      "grad_norm": 1.8674309085915157,
+      "learning_rate": 9.48936170212766e-05,
+      "loss": 4.3312,
+      "step": 1338
+    },
+    {
+      "epoch": 0.28496940675711624,
+      "grad_norm": 1.9239211286583577,
+      "learning_rate": 9.49645390070922e-05,
+      "loss": 4.3212,
+      "step": 1339
+    },
+    {
+      "epoch": 0.28518222931630755,
+      "grad_norm": 1.7627508070715214,
+      "learning_rate": 9.503546099290782e-05,
+      "loss": 4.3078,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2853950518754988,
+      "grad_norm": 2.2022855528304497,
+      "learning_rate": 9.510638297872341e-05,
+      "loss": 4.2872,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2856078744346901,
+      "grad_norm": 1.9091241535016978,
+      "learning_rate": 9.517730496453901e-05,
+      "loss": 4.2952,
+      "step": 1342
+    },
+    {
+      "epoch": 0.28582069699388135,
+      "grad_norm": 2.4311048206449044,
+      "learning_rate": 9.524822695035461e-05,
+      "loss": 4.3084,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2860335195530726,
+      "grad_norm": 1.716014305552909,
+      "learning_rate": 9.531914893617023e-05,
+      "loss": 4.3158,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2862463421122639,
+      "grad_norm": 2.4631395223243318,
+      "learning_rate": 9.539007092198582e-05,
+      "loss": 4.2832,
+      "step": 1345
+    },
+    {
+      "epoch": 0.28645916467145516,
+      "grad_norm": 1.8352110387738065,
+      "learning_rate": 9.546099290780143e-05,
+      "loss": 4.3486,
+      "step": 1346
+    },
+    {
+      "epoch": 0.28667198723064646,
+      "grad_norm": 2.2756616696162917,
+      "learning_rate": 9.553191489361702e-05,
+      "loss": 4.3159,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2868848097898377,
+      "grad_norm": 2.0139118535446188,
+      "learning_rate": 9.560283687943264e-05,
+      "loss": 4.3035,
+      "step": 1348
+    },
+    {
+      "epoch": 0.287097632349029,
+      "grad_norm": 2.154836918643813,
+      "learning_rate": 9.567375886524823e-05,
+      "loss": 4.3293,
+      "step": 1349
+    },
+    {
+      "epoch": 0.28731045490822027,
+      "grad_norm": 1.8680252409133034,
+      "learning_rate": 9.574468085106384e-05,
+      "loss": 4.316,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2875232774674115,
+      "grad_norm": 2.1704724151191104,
+      "learning_rate": 9.581560283687943e-05,
+      "loss": 4.3034,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2877361000266028,
+      "grad_norm": 1.8943400815130151,
+      "learning_rate": 9.588652482269505e-05,
+      "loss": 4.2587,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2879489225857941,
+      "grad_norm": 1.9649498511964072,
+      "learning_rate": 9.595744680851064e-05,
+      "loss": 4.3211,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2881617451449854,
+      "grad_norm": 2.418876915211627,
+      "learning_rate": 9.602836879432625e-05,
+      "loss": 4.3144,
+      "step": 1354
+    },
+    {
+      "epoch": 0.28837456770417663,
+      "grad_norm": 1.8354000463439362,
+      "learning_rate": 9.609929078014184e-05,
+      "loss": 4.2866,
+      "step": 1355
+    },
+    {
+      "epoch": 0.28858739026336794,
+      "grad_norm": 2.089660621063938,
+      "learning_rate": 9.617021276595745e-05,
+      "loss": 4.3238,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2888002128225592,
+      "grad_norm": 2.2365722381997637,
+      "learning_rate": 9.624113475177306e-05,
+      "loss": 4.2891,
+      "step": 1357
+    },
+    {
+      "epoch": 0.28901303538175044,
+      "grad_norm": 2.439215273661668,
+      "learning_rate": 9.631205673758865e-05,
+      "loss": 4.2687,
+      "step": 1358
+    },
+    {
+      "epoch": 0.28922585794094174,
+      "grad_norm": 1.8984721626264625,
+      "learning_rate": 9.638297872340426e-05,
+      "loss": 4.2874,
+      "step": 1359
+    },
+    {
+      "epoch": 0.289438680500133,
+      "grad_norm": 2.167214234995722,
+      "learning_rate": 9.645390070921986e-05,
+      "loss": 4.2621,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2896515030593243,
+      "grad_norm": 1.9900525439879977,
+      "learning_rate": 9.652482269503547e-05,
+      "loss": 4.2713,
+      "step": 1361
+    },
+    {
+      "epoch": 0.28986432561851555,
+      "grad_norm": 1.8225492178518574,
+      "learning_rate": 9.659574468085106e-05,
+      "loss": 4.2314,
+      "step": 1362
+    },
+    {
+      "epoch": 0.29007714817770686,
+      "grad_norm": 2.167901742938815,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 4.3019,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2902899707368981,
+      "grad_norm": 2.174449278978094,
+      "learning_rate": 9.673758865248227e-05,
+      "loss": 4.2906,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2905027932960894,
+      "grad_norm": 2.1590844969872482,
+      "learning_rate": 9.680851063829788e-05,
+      "loss": 4.2545,
+      "step": 1365
+    },
+    {
+      "epoch": 0.29071561585528066,
+      "grad_norm": 1.8449986820528446,
+      "learning_rate": 9.687943262411347e-05,
+      "loss": 4.2588,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2909284384144719,
+      "grad_norm": 2.0784795107126337,
+      "learning_rate": 9.695035460992908e-05,
+      "loss": 4.2661,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2911412609736632,
+      "grad_norm": 1.8597742182150883,
+      "learning_rate": 9.702127659574469e-05,
+      "loss": 4.27,
+      "step": 1368
+    },
+    {
+      "epoch": 0.29135408353285447,
+      "grad_norm": 2.0988121025501334,
+      "learning_rate": 9.709219858156029e-05,
+      "loss": 4.2607,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2915669060920458,
+      "grad_norm": 2.263232933489177,
+      "learning_rate": 9.716312056737588e-05,
+      "loss": 4.304,
+      "step": 1370
+    },
+    {
+      "epoch": 0.291779728651237,
+      "grad_norm": 1.706250476924176,
+      "learning_rate": 9.723404255319149e-05,
+      "loss": 4.2601,
+      "step": 1371
+    },
+    {
+      "epoch": 0.29199255121042833,
+      "grad_norm": 2.124647487415613,
+      "learning_rate": 9.73049645390071e-05,
+      "loss": 4.2523,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2922053737696196,
+      "grad_norm": 1.8520011236820058,
+      "learning_rate": 9.73758865248227e-05,
+      "loss": 4.2363,
+      "step": 1373
+    },
+    {
+      "epoch": 0.29241819632881083,
+      "grad_norm": 2.0206122366137294,
+      "learning_rate": 9.74468085106383e-05,
+      "loss": 4.2941,
+      "step": 1374
+    },
+    {
+      "epoch": 0.29263101888800214,
+      "grad_norm": 2.0421032082164836,
+      "learning_rate": 9.75177304964539e-05,
+      "loss": 4.227,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2928438414471934,
+      "grad_norm": 1.7804524799003993,
+      "learning_rate": 9.758865248226951e-05,
+      "loss": 4.2563,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2930566640063847,
+      "grad_norm": 2.219954629406937,
+      "learning_rate": 9.765957446808512e-05,
+      "loss": 4.2692,
+      "step": 1377
+    },
+    {
+      "epoch": 0.29326948656557594,
+      "grad_norm": 2.0968998865560042,
+      "learning_rate": 9.773049645390071e-05,
+      "loss": 4.2314,
+      "step": 1378
+    },
+    {
+      "epoch": 0.29348230912476725,
+      "grad_norm": 2.148999139785424,
+      "learning_rate": 9.780141843971632e-05,
+      "loss": 4.279,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2936951316839585,
+      "grad_norm": 2.3354141775354926,
+      "learning_rate": 9.787234042553192e-05,
+      "loss": 4.2945,
+      "step": 1380
+    },
+    {
+      "epoch": 0.29390795424314975,
+      "grad_norm": 2.0053446709594316,
+      "learning_rate": 9.794326241134753e-05,
+      "loss": 4.2533,
+      "step": 1381
+    },
+    {
+      "epoch": 0.29412077680234106,
+      "grad_norm": 1.9618987429817016,
+      "learning_rate": 9.801418439716312e-05,
+      "loss": 4.2179,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2943335993615323,
+      "grad_norm": 1.7795023703109905,
+      "learning_rate": 9.808510638297873e-05,
+      "loss": 4.2568,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2945464219207236,
+      "grad_norm": 1.9645209391411815,
+      "learning_rate": 9.815602836879433e-05,
+      "loss": 4.2579,
+      "step": 1384
+    },
+    {
+      "epoch": 0.29475924447991486,
+      "grad_norm": 2.113162675607958,
+      "learning_rate": 9.822695035460994e-05,
+      "loss": 4.1868,
+      "step": 1385
+    },
+    {
+      "epoch": 0.29497206703910617,
+      "grad_norm": 2.040498358336796,
+      "learning_rate": 9.829787234042553e-05,
+      "loss": 4.2827,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2951848895982974,
+      "grad_norm": 2.1459707706213744,
+      "learning_rate": 9.836879432624114e-05,
+      "loss": 4.2675,
+      "step": 1387
+    },
+    {
+      "epoch": 0.29539771215748867,
+      "grad_norm": 1.918885850733912,
+      "learning_rate": 9.843971631205675e-05,
+      "loss": 4.2618,
+      "step": 1388
+    },
+    {
+      "epoch": 0.29561053471668,
+      "grad_norm": 2.2094070473017533,
+      "learning_rate": 9.851063829787235e-05,
+      "loss": 4.2534,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2958233572758712,
+      "grad_norm": 1.8772982094823816,
+      "learning_rate": 9.858156028368794e-05,
+      "loss": 4.2479,
+      "step": 1390
+    },
+    {
+      "epoch": 0.29603617983506253,
+      "grad_norm": 2.4171731193903647,
+      "learning_rate": 9.865248226950355e-05,
+      "loss": 4.2826,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2962490023942538,
+      "grad_norm": 1.8030197124534146,
+      "learning_rate": 9.872340425531916e-05,
+      "loss": 4.2569,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2964618249534451,
+      "grad_norm": 2.7578459794014467,
+      "learning_rate": 9.879432624113476e-05,
+      "loss": 4.2681,
+      "step": 1393
+    },
+    {
+      "epoch": 0.29667464751263634,
+      "grad_norm": 1.645052427989887,
+      "learning_rate": 9.886524822695036e-05,
+      "loss": 4.2204,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2968874700718276,
+      "grad_norm": 2.3562730126035087,
+      "learning_rate": 9.893617021276596e-05,
+      "loss": 4.2236,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2971002926310189,
+      "grad_norm": 1.84073982616707,
+      "learning_rate": 9.900709219858157e-05,
+      "loss": 4.2268,
+      "step": 1396
+    },
+    {
+      "epoch": 0.29731311519021014,
+      "grad_norm": 2.0294494029369,
+      "learning_rate": 9.907801418439718e-05,
+      "loss": 4.2859,
+      "step": 1397
+    },
+    {
+      "epoch": 0.29752593774940145,
+      "grad_norm": 2.357890475158016,
+      "learning_rate": 9.914893617021277e-05,
+      "loss": 4.2546,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2977387603085927,
+      "grad_norm": 1.8173094964045464,
+      "learning_rate": 9.921985815602838e-05,
+      "loss": 4.2706,
+      "step": 1399
+    },
+    {
+      "epoch": 0.297951582867784,
+      "grad_norm": 1.9212096748009289,
+      "learning_rate": 9.929078014184398e-05,
+      "loss": 4.2602,
+      "step": 1400
+    },
+    {
+      "epoch": 0.29816440542697525,
+      "grad_norm": 1.9929197393951634,
+      "learning_rate": 9.936170212765959e-05,
+      "loss": 4.2349,
+      "step": 1401
+    },
+    {
+      "epoch": 0.29837722798616656,
+      "grad_norm": 2.24319956196995,
+      "learning_rate": 9.943262411347518e-05,
+      "loss": 4.2088,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2985900505453578,
+      "grad_norm": 1.8616852976927558,
+      "learning_rate": 9.950354609929079e-05,
+      "loss": 4.2347,
+      "step": 1403
+    },
+    {
+      "epoch": 0.29880287310454906,
+      "grad_norm": 2.177951317983218,
+      "learning_rate": 9.95744680851064e-05,
+      "loss": 4.2422,
+      "step": 1404
+    },
+    {
+      "epoch": 0.29901569566374037,
+      "grad_norm": 1.796248402822251,
+      "learning_rate": 9.9645390070922e-05,
+      "loss": 4.2211,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2992285182229316,
+      "grad_norm": 2.2055550366985304,
+      "learning_rate": 9.971631205673759e-05,
+      "loss": 4.2189,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2994413407821229,
+      "grad_norm": 1.997559274847833,
+      "learning_rate": 9.97872340425532e-05,
+      "loss": 4.2092,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2996541633413142,
+      "grad_norm": 1.809033015796523,
+      "learning_rate": 9.98581560283688e-05,
+      "loss": 4.1961,
+      "step": 1408
+    },
+    {
+      "epoch": 0.2998669859005055,
+      "grad_norm": 2.0302838925305333,
+      "learning_rate": 9.992907801418441e-05,
+      "loss": 4.2112,
+      "step": 1409
+    },
+    {
+      "epoch": 0.30007980845969673,
+      "grad_norm": 2.057210098702343,
+      "learning_rate": 0.0001,
+      "loss": 4.2235,
+      "step": 1410
+    },
+    {
+      "epoch": 0.300292631018888,
+      "grad_norm": 2.1457528855025587,
+      "learning_rate": 9.999999846634638e-05,
+      "loss": 4.2039,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3005054535780793,
+      "grad_norm": 1.9717607009725313,
+      "learning_rate": 9.999999386538561e-05,
+      "loss": 4.251,
+      "step": 1412
+    },
+    {
+      "epoch": 0.30071827613727053,
+      "grad_norm": 1.985340324506948,
+      "learning_rate": 9.999998619711795e-05,
+      "loss": 4.2291,
+      "step": 1413
+    },
+    {
+      "epoch": 0.30093109869646184,
+      "grad_norm": 2.509465890242549,
+      "learning_rate": 9.99999754615439e-05,
+      "loss": 4.1798,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3011439212556531,
+      "grad_norm": 1.9335109898685432,
+      "learning_rate": 9.999996165866412e-05,
+      "loss": 4.2113,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3013567438148444,
+      "grad_norm": 2.1006893155320183,
+      "learning_rate": 9.999994478847943e-05,
+      "loss": 4.2734,
+      "step": 1416
+    },
+    {
+      "epoch": 0.30156956637403565,
+      "grad_norm": 1.8738318137090597,
+      "learning_rate": 9.999992485099088e-05,
+      "loss": 4.2268,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3017823889332269,
+      "grad_norm": 2.220831979330392,
+      "learning_rate": 9.999990184619969e-05,
+      "loss": 4.2393,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3019952114924182,
+      "grad_norm": 1.8697274597157458,
+      "learning_rate": 9.999987577410728e-05,
+      "loss": 4.1969,
+      "step": 1419
+    },
+    {
+      "epoch": 0.30220803405160945,
+      "grad_norm": 2.0232056215698826,
+      "learning_rate": 9.999984663471525e-05,
+      "loss": 4.1762,
+      "step": 1420
+    },
+    {
+      "epoch": 0.30242085661080076,
+      "grad_norm": 1.9742669127351153,
+      "learning_rate": 9.999981442802537e-05,
+      "loss": 4.2402,
+      "step": 1421
+    },
+    {
+      "epoch": 0.302633679169992,
+      "grad_norm": 1.6504141822586853,
+      "learning_rate": 9.999977915403962e-05,
+      "loss": 4.1543,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3028465017291833,
+      "grad_norm": 2.571338736271418,
+      "learning_rate": 9.999974081276018e-05,
+      "loss": 4.2221,
+      "step": 1423
+    },
+    {
+      "epoch": 0.30305932428837457,
+      "grad_norm": 1.4691622128277144,
+      "learning_rate": 9.999969940418939e-05,
+      "loss": 4.209,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3032721468475658,
+      "grad_norm": 2.46972944674285,
+      "learning_rate": 9.99996549283298e-05,
+      "loss": 4.2131,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3034849694067571,
+      "grad_norm": 1.706190121275559,
+      "learning_rate": 9.999960738518413e-05,
+      "loss": 4.2164,
+      "step": 1426
+    },
+    {
+      "epoch": 0.30369779196594837,
+      "grad_norm": 2.210893206340707,
+      "learning_rate": 9.999955677475529e-05,
+      "loss": 4.1912,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3039106145251397,
+      "grad_norm": 2.0227605100429167,
+      "learning_rate": 9.999950309704639e-05,
+      "loss": 4.2092,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3041234370843309,
+      "grad_norm": 1.9890935220013743,
+      "learning_rate": 9.999944635206073e-05,
+      "loss": 4.1698,
+      "step": 1429
+    },
+    {
+      "epoch": 0.30433625964352223,
+      "grad_norm": 1.943359890376191,
+      "learning_rate": 9.99993865398018e-05,
+      "loss": 4.189,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3045490822027135,
+      "grad_norm": 1.992189256042452,
+      "learning_rate": 9.999932366027325e-05,
+      "loss": 4.2139,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 2.1181203920941156,
+      "learning_rate": 9.999925771347894e-05,
+      "loss": 4.197,
+      "step": 1432
+    },
+    {
+      "epoch": 0.30497472732109604,
+      "grad_norm": 2.1473245805733603,
+      "learning_rate": 9.999918869942291e-05,
+      "loss": 4.2088,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3051875498802873,
+      "grad_norm": 2.360729626878775,
+      "learning_rate": 9.99991166181094e-05,
+      "loss": 4.2255,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3054003724394786,
+      "grad_norm": 1.6860919742625724,
+      "learning_rate": 9.999904146954284e-05,
+      "loss": 4.1414,
+      "step": 1435
+    },
+    {
+      "epoch": 0.30561319499866985,
+      "grad_norm": 2.1932565820268763,
+      "learning_rate": 9.999896325372784e-05,
+      "loss": 4.1947,
+      "step": 1436
+    },
+    {
+      "epoch": 0.30582601755786115,
+      "grad_norm": 1.8317275804021,
+      "learning_rate": 9.99988819706692e-05,
+      "loss": 4.1779,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3060388401170524,
+      "grad_norm": 2.265291864346559,
+      "learning_rate": 9.999879762037188e-05,
+      "loss": 4.1917,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3062516626762437,
+      "grad_norm": 1.8941383658626065,
+      "learning_rate": 9.999871020284109e-05,
+      "loss": 4.1786,
+      "step": 1439
+    },
+    {
+      "epoch": 0.30646448523543496,
+      "grad_norm": 2.126259336182579,
+      "learning_rate": 9.999861971808216e-05,
+      "loss": 4.1566,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3066773077946262,
+      "grad_norm": 1.954886568685915,
+      "learning_rate": 9.999852616610068e-05,
+      "loss": 4.1931,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3068901303538175,
+      "grad_norm": 2.043833111543689,
+      "learning_rate": 9.999842954690235e-05,
+      "loss": 4.2133,
+      "step": 1442
+    },
+    {
+      "epoch": 0.30710295291300876,
+      "grad_norm": 2.365146313556929,
+      "learning_rate": 9.999832986049311e-05,
+      "loss": 4.2134,
+      "step": 1443
+    },
+    {
+      "epoch": 0.30731577547220007,
+      "grad_norm": 2.0133185227183987,
+      "learning_rate": 9.999822710687909e-05,
+      "loss": 4.1608,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3075285980313913,
+      "grad_norm": 1.8360943304949244,
+      "learning_rate": 9.999812128606659e-05,
+      "loss": 4.1938,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3077414205905826,
+      "grad_norm": 2.1195509917426283,
+      "learning_rate": 9.999801239806208e-05,
+      "loss": 4.1772,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3079542431497739,
+      "grad_norm": 1.7886154961955596,
+      "learning_rate": 9.999790044287225e-05,
+      "loss": 4.1655,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3081670657089651,
+      "grad_norm": 2.3014249879925113,
+      "learning_rate": 9.999778542050397e-05,
+      "loss": 4.2049,
+      "step": 1448
+    },
+    {
+      "epoch": 0.30837988826815643,
+      "grad_norm": 1.9457746832592913,
+      "learning_rate": 9.999766733096432e-05,
+      "loss": 4.1732,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3085927108273477,
+      "grad_norm": 2.1047993443411706,
+      "learning_rate": 9.999754617426052e-05,
+      "loss": 4.1639,
+      "step": 1450
+    },
+    {
+      "epoch": 0.308805533386539,
+      "grad_norm": 2.048384237611689,
+      "learning_rate": 9.999742195039998e-05,
+      "loss": 4.1712,
+      "step": 1451
+    },
+    {
+      "epoch": 0.30901835594573024,
+      "grad_norm": 1.9224551676110417,
+      "learning_rate": 9.999729465939036e-05,
+      "loss": 4.1376,
+      "step": 1452
+    },
+    {
+      "epoch": 0.30923117850492154,
+      "grad_norm": 1.9190531849245456,
+      "learning_rate": 9.999716430123947e-05,
+      "loss": 4.1981,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3094440010641128,
+      "grad_norm": 2.0262802786648155,
+      "learning_rate": 9.999703087595526e-05,
+      "loss": 4.0792,
+      "step": 1454
+    },
+    {
+      "epoch": 0.30965682362330404,
+      "grad_norm": 2.2005176267807927,
+      "learning_rate": 9.999689438354597e-05,
+      "loss": 4.1867,
+      "step": 1455
+    },
+    {
+      "epoch": 0.30986964618249535,
+      "grad_norm": 1.826690489725022,
+      "learning_rate": 9.999675482401995e-05,
+      "loss": 4.157,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3100824687416866,
+      "grad_norm": 2.2413100348649904,
+      "learning_rate": 9.999661219738576e-05,
+      "loss": 4.1524,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3102952913008779,
+      "grad_norm": 1.8219429953711164,
+      "learning_rate": 9.999646650365214e-05,
+      "loss": 4.1158,
+      "step": 1458
+    },
+    {
+      "epoch": 0.31050811386006916,
+      "grad_norm": 2.0851007987655388,
+      "learning_rate": 9.999631774282804e-05,
+      "loss": 4.1134,
+      "step": 1459
+    },
+    {
+      "epoch": 0.31072093641926046,
+      "grad_norm": 1.825895203639323,
+      "learning_rate": 9.999616591492258e-05,
+      "loss": 4.1384,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3109337589784517,
+      "grad_norm": 1.9173117757438374,
+      "learning_rate": 9.99960110199451e-05,
+      "loss": 4.1272,
+      "step": 1461
+    },
+    {
+      "epoch": 0.31114658153764296,
+      "grad_norm": 1.8540935727909744,
+      "learning_rate": 9.999585305790507e-05,
+      "loss": 4.163,
+      "step": 1462
+    },
+    {
+      "epoch": 0.31135940409683427,
+      "grad_norm": 1.8422027597559865,
+      "learning_rate": 9.999569202881219e-05,
+      "loss": 4.1333,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3115722266560255,
+      "grad_norm": 2.138456154674984,
+      "learning_rate": 9.999552793267634e-05,
+      "loss": 4.1404,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3117850492152168,
+      "grad_norm": 1.8652472162465592,
+      "learning_rate": 9.999536076950759e-05,
+      "loss": 4.1623,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3119978717744081,
+      "grad_norm": 2.002548033401184,
+      "learning_rate": 9.999519053931618e-05,
+      "loss": 4.077,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3122106943335994,
+      "grad_norm": 1.9196229634435173,
+      "learning_rate": 9.999501724211257e-05,
+      "loss": 4.1188,
+      "step": 1467
+    },
+    {
+      "epoch": 0.31242351689279063,
+      "grad_norm": 2.086789327407174,
+      "learning_rate": 9.99948408779074e-05,
+      "loss": 4.0833,
+      "step": 1468
+    },
+    {
+      "epoch": 0.31263633945198194,
+      "grad_norm": 1.8863368811857597,
+      "learning_rate": 9.999466144671145e-05,
+      "loss": 4.1949,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3128491620111732,
+      "grad_norm": 2.224062573988124,
+      "learning_rate": 9.999447894853577e-05,
+      "loss": 4.1148,
+      "step": 1470
+    },
+    {
+      "epoch": 0.31306198457036444,
+      "grad_norm": 1.730943621348111,
+      "learning_rate": 9.999429338339155e-05,
+      "loss": 4.09,
+      "step": 1471
+    },
+    {
+      "epoch": 0.31327480712955574,
+      "grad_norm": 2.452150813906229,
+      "learning_rate": 9.999410475129014e-05,
+      "loss": 4.119,
+      "step": 1472
+    },
+    {
+      "epoch": 0.313487629688747,
+      "grad_norm": 1.867270982885157,
+      "learning_rate": 9.999391305224312e-05,
+      "loss": 4.1668,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3137004522479383,
+      "grad_norm": 2.229152275628279,
+      "learning_rate": 9.99937182862623e-05,
+      "loss": 4.1407,
+      "step": 1474
+    },
+    {
+      "epoch": 0.31391327480712955,
+      "grad_norm": 2.105771475296658,
+      "learning_rate": 9.999352045335957e-05,
+      "loss": 4.1619,
+      "step": 1475
+    },
+    {
+      "epoch": 0.31412609736632086,
+      "grad_norm": 2.10059648087067,
+      "learning_rate": 9.999331955354708e-05,
+      "loss": 4.1118,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3143389199255121,
+      "grad_norm": 2.270607775986675,
+      "learning_rate": 9.999311558683717e-05,
+      "loss": 4.1277,
+      "step": 1477
+    },
+    {
+      "epoch": 0.31455174248470336,
+      "grad_norm": 1.9285098020722493,
+      "learning_rate": 9.999290855324235e-05,
+      "loss": 4.1178,
+      "step": 1478
+    },
+    {
+      "epoch": 0.31476456504389466,
+      "grad_norm": 2.043678326172579,
+      "learning_rate": 9.99926984527753e-05,
+      "loss": 4.1446,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3149773876030859,
+      "grad_norm": 1.9987127716172879,
+      "learning_rate": 9.999248528544894e-05,
+      "loss": 4.1305,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3151902101622772,
+      "grad_norm": 2.014321175400119,
+      "learning_rate": 9.999226905127634e-05,
+      "loss": 4.1065,
+      "step": 1481
+    },
+    {
+      "epoch": 0.31540303272146847,
+      "grad_norm": 1.976907448924241,
+      "learning_rate": 9.999204975027073e-05,
+      "loss": 4.0813,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3156158552806598,
+      "grad_norm": 2.1282442586889907,
+      "learning_rate": 9.999182738244561e-05,
+      "loss": 4.1105,
+      "step": 1483
+    },
+    {
+      "epoch": 0.315828677839851,
+      "grad_norm": 2.080234562058769,
+      "learning_rate": 9.999160194781459e-05,
+      "loss": 4.1351,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3160415003990423,
+      "grad_norm": 1.7217845271174708,
+      "learning_rate": 9.999137344639151e-05,
+      "loss": 4.0259,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3162543229582336,
+      "grad_norm": 1.8797133442217255,
+      "learning_rate": 9.99911418781904e-05,
+      "loss": 4.1122,
+      "step": 1486
+    },
+    {
+      "epoch": 0.31646714551742483,
+      "grad_norm": 2.0162766372711363,
+      "learning_rate": 9.999090724322543e-05,
+      "loss": 4.1497,
+      "step": 1487
+    },
+    {
+      "epoch": 0.31667996807661614,
+      "grad_norm": 1.9079085654606276,
+      "learning_rate": 9.999066954151104e-05,
+      "loss": 4.0624,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3168927906358074,
+      "grad_norm": 1.9528270518170816,
+      "learning_rate": 9.99904287730618e-05,
+      "loss": 4.0733,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3171056131949987,
+      "grad_norm": 1.8773262303001417,
+      "learning_rate": 9.999018493789246e-05,
+      "loss": 4.0721,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31731843575418994,
+      "grad_norm": 1.9719159346972406,
+      "learning_rate": 9.998993803601797e-05,
+      "loss": 4.104,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3175312583133812,
+      "grad_norm": 1.5784956047717422,
+      "learning_rate": 9.99896880674535e-05,
+      "loss": 4.0775,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3177440808725725,
+      "grad_norm": 2.5293251495933657,
+      "learning_rate": 9.998943503221439e-05,
+      "loss": 4.0955,
+      "step": 1493
+    },
+    {
+      "epoch": 0.31795690343176375,
+      "grad_norm": 1.6538696175942666,
+      "learning_rate": 9.998917893031616e-05,
+      "loss": 4.0901,
+      "step": 1494
+    },
+    {
+      "epoch": 0.31816972599095505,
+      "grad_norm": 2.1442304770148217,
+      "learning_rate": 9.998891976177449e-05,
+      "loss": 4.1383,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3183825485501463,
+      "grad_norm": 1.9333531974391915,
+      "learning_rate": 9.998865752660533e-05,
+      "loss": 4.0374,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3185953711093376,
+      "grad_norm": 2.199797421608051,
+      "learning_rate": 9.998839222482473e-05,
+      "loss": 4.1559,
+      "step": 1497
+    },
+    {
+      "epoch": 0.31880819366852886,
+      "grad_norm": 1.8204745277185557,
+      "learning_rate": 9.998812385644896e-05,
+      "loss": 4.1001,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3190210162277201,
+      "grad_norm": 2.052984502250652,
+      "learning_rate": 9.99878524214945e-05,
+      "loss": 4.0904,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3192338387869114,
+      "grad_norm": 1.8372287153271167,
+      "learning_rate": 9.998757791997802e-05,
+      "loss": 4.1138,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31944666134610267,
+      "grad_norm": 2.2747847146122413,
+      "learning_rate": 9.998730035191633e-05,
+      "loss": 4.1144,
+      "step": 1501
+    },
+    {
+      "epoch": 0.319659483905294,
+      "grad_norm": 1.6488286575089979,
+      "learning_rate": 9.998701971732645e-05,
+      "loss": 4.0765,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3198723064644852,
+      "grad_norm": 2.335273697742766,
+      "learning_rate": 9.998673601622563e-05,
+      "loss": 4.0613,
+      "step": 1503
+    },
+    {
+      "epoch": 0.32008512902367653,
+      "grad_norm": 1.6493072145424328,
+      "learning_rate": 9.998644924863126e-05,
+      "loss": 4.1038,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3202979515828678,
+      "grad_norm": 2.2548493249270103,
+      "learning_rate": 9.998615941456091e-05,
+      "loss": 4.0651,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3205107741420591,
+      "grad_norm": 1.8307324786742603,
+      "learning_rate": 9.998586651403239e-05,
+      "loss": 4.0853,
+      "step": 1506
+    },
+    {
+      "epoch": 0.32072359670125034,
+      "grad_norm": 2.1236703723062043,
+      "learning_rate": 9.998557054706365e-05,
+      "loss": 4.0219,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3209364192604416,
+      "grad_norm": 1.6668727405665995,
+      "learning_rate": 9.998527151367287e-05,
+      "loss": 4.0589,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3211492418196329,
+      "grad_norm": 2.279675059874618,
+      "learning_rate": 9.998496941387836e-05,
+      "loss": 4.1102,
+      "step": 1509
+    },
+    {
+      "epoch": 0.32136206437882414,
+      "grad_norm": 1.8322624986430958,
+      "learning_rate": 9.998466424769868e-05,
+      "loss": 4.0602,
+      "step": 1510
+    },
+    {
+      "epoch": 0.32157488693801545,
+      "grad_norm": 1.782482650039212,
+      "learning_rate": 9.998435601515255e-05,
+      "loss": 4.077,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3217877094972067,
+      "grad_norm": 2.2927886076287325,
+      "learning_rate": 9.998404471625886e-05,
+      "loss": 4.0744,
+      "step": 1512
+    },
+    {
+      "epoch": 0.322000532056398,
+      "grad_norm": 1.6692371073252472,
+      "learning_rate": 9.998373035103672e-05,
+      "loss": 4.0805,
+      "step": 1513
+    },
+    {
+      "epoch": 0.32221335461558925,
+      "grad_norm": 2.2499644578629656,
+      "learning_rate": 9.998341291950541e-05,
+      "loss": 4.05,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3224261771747805,
+      "grad_norm": 1.8248337505716161,
+      "learning_rate": 9.998309242168442e-05,
+      "loss": 4.0254,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3226389997339718,
+      "grad_norm": 2.322924932201736,
+      "learning_rate": 9.998276885759339e-05,
+      "loss": 4.0567,
+      "step": 1516
+    },
+    {
+      "epoch": 0.32285182229316306,
+      "grad_norm": 1.6461005218342666,
+      "learning_rate": 9.998244222725217e-05,
+      "loss": 4.0882,
+      "step": 1517
+    },
+    {
+      "epoch": 0.32306464485235437,
+      "grad_norm": 2.4379815476136124,
+      "learning_rate": 9.99821125306808e-05,
+      "loss": 4.0722,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3232774674115456,
+      "grad_norm": 1.6292289795442794,
+      "learning_rate": 9.998177976789952e-05,
+      "loss": 4.043,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3234902899707369,
+      "grad_norm": 2.1171011071672616,
+      "learning_rate": 9.998144393892873e-05,
+      "loss": 4.0124,
+      "step": 1520
+    },
+    {
+      "epoch": 0.32370311252992817,
+      "grad_norm": 2.111550795793769,
+      "learning_rate": 9.998110504378903e-05,
+      "loss": 4.0337,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3239159350891194,
+      "grad_norm": 1.7979938034831229,
+      "learning_rate": 9.998076308250121e-05,
+      "loss": 4.0443,
+      "step": 1522
+    },
+    {
+      "epoch": 0.32412875764831073,
+      "grad_norm": 2.132568433715359,
+      "learning_rate": 9.998041805508626e-05,
+      "loss": 4.0549,
+      "step": 1523
+    },
+    {
+      "epoch": 0.324341580207502,
+      "grad_norm": 1.6079985078350607,
+      "learning_rate": 9.998006996156536e-05,
+      "loss": 4.0716,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3245544027666933,
+      "grad_norm": 2.1376496628924673,
+      "learning_rate": 9.997971880195982e-05,
+      "loss": 4.0434,
+      "step": 1525
+    },
+    {
+      "epoch": 0.32476722532588453,
+      "grad_norm": 1.6548199239112114,
+      "learning_rate": 9.997936457629121e-05,
+      "loss": 4.0303,
+      "step": 1526
+    },
+    {
+      "epoch": 0.32498004788507584,
+      "grad_norm": 2.039533101564343,
+      "learning_rate": 9.997900728458126e-05,
+      "loss": 4.0377,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3251928704442671,
+      "grad_norm": 1.639649581755507,
+      "learning_rate": 9.997864692685189e-05,
+      "loss": 4.0811,
+      "step": 1528
+    },
+    {
+      "epoch": 0.32540569300345834,
+      "grad_norm": 1.9860876293019827,
+      "learning_rate": 9.997828350312519e-05,
+      "loss": 4.0552,
+      "step": 1529
+    },
+    {
+      "epoch": 0.32561851556264965,
+      "grad_norm": 1.4452093273128224,
+      "learning_rate": 9.997791701342347e-05,
+      "loss": 4.0273,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3258313381218409,
+      "grad_norm": 2.0293819583712733,
+      "learning_rate": 9.99775474577692e-05,
+      "loss": 4.0182,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3260441606810322,
+      "grad_norm": 2.0192633082321287,
+      "learning_rate": 9.997717483618508e-05,
+      "loss": 4.0107,
+      "step": 1532
+    },
+    {
+      "epoch": 0.32625698324022345,
+      "grad_norm": 1.5196573950170533,
+      "learning_rate": 9.997679914869392e-05,
+      "loss": 4.0202,
+      "step": 1533
+    },
+    {
+      "epoch": 0.32646980579941476,
+      "grad_norm": 2.4291762572695417,
+      "learning_rate": 9.997642039531881e-05,
+      "loss": 4.0434,
+      "step": 1534
+    },
+    {
+      "epoch": 0.326682628358606,
+      "grad_norm": 1.5831152215169515,
+      "learning_rate": 9.997603857608298e-05,
+      "loss": 4.031,
+      "step": 1535
+    },
+    {
+      "epoch": 0.32689545091779726,
+      "grad_norm": 2.368143953950665,
+      "learning_rate": 9.997565369100983e-05,
+      "loss": 4.03,
+      "step": 1536
+    },
+    {
+      "epoch": 0.32710827347698856,
+      "grad_norm": 1.7191378647574211,
+      "learning_rate": 9.997526574012299e-05,
+      "loss": 4.0313,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3273210960361798,
+      "grad_norm": 2.5686819941899,
+      "learning_rate": 9.997487472344624e-05,
+      "loss": 4.0571,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3275339185953711,
+      "grad_norm": 1.6453158239164551,
+      "learning_rate": 9.997448064100362e-05,
+      "loss": 4.0399,
+      "step": 1539
+    },
+    {
+      "epoch": 0.32774674115456237,
+      "grad_norm": 2.4211188375099044,
+      "learning_rate": 9.997408349281923e-05,
+      "loss": 4.0486,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3279595637137537,
+      "grad_norm": 1.706960389009179,
+      "learning_rate": 9.997368327891747e-05,
+      "loss": 4.0153,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3281723862729449,
+      "grad_norm": 1.8226722048326207,
+      "learning_rate": 9.997327999932291e-05,
+      "loss": 4.0142,
+      "step": 1542
+    },
+    {
+      "epoch": 0.32838520883213623,
+      "grad_norm": 1.9645533893211427,
+      "learning_rate": 9.997287365406025e-05,
+      "loss": 4.0675,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3285980313913275,
+      "grad_norm": 1.9494303477401864,
+      "learning_rate": 9.997246424315447e-05,
+      "loss": 4.0108,
+      "step": 1544
+    },
+    {
+      "epoch": 0.32881085395051873,
+      "grad_norm": 1.8202189925781214,
+      "learning_rate": 9.997205176663063e-05,
+      "loss": 4.0222,
+      "step": 1545
+    },
+    {
+      "epoch": 0.32902367650971004,
+      "grad_norm": 2.2128390010796704,
+      "learning_rate": 9.997163622451407e-05,
+      "loss": 4.0053,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3292364990689013,
+      "grad_norm": 1.8307190264948943,
+      "learning_rate": 9.997121761683026e-05,
+      "loss": 4.0561,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3294493216280926,
+      "grad_norm": 2.320819854594713,
+      "learning_rate": 9.99707959436049e-05,
+      "loss": 4.03,
+      "step": 1548
+    },
+    {
+      "epoch": 0.32966214418728385,
+      "grad_norm": 1.9815394143025524,
+      "learning_rate": 9.997037120486384e-05,
+      "loss": 4.0056,
+      "step": 1549
+    },
+    {
+      "epoch": 0.32987496674647515,
+      "grad_norm": 1.7348624253021843,
+      "learning_rate": 9.996994340063314e-05,
+      "loss": 4.0584,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3300877893056664,
+      "grad_norm": 2.2645085109145318,
+      "learning_rate": 9.996951253093907e-05,
+      "loss": 4.0595,
+      "step": 1551
+    },
+    {
+      "epoch": 0.33030061186485765,
+      "grad_norm": 1.8286506480917346,
+      "learning_rate": 9.996907859580803e-05,
+      "loss": 3.9924,
+      "step": 1552
+    },
+    {
+      "epoch": 0.33051343442404896,
+      "grad_norm": 1.9618177262827263,
+      "learning_rate": 9.996864159526666e-05,
+      "loss": 3.9936,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3307262569832402,
+      "grad_norm": 1.6378844612410797,
+      "learning_rate": 9.996820152934176e-05,
+      "loss": 4.039,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3309390795424315,
+      "grad_norm": 2.1901767370680116,
+      "learning_rate": 9.996775839806031e-05,
+      "loss": 3.9801,
+      "step": 1555
+    },
+    {
+      "epoch": 0.33115190210162276,
+      "grad_norm": 1.5956262494961204,
+      "learning_rate": 9.996731220144952e-05,
+      "loss": 3.9979,
+      "step": 1556
+    },
+    {
+      "epoch": 0.33136472466081407,
+      "grad_norm": 2.1407918299832205,
+      "learning_rate": 9.996686293953675e-05,
+      "loss": 4.0097,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3315775472200053,
+      "grad_norm": 1.750904306071135,
+      "learning_rate": 9.996641061234957e-05,
+      "loss": 4.0264,
+      "step": 1558
+    },
+    {
+      "epoch": 0.33179036977919657,
+      "grad_norm": 2.1602063480300573,
+      "learning_rate": 9.996595521991571e-05,
+      "loss": 3.9746,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3320031923383879,
+      "grad_norm": 1.8915461606346078,
+      "learning_rate": 9.996549676226312e-05,
+      "loss": 4.0038,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3322160148975791,
+      "grad_norm": 1.6959687862864468,
+      "learning_rate": 9.996503523941994e-05,
+      "loss": 4.0035,
+      "step": 1561
+    },
+    {
+      "epoch": 0.33242883745677043,
+      "grad_norm": 2.1954390038337994,
+      "learning_rate": 9.996457065141445e-05,
+      "loss": 3.965,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3326416600159617,
+      "grad_norm": 1.7705369827181163,
+      "learning_rate": 9.996410299827517e-05,
+      "loss": 3.9846,
+      "step": 1563
+    },
+    {
+      "epoch": 0.332854482575153,
+      "grad_norm": 1.9545508445847675,
+      "learning_rate": 9.996363228003079e-05,
+      "loss": 3.9718,
+      "step": 1564
+    },
+    {
+      "epoch": 0.33306730513434424,
+      "grad_norm": 1.72441898725454,
+      "learning_rate": 9.996315849671016e-05,
+      "loss": 3.9948,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3332801276935355,
+      "grad_norm": 2.0610656883398013,
+      "learning_rate": 9.996268164834239e-05,
+      "loss": 3.9807,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3334929502527268,
+      "grad_norm": 1.6686054255854301,
+      "learning_rate": 9.996220173495669e-05,
+      "loss": 3.9544,
+      "step": 1567
+    },
+    {
+      "epoch": 0.33370577281191804,
+      "grad_norm": 2.029885709644651,
+      "learning_rate": 9.996171875658255e-05,
+      "loss": 3.9722,
+      "step": 1568
+    },
+    {
+      "epoch": 0.33391859537110935,
+      "grad_norm": 1.8080401928343206,
+      "learning_rate": 9.996123271324955e-05,
+      "loss": 4.0038,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3341314179303006,
+      "grad_norm": 1.9025959796565706,
+      "learning_rate": 9.996074360498753e-05,
+      "loss": 3.9968,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3343442404894919,
+      "grad_norm": 1.631047338721543,
+      "learning_rate": 9.996025143182648e-05,
+      "loss": 3.9269,
+      "step": 1571
+    },
+    {
+      "epoch": 0.33455706304868316,
+      "grad_norm": 2.3128849796919044,
+      "learning_rate": 9.995975619379662e-05,
+      "loss": 3.9865,
+      "step": 1572
+    },
+    {
+      "epoch": 0.33476988560787446,
+      "grad_norm": 1.5540245792982592,
+      "learning_rate": 9.995925789092831e-05,
+      "loss": 3.9509,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3349827081670657,
+      "grad_norm": 2.4444567757900324,
+      "learning_rate": 9.995875652325212e-05,
+      "loss": 4.02,
+      "step": 1574
+    },
+    {
+      "epoch": 0.33519553072625696,
+      "grad_norm": 1.8327389594916805,
+      "learning_rate": 9.99582520907988e-05,
+      "loss": 3.9644,
+      "step": 1575
+    },
+    {
+      "epoch": 0.33540835328544827,
+      "grad_norm": 2.0778054262080152,
+      "learning_rate": 9.995774459359932e-05,
+      "loss": 4.0156,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3356211758446395,
+      "grad_norm": 1.7723115670804643,
+      "learning_rate": 9.99572340316848e-05,
+      "loss": 3.9597,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3358339984038308,
+      "grad_norm": 1.8803075618550162,
+      "learning_rate": 9.995672040508655e-05,
+      "loss": 3.9739,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3360468209630221,
+      "grad_norm": 1.9451948246193695,
+      "learning_rate": 9.99562037138361e-05,
+      "loss": 4.0006,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3362596435222134,
+      "grad_norm": 1.8047774189043988,
+      "learning_rate": 9.995568395796513e-05,
+      "loss": 3.9464,
+      "step": 1580
+    },
+    {
+      "epoch": 0.33647246608140463,
+      "grad_norm": 1.8297045741092504,
+      "learning_rate": 9.995516113750553e-05,
+      "loss": 4.0152,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3366852886405959,
+      "grad_norm": 1.9766315522751305,
+      "learning_rate": 9.995463525248938e-05,
+      "loss": 3.9502,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3368981111997872,
+      "grad_norm": 1.6506207322021313,
+      "learning_rate": 9.995410630294892e-05,
+      "loss": 3.9603,
+      "step": 1583
+    },
+    {
+      "epoch": 0.33711093375897844,
+      "grad_norm": 2.043159574580509,
+      "learning_rate": 9.995357428891663e-05,
+      "loss": 4.0393,
+      "step": 1584
+    },
+    {
+      "epoch": 0.33732375631816974,
+      "grad_norm": 1.7158391130948287,
+      "learning_rate": 9.995303921042513e-05,
+      "loss": 3.9531,
+      "step": 1585
+    },
+    {
+      "epoch": 0.337536578877361,
+      "grad_norm": 1.7292297930127836,
+      "learning_rate": 9.995250106750724e-05,
+      "loss": 3.9568,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3377494014365523,
+      "grad_norm": 1.7528588062383383,
+      "learning_rate": 9.995195986019597e-05,
+      "loss": 3.9287,
+      "step": 1587
+    },
+    {
+      "epoch": 0.33796222399574355,
+      "grad_norm": 1.853086521682465,
+      "learning_rate": 9.995141558852455e-05,
+      "loss": 3.9854,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3381750465549348,
+      "grad_norm": 1.7689834044870927,
+      "learning_rate": 9.995086825252633e-05,
+      "loss": 3.9638,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3383878691141261,
+      "grad_norm": 2.0037045208774225,
+      "learning_rate": 9.995031785223492e-05,
+      "loss": 3.9392,
+      "step": 1590
+    },
+    {
+      "epoch": 0.33860069167331736,
+      "grad_norm": 1.7400630185426285,
+      "learning_rate": 9.994976438768406e-05,
+      "loss": 3.9797,
+      "step": 1591
+    },
+    {
+      "epoch": 0.33881351423250866,
+      "grad_norm": 2.2518005867626063,
+      "learning_rate": 9.994920785890771e-05,
+      "loss": 3.941,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3390263367916999,
+      "grad_norm": 1.7293128119993444,
+      "learning_rate": 9.994864826594002e-05,
+      "loss": 3.9373,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3392391593508912,
+      "grad_norm": 2.1843116343487523,
+      "learning_rate": 9.994808560881532e-05,
+      "loss": 3.9662,
+      "step": 1594
+    },
+    {
+      "epoch": 0.33945198191008247,
+      "grad_norm": 1.65345043320922,
+      "learning_rate": 9.994751988756812e-05,
+      "loss": 3.9235,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3396648044692737,
+      "grad_norm": 1.9173993820196897,
+      "learning_rate": 9.994695110223311e-05,
+      "loss": 3.8875,
+      "step": 1596
+    },
+    {
+      "epoch": 0.339877627028465,
+      "grad_norm": 1.6571179472119826,
+      "learning_rate": 9.994637925284521e-05,
+      "loss": 3.9474,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3400904495876563,
+      "grad_norm": 1.9182037141629453,
+      "learning_rate": 9.99458043394395e-05,
+      "loss": 3.9494,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3403032721468476,
+      "grad_norm": 1.6571262901338155,
+      "learning_rate": 9.994522636205122e-05,
+      "loss": 3.9048,
+      "step": 1599
+    },
+    {
+      "epoch": 0.34051609470603883,
+      "grad_norm": 1.864264264814133,
+      "learning_rate": 9.994464532071584e-05,
+      "loss": 3.9836,
+      "step": 1600
+    },
+    {
+      "epoch": 0.34072891726523014,
+      "grad_norm": 1.92275179644132,
+      "learning_rate": 9.994406121546902e-05,
+      "loss": 3.9213,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3409417398244214,
+      "grad_norm": 1.7436767748535984,
+      "learning_rate": 9.994347404634657e-05,
+      "loss": 3.9507,
+      "step": 1602
+    },
+    {
+      "epoch": 0.34115456238361264,
+      "grad_norm": 1.9960817786678253,
+      "learning_rate": 9.994288381338452e-05,
+      "loss": 3.9751,
+      "step": 1603
+    },
+    {
+      "epoch": 0.34136738494280394,
+      "grad_norm": 1.875605777235697,
+      "learning_rate": 9.99422905166191e-05,
+      "loss": 3.97,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3415802075019952,
+      "grad_norm": 1.9707407213256385,
+      "learning_rate": 9.994169415608668e-05,
+      "loss": 3.9282,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3417930300611865,
+      "grad_norm": 1.889649250180927,
+      "learning_rate": 9.994109473182385e-05,
+      "loss": 3.9551,
+      "step": 1606
+    },
+    {
+      "epoch": 0.34200585262037775,
+      "grad_norm": 1.668911276839344,
+      "learning_rate": 9.994049224386737e-05,
+      "loss": 3.9625,
+      "step": 1607
+    },
+    {
+      "epoch": 0.34221867517956905,
+      "grad_norm": 2.0976103150899568,
+      "learning_rate": 9.993988669225423e-05,
+      "loss": 3.9232,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3424314977387603,
+      "grad_norm": 1.4669735967608155,
+      "learning_rate": 9.993927807702155e-05,
+      "loss": 3.9244,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3426443202979516,
+      "grad_norm": 2.2727445519077616,
+      "learning_rate": 9.99386663982067e-05,
+      "loss": 3.9355,
+      "step": 1610
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 1.3299459565282938,
+      "learning_rate": 9.993805165584715e-05,
+      "loss": 3.8879,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3430699654163341,
+      "grad_norm": 2.314944598527498,
+      "learning_rate": 9.993743384998067e-05,
+      "loss": 3.9644,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3432827879755254,
+      "grad_norm": 1.8339970524438334,
+      "learning_rate": 9.993681298064512e-05,
+      "loss": 3.9689,
+      "step": 1613
+    },
+    {
+      "epoch": 0.34349561053471667,
+      "grad_norm": 2.0907229862424783,
+      "learning_rate": 9.993618904787862e-05,
+      "loss": 3.9091,
+      "step": 1614
+    },
+    {
+      "epoch": 0.343708433093908,
+      "grad_norm": 1.681861118437772,
+      "learning_rate": 9.993556205171941e-05,
+      "loss": 3.9545,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3439212556530992,
+      "grad_norm": 1.7521423313772648,
+      "learning_rate": 9.993493199220598e-05,
+      "loss": 3.9152,
+      "step": 1616
+    },
+    {
+      "epoch": 0.34413407821229053,
+      "grad_norm": 1.895711476680671,
+      "learning_rate": 9.993429886937697e-05,
+      "loss": 3.8937,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3443469007714818,
+      "grad_norm": 1.815113109027338,
+      "learning_rate": 9.993366268327122e-05,
+      "loss": 3.9423,
+      "step": 1618
+    },
+    {
+      "epoch": 0.34455972333067303,
+      "grad_norm": 1.9351608683031696,
+      "learning_rate": 9.993302343392779e-05,
+      "loss": 3.9425,
+      "step": 1619
+    },
+    {
+      "epoch": 0.34477254588986433,
+      "grad_norm": 1.6974566149769976,
+      "learning_rate": 9.993238112138583e-05,
+      "loss": 3.9319,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3449853684490556,
+      "grad_norm": 1.9609388835156845,
+      "learning_rate": 9.99317357456848e-05,
+      "loss": 3.8935,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3451981910082469,
+      "grad_norm": 2.070503888402769,
+      "learning_rate": 9.993108730686427e-05,
+      "loss": 3.9143,
+      "step": 1622
+    },
+    {
+      "epoch": 0.34541101356743814,
+      "grad_norm": 1.4478931594970357,
+      "learning_rate": 9.9930435804964e-05,
+      "loss": 3.9311,
+      "step": 1623
+    },
+    {
+      "epoch": 0.34562383612662945,
+      "grad_norm": 2.051722026772044,
+      "learning_rate": 9.9929781240024e-05,
+      "loss": 3.9299,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3458366586858207,
+      "grad_norm": 1.4885138570862244,
+      "learning_rate": 9.992912361208439e-05,
+      "loss": 3.9177,
+      "step": 1625
+    },
+    {
+      "epoch": 0.34604948124501195,
+      "grad_norm": 2.5219141980128055,
+      "learning_rate": 9.992846292118554e-05,
+      "loss": 3.9845,
+      "step": 1626
+    },
+    {
+      "epoch": 0.34626230380420325,
+      "grad_norm": 1.5199798054088085,
+      "learning_rate": 9.992779916736795e-05,
+      "loss": 3.926,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3464751263633945,
+      "grad_norm": 2.073619812345647,
+      "learning_rate": 9.992713235067237e-05,
+      "loss": 3.9091,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3466879489225858,
+      "grad_norm": 1.7941087595844345,
+      "learning_rate": 9.992646247113968e-05,
+      "loss": 3.8933,
+      "step": 1629
+    },
+    {
+      "epoch": 0.34690077148177706,
+      "grad_norm": 2.2216638144620973,
+      "learning_rate": 9.9925789528811e-05,
+      "loss": 3.9283,
+      "step": 1630
+    },
+    {
+      "epoch": 0.34711359404096837,
+      "grad_norm": 1.7924725829921953,
+      "learning_rate": 9.992511352372757e-05,
+      "loss": 3.925,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3473264166001596,
+      "grad_norm": 1.967036938454875,
+      "learning_rate": 9.992443445593092e-05,
+      "loss": 3.9241,
+      "step": 1632
+    },
+    {
+      "epoch": 0.34753923915935087,
+      "grad_norm": 1.825151093692288,
+      "learning_rate": 9.992375232546267e-05,
+      "loss": 3.9691,
+      "step": 1633
+    },
+    {
+      "epoch": 0.34775206171854217,
+      "grad_norm": 2.0045495118437278,
+      "learning_rate": 9.992306713236466e-05,
+      "loss": 3.8903,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3479648842777334,
+      "grad_norm": 1.543791223560812,
+      "learning_rate": 9.992237887667894e-05,
+      "loss": 3.9108,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3481777068369247,
+      "grad_norm": 2.2782065312027915,
+      "learning_rate": 9.992168755844775e-05,
+      "loss": 3.9549,
+      "step": 1636
+    },
+    {
+      "epoch": 0.348390529396116,
+      "grad_norm": 1.6896473295319119,
+      "learning_rate": 9.992099317771346e-05,
+      "loss": 3.8474,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3486033519553073,
+      "grad_norm": 2.0422128867796103,
+      "learning_rate": 9.99202957345187e-05,
+      "loss": 3.9331,
+      "step": 1638
+    },
+    {
+      "epoch": 0.34881617451449853,
+      "grad_norm": 1.7586921890741372,
+      "learning_rate": 9.991959522890622e-05,
+      "loss": 3.8913,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3490289970736898,
+      "grad_norm": 2.0710973892643247,
+      "learning_rate": 9.991889166091904e-05,
+      "loss": 3.9156,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3492418196328811,
+      "grad_norm": 1.5660044561273896,
+      "learning_rate": 9.991818503060027e-05,
+      "loss": 3.8958,
+      "step": 1641
+    },
+    {
+      "epoch": 0.34945464219207234,
+      "grad_norm": 2.1784238419556883,
+      "learning_rate": 9.99174753379933e-05,
+      "loss": 3.9513,
+      "step": 1642
+    },
+    {
+      "epoch": 0.34966746475126365,
+      "grad_norm": 1.2493410858289327,
+      "learning_rate": 9.991676258314166e-05,
+      "loss": 3.8903,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3498802873104549,
+      "grad_norm": 2.3888183394675875,
+      "learning_rate": 9.991604676608906e-05,
+      "loss": 3.9045,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3500931098696462,
+      "grad_norm": 1.537292183156622,
+      "learning_rate": 9.991532788687941e-05,
+      "loss": 3.9069,
+      "step": 1645
+    },
+    {
+      "epoch": 0.35030593242883745,
+      "grad_norm": 2.02387969086736,
+      "learning_rate": 9.991460594555684e-05,
+      "loss": 3.8427,
+      "step": 1646
+    },
+    {
+      "epoch": 0.35051875498802876,
+      "grad_norm": 1.781952131596719,
+      "learning_rate": 9.99138809421656e-05,
+      "loss": 3.9433,
+      "step": 1647
+    },
+    {
+      "epoch": 0.35073157754722,
+      "grad_norm": 1.8801042249902942,
+      "learning_rate": 9.991315287675019e-05,
+      "loss": 3.9216,
+      "step": 1648
+    },
+    {
+      "epoch": 0.35094440010641126,
+      "grad_norm": 1.927014719192349,
+      "learning_rate": 9.991242174935526e-05,
+      "loss": 3.9013,
+      "step": 1649
+    },
+    {
+      "epoch": 0.35115722266560256,
+      "grad_norm": 1.7449291446366315,
+      "learning_rate": 9.991168756002569e-05,
+      "loss": 3.9466,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3513700452247938,
+      "grad_norm": 1.892730361750442,
+      "learning_rate": 9.99109503088065e-05,
+      "loss": 3.892,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3515828677839851,
+      "grad_norm": 1.9815671507160142,
+      "learning_rate": 9.991020999574289e-05,
+      "loss": 3.9092,
+      "step": 1652
+    },
+    {
+      "epoch": 0.35179569034317637,
+      "grad_norm": 1.576907138950381,
+      "learning_rate": 9.990946662088033e-05,
+      "loss": 3.8982,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3520085129023677,
+      "grad_norm": 1.9246689330646005,
+      "learning_rate": 9.990872018426438e-05,
+      "loss": 3.8514,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3522213354615589,
+      "grad_norm": 1.7989351169322727,
+      "learning_rate": 9.990797068594086e-05,
+      "loss": 3.8991,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3524341580207502,
+      "grad_norm": 2.0062554249437032,
+      "learning_rate": 9.990721812595574e-05,
+      "loss": 3.8868,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3526469805799415,
+      "grad_norm": 1.7657781054622028,
+      "learning_rate": 9.990646250435517e-05,
+      "loss": 3.9021,
+      "step": 1657
+    },
+    {
+      "epoch": 0.35285980313913273,
+      "grad_norm": 2.104994147164168,
+      "learning_rate": 9.990570382118551e-05,
+      "loss": 3.8823,
+      "step": 1658
+    },
+    {
+      "epoch": 0.35307262569832404,
+      "grad_norm": 1.3588272633041238,
+      "learning_rate": 9.990494207649332e-05,
+      "loss": 3.8835,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3532854482575153,
+      "grad_norm": 2.6117825169461972,
+      "learning_rate": 9.990417727032533e-05,
+      "loss": 3.9128,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3534982708167066,
+      "grad_norm": 1.4557521651665584,
+      "learning_rate": 9.990340940272844e-05,
+      "loss": 3.8709,
+      "step": 1661
+    },
+    {
+      "epoch": 0.35371109337589784,
+      "grad_norm": 2.325843131587253,
+      "learning_rate": 9.990263847374976e-05,
+      "loss": 3.8668,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3539239159350891,
+      "grad_norm": 1.468311064530581,
+      "learning_rate": 9.990186448343658e-05,
+      "loss": 3.8388,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3541367384942804,
+      "grad_norm": 1.9432180653679663,
+      "learning_rate": 9.990108743183639e-05,
+      "loss": 3.8761,
+      "step": 1664
+    },
+    {
+      "epoch": 0.35434956105347165,
+      "grad_norm": 1.7988047833069707,
+      "learning_rate": 9.990030731899687e-05,
+      "loss": 3.8576,
+      "step": 1665
+    },
+    {
+      "epoch": 0.35456238361266296,
+      "grad_norm": 1.4518409960239969,
+      "learning_rate": 9.989952414496586e-05,
+      "loss": 3.8827,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3547752061718542,
+      "grad_norm": 1.9540800574194195,
+      "learning_rate": 9.98987379097914e-05,
+      "loss": 3.8714,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3549880287310455,
+      "grad_norm": 1.8970186920519527,
+      "learning_rate": 9.989794861352173e-05,
+      "loss": 3.9046,
+      "step": 1668
+    },
+    {
+      "epoch": 0.35520085129023676,
+      "grad_norm": 1.6167077274577983,
+      "learning_rate": 9.989715625620528e-05,
+      "loss": 3.8801,
+      "step": 1669
+    },
+    {
+      "epoch": 0.355413673849428,
+      "grad_norm": 2.202831911981558,
+      "learning_rate": 9.989636083789065e-05,
+      "loss": 3.9396,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3556264964086193,
+      "grad_norm": 1.6685026425453673,
+      "learning_rate": 9.989556235862663e-05,
+      "loss": 3.8725,
+      "step": 1671
+    },
+    {
+      "epoch": 0.35583931896781057,
+      "grad_norm": 1.7870858745289626,
+      "learning_rate": 9.98947608184622e-05,
+      "loss": 3.8769,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3560521415270019,
+      "grad_norm": 1.6903910193426572,
+      "learning_rate": 9.989395621744657e-05,
+      "loss": 3.9109,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3562649640861931,
+      "grad_norm": 1.8225080105238003,
+      "learning_rate": 9.989314855562906e-05,
+      "loss": 3.8524,
+      "step": 1674
+    },
+    {
+      "epoch": 0.35647778664538443,
+      "grad_norm": 1.9673046184710226,
+      "learning_rate": 9.98923378330592e-05,
+      "loss": 3.8785,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3566906092045757,
+      "grad_norm": 1.9016182839211988,
+      "learning_rate": 9.989152404978678e-05,
+      "loss": 3.9093,
+      "step": 1676
+    },
+    {
+      "epoch": 0.35690343176376693,
+      "grad_norm": 1.558301931496136,
+      "learning_rate": 9.989070720586168e-05,
+      "loss": 3.8533,
+      "step": 1677
+    },
+    {
+      "epoch": 0.35711625432295824,
+      "grad_norm": 2.154760430102193,
+      "learning_rate": 9.988988730133403e-05,
+      "loss": 3.8362,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3573290768821495,
+      "grad_norm": 1.482842587350976,
+      "learning_rate": 9.988906433625412e-05,
+      "loss": 3.821,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3575418994413408,
+      "grad_norm": 2.271860102873632,
+      "learning_rate": 9.988823831067244e-05,
+      "loss": 3.8665,
+      "step": 1680
+    },
+    {
+      "epoch": 0.35775472200053204,
+      "grad_norm": 1.7326952603610917,
+      "learning_rate": 9.988740922463966e-05,
+      "loss": 3.828,
+      "step": 1681
+    },
+    {
+      "epoch": 0.35796754455972335,
+      "grad_norm": 1.8369944308840185,
+      "learning_rate": 9.988657707820663e-05,
+      "loss": 3.8697,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3581803671189146,
+      "grad_norm": 1.642012936224087,
+      "learning_rate": 9.988574187142444e-05,
+      "loss": 3.8664,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3583931896781059,
+      "grad_norm": 2.003094487285699,
+      "learning_rate": 9.988490360434426e-05,
+      "loss": 3.8844,
+      "step": 1684
+    },
+    {
+      "epoch": 0.35860601223729716,
+      "grad_norm": 1.5355673675953914,
+      "learning_rate": 9.988406227701759e-05,
+      "loss": 3.8188,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3588188347964884,
+      "grad_norm": 1.8861939514882886,
+      "learning_rate": 9.988321788949598e-05,
+      "loss": 3.8444,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3590316573556797,
+      "grad_norm": 1.5009629383156755,
+      "learning_rate": 9.988237044183127e-05,
+      "loss": 3.8792,
+      "step": 1687
+    },
+    {
+      "epoch": 0.35924447991487096,
+      "grad_norm": 1.9225237623523472,
+      "learning_rate": 9.988151993407542e-05,
+      "loss": 3.9044,
+      "step": 1688
+    },
+    {
+      "epoch": 0.35945730247406227,
+      "grad_norm": 1.7181934508257939,
+      "learning_rate": 9.988066636628062e-05,
+      "loss": 3.8471,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3596701250332535,
+      "grad_norm": 1.6789398875309485,
+      "learning_rate": 9.987980973849923e-05,
+      "loss": 3.8062,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3598829475924448,
+      "grad_norm": 1.9551409984969332,
+      "learning_rate": 9.98789500507838e-05,
+      "loss": 3.8883,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3600957701516361,
+      "grad_norm": 1.437624234398852,
+      "learning_rate": 9.987808730318708e-05,
+      "loss": 3.8389,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3603085927108273,
+      "grad_norm": 2.1126580508453454,
+      "learning_rate": 9.987722149576198e-05,
+      "loss": 3.8174,
+      "step": 1693
+    },
+    {
+      "epoch": 0.36052141527001863,
+      "grad_norm": 1.4749175324408836,
+      "learning_rate": 9.987635262856162e-05,
+      "loss": 3.8122,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3607342378292099,
+      "grad_norm": 1.8757659358823107,
+      "learning_rate": 9.98754807016393e-05,
+      "loss": 3.8098,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3609470603884012,
+      "grad_norm": 1.6229162244576096,
+      "learning_rate": 9.987460571504853e-05,
+      "loss": 3.8795,
+      "step": 1696
+    },
+    {
+      "epoch": 0.36115988294759244,
+      "grad_norm": 2.221964426078539,
+      "learning_rate": 9.987372766884294e-05,
+      "loss": 3.8602,
+      "step": 1697
+    },
+    {
+      "epoch": 0.36137270550678374,
+      "grad_norm": 1.2390446386450902,
+      "learning_rate": 9.987284656307644e-05,
+      "loss": 3.9024,
+      "step": 1698
+    },
+    {
+      "epoch": 0.361585528065975,
+      "grad_norm": 2.3778276000249017,
+      "learning_rate": 9.987196239780306e-05,
+      "loss": 3.8539,
+      "step": 1699
+    },
+    {
+      "epoch": 0.36179835062516624,
+      "grad_norm": 1.5139977863730871,
+      "learning_rate": 9.987107517307705e-05,
+      "loss": 3.8361,
+      "step": 1700
+    },
+    {
+      "epoch": 0.36201117318435755,
+      "grad_norm": 2.1468124081106237,
+      "learning_rate": 9.987018488895284e-05,
+      "loss": 3.8552,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3622239957435488,
+      "grad_norm": 1.7569178571153585,
+      "learning_rate": 9.986929154548502e-05,
+      "loss": 3.8339,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3624368183027401,
+      "grad_norm": 1.7226922571092818,
+      "learning_rate": 9.986839514272842e-05,
+      "loss": 3.8378,
+      "step": 1703
+    },
+    {
+      "epoch": 0.36264964086193135,
+      "grad_norm": 2.135572642071241,
+      "learning_rate": 9.986749568073803e-05,
+      "loss": 3.8655,
+      "step": 1704
+    },
+    {
+      "epoch": 0.36286246342112266,
+      "grad_norm": 2.202624251662526,
+      "learning_rate": 9.986659315956901e-05,
+      "loss": 3.8798,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3630752859803139,
+      "grad_norm": 1.6264517484012144,
+      "learning_rate": 9.986568757927675e-05,
+      "loss": 3.8797,
+      "step": 1706
+    },
+    {
+      "epoch": 0.36328810853950516,
+      "grad_norm": 1.7937933446350038,
+      "learning_rate": 9.986477893991679e-05,
+      "loss": 3.8457,
+      "step": 1707
+    },
+    {
+      "epoch": 0.36350093109869647,
+      "grad_norm": 1.6847206791156795,
+      "learning_rate": 9.986386724154486e-05,
+      "loss": 3.8232,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3637137536578877,
+      "grad_norm": 1.6846222063993437,
+      "learning_rate": 9.98629524842169e-05,
+      "loss": 3.7739,
+      "step": 1709
+    },
+    {
+      "epoch": 0.363926576217079,
+      "grad_norm": 1.7646856024981903,
+      "learning_rate": 9.986203466798906e-05,
+      "loss": 3.8346,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3641393987762703,
+      "grad_norm": 1.646552660477093,
+      "learning_rate": 9.986111379291759e-05,
+      "loss": 3.8568,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3643522213354616,
+      "grad_norm": 1.918102293998163,
+      "learning_rate": 9.986018985905901e-05,
+      "loss": 3.8387,
+      "step": 1712
+    },
+    {
+      "epoch": 0.36456504389465283,
+      "grad_norm": 1.8273863785640896,
+      "learning_rate": 9.985926286647e-05,
+      "loss": 3.8434,
+      "step": 1713
+    },
+    {
+      "epoch": 0.36477786645384414,
+      "grad_norm": 1.699754337248037,
+      "learning_rate": 9.985833281520742e-05,
+      "loss": 3.8108,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3649906890130354,
+      "grad_norm": 1.9214483996123692,
+      "learning_rate": 9.985739970532833e-05,
+      "loss": 3.8084,
+      "step": 1715
+    },
+    {
+      "epoch": 0.36520351157222664,
+      "grad_norm": 1.7114751534218267,
+      "learning_rate": 9.985646353688996e-05,
+      "loss": 3.7797,
+      "step": 1716
+    },
+    {
+      "epoch": 0.36541633413141794,
+      "grad_norm": 1.524916572542435,
+      "learning_rate": 9.985552430994976e-05,
+      "loss": 3.8444,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3656291566906092,
+      "grad_norm": 1.8047564545058756,
+      "learning_rate": 9.985458202456534e-05,
+      "loss": 3.8747,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3658419792498005,
+      "grad_norm": 1.4840008684598753,
+      "learning_rate": 9.98536366807945e-05,
+      "loss": 3.831,
+      "step": 1719
+    },
+    {
+      "epoch": 0.36605480180899175,
+      "grad_norm": 2.1018427582169674,
+      "learning_rate": 9.985268827869523e-05,
+      "loss": 3.8495,
+      "step": 1720
+    },
+    {
+      "epoch": 0.36626762436818305,
+      "grad_norm": 1.6809806090522024,
+      "learning_rate": 9.985173681832573e-05,
+      "loss": 3.8194,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3664804469273743,
+      "grad_norm": 1.736421685961997,
+      "learning_rate": 9.985078229974437e-05,
+      "loss": 3.8284,
+      "step": 1722
+    },
+    {
+      "epoch": 0.36669326948656555,
+      "grad_norm": 1.6964940932297505,
+      "learning_rate": 9.984982472300968e-05,
+      "loss": 3.8182,
+      "step": 1723
+    },
+    {
+      "epoch": 0.36690609204575686,
+      "grad_norm": 1.693471952359889,
+      "learning_rate": 9.984886408818041e-05,
+      "loss": 3.7793,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3671189146049481,
+      "grad_norm": 1.8683097943093017,
+      "learning_rate": 9.984790039531551e-05,
+      "loss": 3.8269,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3673317371641394,
+      "grad_norm": 1.3676380630107463,
+      "learning_rate": 9.984693364447408e-05,
+      "loss": 3.8134,
+      "step": 1726
+    },
+    {
+      "epoch": 0.36754455972333067,
+      "grad_norm": 1.7975765980586877,
+      "learning_rate": 9.984596383571543e-05,
+      "loss": 3.817,
+      "step": 1727
+    },
+    {
+      "epoch": 0.36775738228252197,
+      "grad_norm": 1.5686397067195228,
+      "learning_rate": 9.984499096909904e-05,
+      "loss": 3.7652,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3679702048417132,
+      "grad_norm": 1.7171496204107006,
+      "learning_rate": 9.984401504468464e-05,
+      "loss": 3.7925,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3681830274009045,
+      "grad_norm": 1.6814618738255545,
+      "learning_rate": 9.984303606253207e-05,
+      "loss": 3.8254,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3683958499600958,
+      "grad_norm": 1.8986249004668017,
+      "learning_rate": 9.984205402270138e-05,
+      "loss": 3.8147,
+      "step": 1731
+    },
+    {
+      "epoch": 0.36860867251928703,
+      "grad_norm": 1.759706759480557,
+      "learning_rate": 9.98410689252528e-05,
+      "loss": 3.8103,
+      "step": 1732
+    },
+    {
+      "epoch": 0.36882149507847833,
+      "grad_norm": 1.9130421284471273,
+      "learning_rate": 9.98400807702468e-05,
+      "loss": 3.7859,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3690343176376696,
+      "grad_norm": 1.3744085875443373,
+      "learning_rate": 9.983908955774397e-05,
+      "loss": 3.8217,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3692471401968609,
+      "grad_norm": 1.97363123285155,
+      "learning_rate": 9.983809528780513e-05,
+      "loss": 3.7768,
+      "step": 1735
+    },
+    {
+      "epoch": 0.36945996275605214,
+      "grad_norm": 1.4944672577191975,
+      "learning_rate": 9.983709796049128e-05,
+      "loss": 3.823,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3696727853152434,
+      "grad_norm": 1.7729658751483626,
+      "learning_rate": 9.983609757586358e-05,
+      "loss": 3.8488,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3698856078744347,
+      "grad_norm": 1.3192893730356077,
+      "learning_rate": 9.983509413398341e-05,
+      "loss": 3.7695,
+      "step": 1738
+    },
+    {
+      "epoch": 0.37009843043362595,
+      "grad_norm": 1.9334695354507512,
+      "learning_rate": 9.983408763491235e-05,
+      "loss": 3.8006,
+      "step": 1739
+    },
+    {
+      "epoch": 0.37031125299281725,
+      "grad_norm": 1.6272682454590983,
+      "learning_rate": 9.983307807871212e-05,
+      "loss": 3.8178,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3705240755520085,
+      "grad_norm": 2.287378244480739,
+      "learning_rate": 9.983206546544466e-05,
+      "loss": 3.7888,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3707368981111998,
+      "grad_norm": 1.5629055025227971,
+      "learning_rate": 9.98310497951721e-05,
+      "loss": 3.8513,
+      "step": 1742
+    },
+    {
+      "epoch": 0.37094972067039106,
+      "grad_norm": 1.849790344955644,
+      "learning_rate": 9.983003106795672e-05,
+      "loss": 3.7801,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3711625432295823,
+      "grad_norm": 1.7863975549279456,
+      "learning_rate": 9.982900928386103e-05,
+      "loss": 3.7606,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3713753657887736,
+      "grad_norm": 1.4690715496158369,
+      "learning_rate": 9.982798444294772e-05,
+      "loss": 3.7888,
+      "step": 1745
+    },
+    {
+      "epoch": 0.37158818834796487,
+      "grad_norm": 2.2464236029531417,
+      "learning_rate": 9.982695654527965e-05,
+      "loss": 3.7862,
+      "step": 1746
+    },
+    {
+      "epoch": 0.37180101090715617,
+      "grad_norm": 1.4000703441149394,
+      "learning_rate": 9.982592559091988e-05,
+      "loss": 3.8093,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3720138334663474,
+      "grad_norm": 2.02600717076753,
+      "learning_rate": 9.982489157993167e-05,
+      "loss": 3.8187,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3722266560255387,
+      "grad_norm": 1.8168624605287658,
+      "learning_rate": 9.982385451237843e-05,
+      "loss": 3.8707,
+      "step": 1749
+    },
+    {
+      "epoch": 0.37243947858473,
+      "grad_norm": 1.6057227793860889,
+      "learning_rate": 9.982281438832378e-05,
+      "loss": 3.8119,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3726523011439213,
+      "grad_norm": 1.8001064397577562,
+      "learning_rate": 9.982177120783154e-05,
+      "loss": 3.7938,
+      "step": 1751
+    },
+    {
+      "epoch": 0.37286512370311253,
+      "grad_norm": 1.525581752672076,
+      "learning_rate": 9.982072497096571e-05,
+      "loss": 3.7934,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3730779462623038,
+      "grad_norm": 1.8027899807131622,
+      "learning_rate": 9.981967567779047e-05,
+      "loss": 3.7763,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3732907688214951,
+      "grad_norm": 1.507584838900792,
+      "learning_rate": 9.981862332837017e-05,
+      "loss": 3.7423,
+      "step": 1754
+    },
+    {
+      "epoch": 0.37350359138068634,
+      "grad_norm": 1.5664827320102415,
+      "learning_rate": 9.981756792276938e-05,
+      "loss": 3.7952,
+      "step": 1755
+    },
+    {
+      "epoch": 0.37371641393987765,
+      "grad_norm": 1.924427411079622,
+      "learning_rate": 9.981650946105285e-05,
+      "loss": 3.7375,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3739292364990689,
+      "grad_norm": 1.7275984526233905,
+      "learning_rate": 9.981544794328552e-05,
+      "loss": 3.8201,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3741420590582602,
+      "grad_norm": 1.6187509929526644,
+      "learning_rate": 9.98143833695325e-05,
+      "loss": 3.8041,
+      "step": 1758
+    },
+    {
+      "epoch": 0.37435488161745145,
+      "grad_norm": 1.808084723593486,
+      "learning_rate": 9.981331573985908e-05,
+      "loss": 3.7592,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3745677041766427,
+      "grad_norm": 1.3466182397438482,
+      "learning_rate": 9.981224505433079e-05,
+      "loss": 3.7796,
+      "step": 1760
+    },
+    {
+      "epoch": 0.374780526735834,
+      "grad_norm": 1.9935447081180317,
+      "learning_rate": 9.981117131301329e-05,
+      "loss": 3.7619,
+      "step": 1761
+    },
+    {
+      "epoch": 0.37499334929502526,
+      "grad_norm": 1.6023905594255279,
+      "learning_rate": 9.981009451597245e-05,
+      "loss": 3.7375,
+      "step": 1762
+    },
+    {
+      "epoch": 0.37520617185421656,
+      "grad_norm": 1.8995968162081127,
+      "learning_rate": 9.980901466327433e-05,
+      "loss": 3.8276,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3754189944134078,
+      "grad_norm": 1.5538422693401093,
+      "learning_rate": 9.980793175498517e-05,
+      "loss": 3.8254,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3756318169725991,
+      "grad_norm": 2.058792797979574,
+      "learning_rate": 9.980684579117143e-05,
+      "loss": 3.7705,
+      "step": 1765
+    },
+    {
+      "epoch": 0.37584463953179037,
+      "grad_norm": 1.4707986219228755,
+      "learning_rate": 9.98057567718997e-05,
+      "loss": 3.7513,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3760574620909816,
+      "grad_norm": 1.8278298323690512,
+      "learning_rate": 9.98046646972368e-05,
+      "loss": 3.7634,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3762702846501729,
+      "grad_norm": 1.6121568880804906,
+      "learning_rate": 9.98035695672497e-05,
+      "loss": 3.7674,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3764831072093642,
+      "grad_norm": 1.8606006798395798,
+      "learning_rate": 9.980247138200562e-05,
+      "loss": 3.7615,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3766959297685555,
+      "grad_norm": 1.6858770064711732,
+      "learning_rate": 9.98013701415719e-05,
+      "loss": 3.7817,
+      "step": 1770
+    },
+    {
+      "epoch": 0.37690875232774673,
+      "grad_norm": 1.9570305994979096,
+      "learning_rate": 9.980026584601611e-05,
+      "loss": 3.8009,
+      "step": 1771
+    },
+    {
+      "epoch": 0.37712157488693804,
+      "grad_norm": 1.6043611091748293,
+      "learning_rate": 9.979915849540599e-05,
+      "loss": 3.7677,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3773343974461293,
+      "grad_norm": 1.6747147028587241,
+      "learning_rate": 9.979804808980949e-05,
+      "loss": 3.7691,
+      "step": 1773
+    },
+    {
+      "epoch": 0.37754722000532054,
+      "grad_norm": 1.8220719497332698,
+      "learning_rate": 9.979693462929471e-05,
+      "loss": 3.7839,
+      "step": 1774
+    },
+    {
+      "epoch": 0.37776004256451184,
+      "grad_norm": 1.563164796353498,
+      "learning_rate": 9.979581811392996e-05,
+      "loss": 3.8068,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3779728651237031,
+      "grad_norm": 1.769370315566917,
+      "learning_rate": 9.979469854378373e-05,
+      "loss": 3.8268,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3781856876828944,
+      "grad_norm": 1.4706547071000464,
+      "learning_rate": 9.97935759189247e-05,
+      "loss": 3.747,
+      "step": 1777
+    },
+    {
+      "epoch": 0.37839851024208565,
+      "grad_norm": 1.5567877064997215,
+      "learning_rate": 9.979245023942174e-05,
+      "loss": 3.7284,
+      "step": 1778
+    },
+    {
+      "epoch": 0.37861133280127696,
+      "grad_norm": 1.576287211612639,
+      "learning_rate": 9.979132150534392e-05,
+      "loss": 3.7706,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3788241553604682,
+      "grad_norm": 1.6534587798066098,
+      "learning_rate": 9.979018971676048e-05,
+      "loss": 3.7846,
+      "step": 1780
+    },
+    {
+      "epoch": 0.37903697791965946,
+      "grad_norm": 1.4127079631148325,
+      "learning_rate": 9.978905487374084e-05,
+      "loss": 3.7682,
+      "step": 1781
+    },
+    {
+      "epoch": 0.37924980047885076,
+      "grad_norm": 1.8978072921365683,
+      "learning_rate": 9.978791697635461e-05,
+      "loss": 3.7889,
+      "step": 1782
+    },
+    {
+      "epoch": 0.379462623038042,
+      "grad_norm": 1.587895099493056,
+      "learning_rate": 9.978677602467163e-05,
+      "loss": 3.7871,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3796754455972333,
+      "grad_norm": 2.047542694680888,
+      "learning_rate": 9.978563201876185e-05,
+      "loss": 3.7356,
+      "step": 1784
+    },
+    {
+      "epoch": 0.37988826815642457,
+      "grad_norm": 1.476608348518707,
+      "learning_rate": 9.978448495869548e-05,
+      "loss": 3.7747,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3801010907156159,
+      "grad_norm": 1.7893968546972998,
+      "learning_rate": 9.978333484454288e-05,
+      "loss": 3.7181,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3803139132748071,
+      "grad_norm": 1.7105911977379014,
+      "learning_rate": 9.978218167637462e-05,
+      "loss": 3.7602,
+      "step": 1787
+    },
+    {
+      "epoch": 0.38052673583399843,
+      "grad_norm": 1.651464771403367,
+      "learning_rate": 9.97810254542614e-05,
+      "loss": 3.7702,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3807395583931897,
+      "grad_norm": 1.6469184831750336,
+      "learning_rate": 9.97798661782742e-05,
+      "loss": 3.7612,
+      "step": 1789
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.4747300202017595,
+      "learning_rate": 9.977870384848409e-05,
+      "loss": 3.7577,
+      "step": 1790
+    },
+    {
+      "epoch": 0.38116520351157224,
+      "grad_norm": 1.8702289474329212,
+      "learning_rate": 9.97775384649624e-05,
+      "loss": 3.736,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3813780260707635,
+      "grad_norm": 1.7352087719215192,
+      "learning_rate": 9.977637002778063e-05,
+      "loss": 3.7657,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3815908486299548,
+      "grad_norm": 1.776932037645889,
+      "learning_rate": 9.977519853701047e-05,
+      "loss": 3.7533,
+      "step": 1793
+    },
+    {
+      "epoch": 0.38180367118914604,
+      "grad_norm": 1.4170026831604106,
+      "learning_rate": 9.977402399272373e-05,
+      "loss": 3.7705,
+      "step": 1794
+    },
+    {
+      "epoch": 0.38201649374833735,
+      "grad_norm": 1.6094755247013082,
+      "learning_rate": 9.977284639499251e-05,
+      "loss": 3.7076,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3822293163075286,
+      "grad_norm": 1.6041585126352136,
+      "learning_rate": 9.977166574388906e-05,
+      "loss": 3.7838,
+      "step": 1796
+    },
+    {
+      "epoch": 0.38244213886671985,
+      "grad_norm": 1.6234324518703585,
+      "learning_rate": 9.977048203948576e-05,
+      "loss": 3.7572,
+      "step": 1797
+    },
+    {
+      "epoch": 0.38265496142591116,
+      "grad_norm": 1.7369089824702106,
+      "learning_rate": 9.976929528185528e-05,
+      "loss": 3.7319,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3828677839851024,
+      "grad_norm": 1.7901400338848223,
+      "learning_rate": 9.976810547107038e-05,
+      "loss": 3.7617,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3830806065442937,
+      "grad_norm": 1.7610806795560399,
+      "learning_rate": 9.976691260720407e-05,
+      "loss": 3.7991,
+      "step": 1800
+    },
+    {
+      "epoch": 0.38329342910348496,
+      "grad_norm": 1.4548045846173423,
+      "learning_rate": 9.976571669032954e-05,
+      "loss": 3.7024,
+      "step": 1801
+    },
+    {
+      "epoch": 0.38350625166267627,
+      "grad_norm": 1.968119329613715,
+      "learning_rate": 9.976451772052013e-05,
+      "loss": 3.781,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3837190742218675,
+      "grad_norm": 1.565625326765906,
+      "learning_rate": 9.976331569784939e-05,
+      "loss": 3.7749,
+      "step": 1803
+    },
+    {
+      "epoch": 0.38393189678105877,
+      "grad_norm": 2.154938126011724,
+      "learning_rate": 9.976211062239109e-05,
+      "loss": 3.7463,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3841447193402501,
+      "grad_norm": 1.3516405845638617,
+      "learning_rate": 9.976090249421912e-05,
+      "loss": 3.7547,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3843575418994413,
+      "grad_norm": 1.7527624059396612,
+      "learning_rate": 9.975969131340763e-05,
+      "loss": 3.7184,
+      "step": 1806
+    },
+    {
+      "epoch": 0.38457036445863263,
+      "grad_norm": 1.50986208112751,
+      "learning_rate": 9.975847708003089e-05,
+      "loss": 3.7957,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3847831870178239,
+      "grad_norm": 2.0345269460751334,
+      "learning_rate": 9.975725979416341e-05,
+      "loss": 3.7735,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3849960095770152,
+      "grad_norm": 1.4542012155309514,
+      "learning_rate": 9.975603945587987e-05,
+      "loss": 3.7235,
+      "step": 1809
+    },
+    {
+      "epoch": 0.38520883213620644,
+      "grad_norm": 1.8142914627852795,
+      "learning_rate": 9.97548160652551e-05,
+      "loss": 3.755,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3854216546953977,
+      "grad_norm": 1.372375166520578,
+      "learning_rate": 9.975358962236419e-05,
+      "loss": 3.7483,
+      "step": 1811
+    },
+    {
+      "epoch": 0.385634477254589,
+      "grad_norm": 1.8592910966583844,
+      "learning_rate": 9.975236012728235e-05,
+      "loss": 3.7545,
+      "step": 1812
+    },
+    {
+      "epoch": 0.38584729981378024,
+      "grad_norm": 1.637494507094227,
+      "learning_rate": 9.975112758008502e-05,
+      "loss": 3.7409,
+      "step": 1813
+    },
+    {
+      "epoch": 0.38606012237297155,
+      "grad_norm": 1.6144476127391192,
+      "learning_rate": 9.97498919808478e-05,
+      "loss": 3.7141,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3862729449321628,
+      "grad_norm": 1.8098260412976042,
+      "learning_rate": 9.97486533296465e-05,
+      "loss": 3.7366,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3864857674913541,
+      "grad_norm": 1.4265314894124559,
+      "learning_rate": 9.974741162655709e-05,
+      "loss": 3.7508,
+      "step": 1816
+    },
+    {
+      "epoch": 0.38669859005054535,
+      "grad_norm": 1.9632263808498074,
+      "learning_rate": 9.974616687165577e-05,
+      "loss": 3.7844,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3869114126097366,
+      "grad_norm": 1.371049769643636,
+      "learning_rate": 9.974491906501887e-05,
+      "loss": 3.7253,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3871242351689279,
+      "grad_norm": 1.9418052953411442,
+      "learning_rate": 9.974366820672296e-05,
+      "loss": 3.7477,
+      "step": 1819
+    },
+    {
+      "epoch": 0.38733705772811916,
+      "grad_norm": 1.325286038757705,
+      "learning_rate": 9.974241429684478e-05,
+      "loss": 3.7333,
+      "step": 1820
+    },
+    {
+      "epoch": 0.38754988028731047,
+      "grad_norm": 1.9365208651745391,
+      "learning_rate": 9.974115733546123e-05,
+      "loss": 3.7623,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3877627028465017,
+      "grad_norm": 1.5483594213983596,
+      "learning_rate": 9.973989732264944e-05,
+      "loss": 3.749,
+      "step": 1822
+    },
+    {
+      "epoch": 0.387975525405693,
+      "grad_norm": 1.7462746849017645,
+      "learning_rate": 9.973863425848671e-05,
+      "loss": 3.7291,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3881883479648843,
+      "grad_norm": 1.5804888331637794,
+      "learning_rate": 9.973736814305049e-05,
+      "loss": 3.7269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3884011705240756,
+      "grad_norm": 1.7742301146970798,
+      "learning_rate": 9.973609897641849e-05,
+      "loss": 3.7064,
+      "step": 1825
+    },
+    {
+      "epoch": 0.38861399308326683,
+      "grad_norm": 1.9934849762445799,
+      "learning_rate": 9.973482675866854e-05,
+      "loss": 3.7394,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3888268156424581,
+      "grad_norm": 1.4399392541720046,
+      "learning_rate": 9.973355148987871e-05,
+      "loss": 3.719,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3890396382016494,
+      "grad_norm": 1.9658513642821551,
+      "learning_rate": 9.973227317012721e-05,
+      "loss": 3.7608,
+      "step": 1828
+    },
+    {
+      "epoch": 0.38925246076084064,
+      "grad_norm": 1.5402228308350552,
+      "learning_rate": 9.973099179949248e-05,
+      "loss": 3.7675,
+      "step": 1829
+    },
+    {
+      "epoch": 0.38946528332003194,
+      "grad_norm": 1.8375124254441477,
+      "learning_rate": 9.972970737805311e-05,
+      "loss": 3.7322,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3896781058792232,
+      "grad_norm": 1.5653383146276192,
+      "learning_rate": 9.972841990588791e-05,
+      "loss": 3.7256,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3898909284384145,
+      "grad_norm": 1.713703013856306,
+      "learning_rate": 9.972712938307586e-05,
+      "loss": 3.7353,
+      "step": 1832
+    },
+    {
+      "epoch": 0.39010375099760575,
+      "grad_norm": 1.5688211255416356,
+      "learning_rate": 9.972583580969612e-05,
+      "loss": 3.7415,
+      "step": 1833
+    },
+    {
+      "epoch": 0.390316573556797,
+      "grad_norm": 1.5291940027145485,
+      "learning_rate": 9.972453918582803e-05,
+      "loss": 3.738,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3905293961159883,
+      "grad_norm": 1.4630102263504914,
+      "learning_rate": 9.972323951155118e-05,
+      "loss": 3.7292,
+      "step": 1835
+    },
+    {
+      "epoch": 0.39074221867517955,
+      "grad_norm": 1.7554213320474121,
+      "learning_rate": 9.972193678694526e-05,
+      "loss": 3.7402,
+      "step": 1836
+    },
+    {
+      "epoch": 0.39095504123437086,
+      "grad_norm": 1.6425954659256692,
+      "learning_rate": 9.97206310120902e-05,
+      "loss": 3.6953,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3911678637935621,
+      "grad_norm": 1.833146715381489,
+      "learning_rate": 9.971932218706609e-05,
+      "loss": 3.6772,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3913806863527534,
+      "grad_norm": 1.2910515127912223,
+      "learning_rate": 9.971801031195325e-05,
+      "loss": 3.7275,
+      "step": 1839
+    },
+    {
+      "epoch": 0.39159350891194467,
+      "grad_norm": 1.9189789298711883,
+      "learning_rate": 9.971669538683212e-05,
+      "loss": 3.7673,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3918063314711359,
+      "grad_norm": 1.6213397757110024,
+      "learning_rate": 9.97153774117834e-05,
+      "loss": 3.7314,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3920191540303272,
+      "grad_norm": 1.675045789868007,
+      "learning_rate": 9.971405638688794e-05,
+      "loss": 3.6904,
+      "step": 1842
+    },
+    {
+      "epoch": 0.39223197658951847,
+      "grad_norm": 1.737781512624984,
+      "learning_rate": 9.971273231222678e-05,
+      "loss": 3.7007,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3924447991487098,
+      "grad_norm": 1.5861833356503339,
+      "learning_rate": 9.971140518788111e-05,
+      "loss": 3.6695,
+      "step": 1844
+    },
+    {
+      "epoch": 0.39265762170790103,
+      "grad_norm": 1.8392325841545267,
+      "learning_rate": 9.97100750139324e-05,
+      "loss": 3.6901,
+      "step": 1845
+    },
+    {
+      "epoch": 0.39287044426709233,
+      "grad_norm": 1.4452015454590998,
+      "learning_rate": 9.970874179046219e-05,
+      "loss": 3.7438,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3930832668262836,
+      "grad_norm": 2.1458133151134127,
+      "learning_rate": 9.970740551755233e-05,
+      "loss": 3.6856,
+      "step": 1847
+    },
+    {
+      "epoch": 0.39329608938547483,
+      "grad_norm": 1.1460271968296498,
+      "learning_rate": 9.970606619528475e-05,
+      "loss": 3.7594,
+      "step": 1848
+    },
+    {
+      "epoch": 0.39350891194466614,
+      "grad_norm": 2.183219453704447,
+      "learning_rate": 9.970472382374163e-05,
+      "loss": 3.7354,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3937217345038574,
+      "grad_norm": 1.5319634498131545,
+      "learning_rate": 9.970337840300532e-05,
+      "loss": 3.7448,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3939345570630487,
+      "grad_norm": 1.642115194090818,
+      "learning_rate": 9.970202993315836e-05,
+      "loss": 3.7086,
+      "step": 1851
+    },
+    {
+      "epoch": 0.39414737962223995,
+      "grad_norm": 1.397689013927998,
+      "learning_rate": 9.970067841428346e-05,
+      "loss": 3.7133,
+      "step": 1852
+    },
+    {
+      "epoch": 0.39436020218143125,
+      "grad_norm": 1.788725127482901,
+      "learning_rate": 9.969932384646354e-05,
+      "loss": 3.6868,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3945730247406225,
+      "grad_norm": 1.6302990275655793,
+      "learning_rate": 9.96979662297817e-05,
+      "loss": 3.7188,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3947858472998138,
+      "grad_norm": 1.6659765698741593,
+      "learning_rate": 9.969660556432121e-05,
+      "loss": 3.7189,
+      "step": 1855
+    },
+    {
+      "epoch": 0.39499866985900506,
+      "grad_norm": 1.3308742093949368,
+      "learning_rate": 9.969524185016555e-05,
+      "loss": 3.7305,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3952114924181963,
+      "grad_norm": 1.6679375874832303,
+      "learning_rate": 9.969387508739839e-05,
+      "loss": 3.6823,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3954243149773876,
+      "grad_norm": 1.6004808394005565,
+      "learning_rate": 9.969250527610356e-05,
+      "loss": 3.6797,
+      "step": 1858
+    },
+    {
+      "epoch": 0.39563713753657886,
+      "grad_norm": 1.5874608666964882,
+      "learning_rate": 9.96911324163651e-05,
+      "loss": 3.704,
+      "step": 1859
+    },
+    {
+      "epoch": 0.39584996009577017,
+      "grad_norm": 1.6963141487940052,
+      "learning_rate": 9.968975650826722e-05,
+      "loss": 3.7399,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3960627826549614,
+      "grad_norm": 1.5849534323350822,
+      "learning_rate": 9.968837755189436e-05,
+      "loss": 3.6747,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3962756052141527,
+      "grad_norm": 1.4631675402043787,
+      "learning_rate": 9.968699554733105e-05,
+      "loss": 3.6913,
+      "step": 1862
+    },
+    {
+      "epoch": 0.396488427773344,
+      "grad_norm": 1.6691115981818123,
+      "learning_rate": 9.968561049466214e-05,
+      "loss": 3.7435,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3967012503325352,
+      "grad_norm": 1.6396965892456417,
+      "learning_rate": 9.968422239397254e-05,
+      "loss": 3.7469,
+      "step": 1864
+    },
+    {
+      "epoch": 0.39691407289172653,
+      "grad_norm": 1.6349057043237116,
+      "learning_rate": 9.968283124534746e-05,
+      "loss": 3.695,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3971268954509178,
+      "grad_norm": 1.5949663961559057,
+      "learning_rate": 9.96814370488722e-05,
+      "loss": 3.7079,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3973397180101091,
+      "grad_norm": 1.6596391669192616,
+      "learning_rate": 9.96800398046323e-05,
+      "loss": 3.6912,
+      "step": 1867
+    },
+    {
+      "epoch": 0.39755254056930034,
+      "grad_norm": 1.633226068703343,
+      "learning_rate": 9.967863951271347e-05,
+      "loss": 3.7054,
+      "step": 1868
+    },
+    {
+      "epoch": 0.39776536312849164,
+      "grad_norm": 1.4731845874064953,
+      "learning_rate": 9.967723617320164e-05,
+      "loss": 3.6635,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3979781856876829,
+      "grad_norm": 1.6409359953429168,
+      "learning_rate": 9.967582978618286e-05,
+      "loss": 3.7395,
+      "step": 1870
+    },
+    {
+      "epoch": 0.39819100824687415,
+      "grad_norm": 1.6165909559603486,
+      "learning_rate": 9.967442035174344e-05,
+      "loss": 3.6893,
+      "step": 1871
+    },
+    {
+      "epoch": 0.39840383080606545,
+      "grad_norm": 1.6729824678135579,
+      "learning_rate": 9.96730078699698e-05,
+      "loss": 3.6736,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3986166533652567,
+      "grad_norm": 1.6628884940035733,
+      "learning_rate": 9.967159234094864e-05,
+      "loss": 3.6945,
+      "step": 1873
+    },
+    {
+      "epoch": 0.398829475924448,
+      "grad_norm": 1.4966241545818224,
+      "learning_rate": 9.967017376476679e-05,
+      "loss": 3.7032,
+      "step": 1874
+    },
+    {
+      "epoch": 0.39904229848363926,
+      "grad_norm": 1.4861324904578115,
+      "learning_rate": 9.966875214151123e-05,
+      "loss": 3.7177,
+      "step": 1875
+    },
+    {
+      "epoch": 0.39925512104283056,
+      "grad_norm": 2.0635148578440443,
+      "learning_rate": 9.966732747126924e-05,
+      "loss": 3.6726,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3994679436020218,
+      "grad_norm": 1.3048580390661078,
+      "learning_rate": 9.966589975412814e-05,
+      "loss": 3.6613,
+      "step": 1877
+    },
+    {
+      "epoch": 0.39968076616121306,
+      "grad_norm": 1.845486094281302,
+      "learning_rate": 9.966446899017558e-05,
+      "loss": 3.7014,
+      "step": 1878
+    },
+    {
+      "epoch": 0.39989358872040437,
+      "grad_norm": 1.6308465157837442,
+      "learning_rate": 9.96630351794993e-05,
+      "loss": 3.7061,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4001064112795956,
+      "grad_norm": 1.9275143939652184,
+      "learning_rate": 9.966159832218727e-05,
+      "loss": 3.717,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4003192338387869,
+      "grad_norm": 1.5364373765458383,
+      "learning_rate": 9.966015841832761e-05,
+      "loss": 3.6551,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4005320563979782,
+      "grad_norm": 1.6281320446258984,
+      "learning_rate": 9.965871546800869e-05,
+      "loss": 3.6873,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4007448789571695,
+      "grad_norm": 1.5758765160129375,
+      "learning_rate": 9.965726947131901e-05,
+      "loss": 3.6994,
+      "step": 1883
+    },
+    {
+      "epoch": 0.40095770151636073,
+      "grad_norm": 1.298754924388362,
+      "learning_rate": 9.965582042834728e-05,
+      "loss": 3.6685,
+      "step": 1884
+    },
+    {
+      "epoch": 0.401170524075552,
+      "grad_norm": 2.2454227340595603,
+      "learning_rate": 9.965436833918239e-05,
+      "loss": 3.6975,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4013833466347433,
+      "grad_norm": 1.2420812499015241,
+      "learning_rate": 9.965291320391342e-05,
+      "loss": 3.6432,
+      "step": 1886
+    },
+    {
+      "epoch": 0.40159616919393454,
+      "grad_norm": 2.0034145795058036,
+      "learning_rate": 9.965145502262965e-05,
+      "loss": 3.6721,
+      "step": 1887
+    },
+    {
+      "epoch": 0.40180899175312584,
+      "grad_norm": 1.3767123968750452,
+      "learning_rate": 9.96499937954205e-05,
+      "loss": 3.7323,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4020218143123171,
+      "grad_norm": 2.0767410074919432,
+      "learning_rate": 9.964852952237566e-05,
+      "loss": 3.6861,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4022346368715084,
+      "grad_norm": 1.3357772447141618,
+      "learning_rate": 9.964706220358492e-05,
+      "loss": 3.7031,
+      "step": 1890
+    },
+    {
+      "epoch": 0.40244745943069965,
+      "grad_norm": 1.8458302943478273,
+      "learning_rate": 9.964559183913831e-05,
+      "loss": 3.7025,
+      "step": 1891
+    },
+    {
+      "epoch": 0.40266028198989096,
+      "grad_norm": 1.5649156607176204,
+      "learning_rate": 9.964411842912603e-05,
+      "loss": 3.7256,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4028731045490822,
+      "grad_norm": 1.7199119686625721,
+      "learning_rate": 9.964264197363845e-05,
+      "loss": 3.6669,
+      "step": 1893
+    },
+    {
+      "epoch": 0.40308592710827346,
+      "grad_norm": 1.5842717431363056,
+      "learning_rate": 9.964116247276617e-05,
+      "loss": 3.7121,
+      "step": 1894
+    },
+    {
+      "epoch": 0.40329874966746476,
+      "grad_norm": 1.526177166344799,
+      "learning_rate": 9.963967992659994e-05,
+      "loss": 3.7092,
+      "step": 1895
+    },
+    {
+      "epoch": 0.403511572226656,
+      "grad_norm": 1.6895233986698932,
+      "learning_rate": 9.963819433523071e-05,
+      "loss": 3.7472,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4037243947858473,
+      "grad_norm": 1.489999917932121,
+      "learning_rate": 9.96367056987496e-05,
+      "loss": 3.744,
+      "step": 1897
+    },
+    {
+      "epoch": 0.40393721734503857,
+      "grad_norm": 1.7901678087111785,
+      "learning_rate": 9.963521401724798e-05,
+      "loss": 3.6662,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4041500399042299,
+      "grad_norm": 1.2605668521234672,
+      "learning_rate": 9.963371929081731e-05,
+      "loss": 3.6689,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4043628624634211,
+      "grad_norm": 1.7467178170431932,
+      "learning_rate": 9.96322215195493e-05,
+      "loss": 3.6617,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4045756850226124,
+      "grad_norm": 1.4441818386003393,
+      "learning_rate": 9.963072070353582e-05,
+      "loss": 3.6754,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4047885075818037,
+      "grad_norm": 1.5749447346531065,
+      "learning_rate": 9.962921684286897e-05,
+      "loss": 3.6361,
+      "step": 1902
+    },
+    {
+      "epoch": 0.40500133014099493,
+      "grad_norm": 1.3255487942449569,
+      "learning_rate": 9.962770993764098e-05,
+      "loss": 3.7329,
+      "step": 1903
+    },
+    {
+      "epoch": 0.40521415270018624,
+      "grad_norm": 1.9631826888325186,
+      "learning_rate": 9.962619998794431e-05,
+      "loss": 3.6274,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4054269752593775,
+      "grad_norm": 1.233422053703425,
+      "learning_rate": 9.96246869938716e-05,
+      "loss": 3.6962,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4056397978185688,
+      "grad_norm": 2.083502590495849,
+      "learning_rate": 9.962317095551562e-05,
+      "loss": 3.6941,
+      "step": 1906
+    },
+    {
+      "epoch": 0.40585262037776004,
+      "grad_norm": 1.2866603022144822,
+      "learning_rate": 9.962165187296941e-05,
+      "loss": 3.6366,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4060654429369513,
+      "grad_norm": 1.9243468610438617,
+      "learning_rate": 9.962012974632615e-05,
+      "loss": 3.6879,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4062782654961426,
+      "grad_norm": 1.394083752187571,
+      "learning_rate": 9.961860457567923e-05,
+      "loss": 3.716,
+      "step": 1909
+    },
+    {
+      "epoch": 0.40649108805533385,
+      "grad_norm": 1.7690198631173935,
+      "learning_rate": 9.961707636112218e-05,
+      "loss": 3.6625,
+      "step": 1910
+    },
+    {
+      "epoch": 0.40670391061452515,
+      "grad_norm": 1.5539812282471288,
+      "learning_rate": 9.96155451027488e-05,
+      "loss": 3.6952,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4069167331737164,
+      "grad_norm": 1.4753011037056616,
+      "learning_rate": 9.961401080065297e-05,
+      "loss": 3.6469,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4071295557329077,
+      "grad_norm": 1.7417495727664285,
+      "learning_rate": 9.961247345492886e-05,
+      "loss": 3.6934,
+      "step": 1913
+    },
+    {
+      "epoch": 0.40734237829209896,
+      "grad_norm": 1.5242946625840743,
+      "learning_rate": 9.961093306567075e-05,
+      "loss": 3.6702,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4075552008512902,
+      "grad_norm": 1.3939557867765446,
+      "learning_rate": 9.960938963297317e-05,
+      "loss": 3.6794,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4077680234104815,
+      "grad_norm": 1.8026863138528022,
+      "learning_rate": 9.960784315693077e-05,
+      "loss": 3.6916,
+      "step": 1916
+    },
+    {
+      "epoch": 0.40798084596967277,
+      "grad_norm": 1.4735507387680824,
+      "learning_rate": 9.960629363763842e-05,
+      "loss": 3.6789,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4081936685288641,
+      "grad_norm": 1.8173259099732257,
+      "learning_rate": 9.96047410751912e-05,
+      "loss": 3.6762,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4084064910880553,
+      "grad_norm": 1.479987153138643,
+      "learning_rate": 9.960318546968434e-05,
+      "loss": 3.6685,
+      "step": 1919
+    },
+    {
+      "epoch": 0.40861931364724663,
+      "grad_norm": 1.6575831619016985,
+      "learning_rate": 9.960162682121327e-05,
+      "loss": 3.6921,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4088321362064379,
+      "grad_norm": 1.481258679635087,
+      "learning_rate": 9.960006512987363e-05,
+      "loss": 3.6807,
+      "step": 1921
+    },
+    {
+      "epoch": 0.40904495876562913,
+      "grad_norm": 1.8056683284872226,
+      "learning_rate": 9.95985003957612e-05,
+      "loss": 3.645,
+      "step": 1922
+    },
+    {
+      "epoch": 0.40925778132482044,
+      "grad_norm": 1.4893226979979153,
+      "learning_rate": 9.959693261897197e-05,
+      "loss": 3.6562,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4094706038840117,
+      "grad_norm": 1.8267107411799268,
+      "learning_rate": 9.959536179960212e-05,
+      "loss": 3.6765,
+      "step": 1924
+    },
+    {
+      "epoch": 0.409683426443203,
+      "grad_norm": 1.331287597457362,
+      "learning_rate": 9.959378793774803e-05,
+      "loss": 3.7251,
+      "step": 1925
+    },
+    {
+      "epoch": 0.40989624900239424,
+      "grad_norm": 1.8216539696404221,
+      "learning_rate": 9.959221103350623e-05,
+      "loss": 3.6734,
+      "step": 1926
+    },
+    {
+      "epoch": 0.41010907156158555,
+      "grad_norm": 1.4883831483399959,
+      "learning_rate": 9.959063108697344e-05,
+      "loss": 3.677,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4103218941207768,
+      "grad_norm": 1.6794628924583943,
+      "learning_rate": 9.958904809824663e-05,
+      "loss": 3.6248,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4105347166799681,
+      "grad_norm": 1.3268106192382203,
+      "learning_rate": 9.958746206742289e-05,
+      "loss": 3.6078,
+      "step": 1929
+    },
+    {
+      "epoch": 0.41074753923915935,
+      "grad_norm": 1.6486090231702455,
+      "learning_rate": 9.958587299459949e-05,
+      "loss": 3.6651,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4109603617983506,
+      "grad_norm": 1.4253971113605197,
+      "learning_rate": 9.958428087987397e-05,
+      "loss": 3.6603,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4111731843575419,
+      "grad_norm": 1.9471524285216777,
+      "learning_rate": 9.958268572334395e-05,
+      "loss": 3.6219,
+      "step": 1932
+    },
+    {
+      "epoch": 0.41138600691673316,
+      "grad_norm": 1.2376686814086857,
+      "learning_rate": 9.958108752510731e-05,
+      "loss": 3.6761,
+      "step": 1933
+    },
+    {
+      "epoch": 0.41159882947592447,
+      "grad_norm": 1.6399492905963113,
+      "learning_rate": 9.957948628526207e-05,
+      "loss": 3.664,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4118116520351157,
+      "grad_norm": 1.6411112009136748,
+      "learning_rate": 9.95778820039065e-05,
+      "loss": 3.6429,
+      "step": 1935
+    },
+    {
+      "epoch": 0.412024474594307,
+      "grad_norm": 1.558241643172131,
+      "learning_rate": 9.9576274681139e-05,
+      "loss": 3.6427,
+      "step": 1936
+    },
+    {
+      "epoch": 0.41223729715349827,
+      "grad_norm": 1.607509981710088,
+      "learning_rate": 9.957466431705814e-05,
+      "loss": 3.6678,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4124501197126895,
+      "grad_norm": 1.7655517533617837,
+      "learning_rate": 9.957305091176274e-05,
+      "loss": 3.6924,
+      "step": 1938
+    },
+    {
+      "epoch": 0.41266294227188083,
+      "grad_norm": 1.2790068758782143,
+      "learning_rate": 9.95714344653518e-05,
+      "loss": 3.6073,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4128757648310721,
+      "grad_norm": 1.7827562254098706,
+      "learning_rate": 9.956981497792444e-05,
+      "loss": 3.6162,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4130885873902634,
+      "grad_norm": 1.6025624671726932,
+      "learning_rate": 9.956819244958003e-05,
+      "loss": 3.6532,
+      "step": 1941
+    },
+    {
+      "epoch": 0.41330140994945463,
+      "grad_norm": 1.4539927289128276,
+      "learning_rate": 9.956656688041808e-05,
+      "loss": 3.6663,
+      "step": 1942
+    },
+    {
+      "epoch": 0.41351423250864594,
+      "grad_norm": 1.5660778226022591,
+      "learning_rate": 9.956493827053836e-05,
+      "loss": 3.6775,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4137270550678372,
+      "grad_norm": 1.5889295715662668,
+      "learning_rate": 9.956330662004074e-05,
+      "loss": 3.676,
+      "step": 1944
+    },
+    {
+      "epoch": 0.41393987762702844,
+      "grad_norm": 1.4665762449393704,
+      "learning_rate": 9.956167192902532e-05,
+      "loss": 3.6327,
+      "step": 1945
+    },
+    {
+      "epoch": 0.41415270018621975,
+      "grad_norm": 1.5759922525627934,
+      "learning_rate": 9.95600341975924e-05,
+      "loss": 3.6407,
+      "step": 1946
+    },
+    {
+      "epoch": 0.414365522745411,
+      "grad_norm": 1.3950824796900492,
+      "learning_rate": 9.955839342584244e-05,
+      "loss": 3.6441,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4145783453046023,
+      "grad_norm": 1.6393429352759517,
+      "learning_rate": 9.955674961387609e-05,
+      "loss": 3.6538,
+      "step": 1948
+    },
+    {
+      "epoch": 0.41479116786379355,
+      "grad_norm": 1.3105866415956067,
+      "learning_rate": 9.95551027617942e-05,
+      "loss": 3.6767,
+      "step": 1949
+    },
+    {
+      "epoch": 0.41500399042298486,
+      "grad_norm": 1.718667326247089,
+      "learning_rate": 9.955345286969778e-05,
+      "loss": 3.6387,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4152168129821761,
+      "grad_norm": 1.40108434348274,
+      "learning_rate": 9.955179993768807e-05,
+      "loss": 3.6121,
+      "step": 1951
+    },
+    {
+      "epoch": 0.41542963554136736,
+      "grad_norm": 1.8898936253222636,
+      "learning_rate": 9.955014396586644e-05,
+      "loss": 3.6441,
+      "step": 1952
+    },
+    {
+      "epoch": 0.41564245810055866,
+      "grad_norm": 1.2561332143899129,
+      "learning_rate": 9.954848495433452e-05,
+      "loss": 3.6339,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4158552806597499,
+      "grad_norm": 1.8403872671554398,
+      "learning_rate": 9.954682290319406e-05,
+      "loss": 3.6726,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4160681032189412,
+      "grad_norm": 1.443391856013396,
+      "learning_rate": 9.9545157812547e-05,
+      "loss": 3.6491,
+      "step": 1955
+    },
+    {
+      "epoch": 0.41628092577813247,
+      "grad_norm": 1.5200314660730185,
+      "learning_rate": 9.954348968249551e-05,
+      "loss": 3.6337,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4164937483373238,
+      "grad_norm": 1.85673977914938,
+      "learning_rate": 9.954181851314192e-05,
+      "loss": 3.65,
+      "step": 1957
+    },
+    {
+      "epoch": 0.416706570896515,
+      "grad_norm": 1.6174484842550616,
+      "learning_rate": 9.954014430458877e-05,
+      "loss": 3.6077,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4169193934557063,
+      "grad_norm": 1.4887340248563563,
+      "learning_rate": 9.953846705693875e-05,
+      "loss": 3.6253,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4171322160148976,
+      "grad_norm": 1.756381995726698,
+      "learning_rate": 9.953678677029472e-05,
+      "loss": 3.6486,
+      "step": 1960
+    },
+    {
+      "epoch": 0.41734503857408883,
+      "grad_norm": 1.3079470211216147,
+      "learning_rate": 9.95351034447598e-05,
+      "loss": 3.6789,
+      "step": 1961
+    },
+    {
+      "epoch": 0.41755786113328014,
+      "grad_norm": 2.239029506370235,
+      "learning_rate": 9.953341708043724e-05,
+      "loss": 3.6247,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4177706836924714,
+      "grad_norm": 1.2242454422508438,
+      "learning_rate": 9.953172767743051e-05,
+      "loss": 3.6386,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4179835062516627,
+      "grad_norm": 1.949336687096158,
+      "learning_rate": 9.953003523584323e-05,
+      "loss": 3.6467,
+      "step": 1964
+    },
+    {
+      "epoch": 0.41819632881085395,
+      "grad_norm": 1.3936171298103701,
+      "learning_rate": 9.952833975577922e-05,
+      "loss": 3.6424,
+      "step": 1965
+    },
+    {
+      "epoch": 0.41840915137004525,
+      "grad_norm": 1.8059463092059198,
+      "learning_rate": 9.95266412373425e-05,
+      "loss": 3.6681,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4186219739292365,
+      "grad_norm": 1.74332569047388,
+      "learning_rate": 9.952493968063728e-05,
+      "loss": 3.6683,
+      "step": 1967
+    },
+    {
+      "epoch": 0.41883479648842775,
+      "grad_norm": 1.528199514822353,
+      "learning_rate": 9.952323508576791e-05,
+      "loss": 3.624,
+      "step": 1968
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 1.4509306439322318,
+      "learning_rate": 9.9521527452839e-05,
+      "loss": 3.6396,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4192604416068103,
+      "grad_norm": 1.6827199852394896,
+      "learning_rate": 9.951981678195528e-05,
+      "loss": 3.5933,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4194732641660016,
+      "grad_norm": 1.349791465911406,
+      "learning_rate": 9.951810307322172e-05,
+      "loss": 3.6175,
+      "step": 1971
+    },
+    {
+      "epoch": 0.41968608672519286,
+      "grad_norm": 1.623836375529104,
+      "learning_rate": 9.951638632674341e-05,
+      "loss": 3.6584,
+      "step": 1972
+    },
+    {
+      "epoch": 0.41989890928438417,
+      "grad_norm": 1.6346181892787703,
+      "learning_rate": 9.95146665426257e-05,
+      "loss": 3.6218,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4201117318435754,
+      "grad_norm": 1.5283805091008575,
+      "learning_rate": 9.951294372097407e-05,
+      "loss": 3.5969,
+      "step": 1974
+    },
+    {
+      "epoch": 0.42032455440276667,
+      "grad_norm": 1.484333507267646,
+      "learning_rate": 9.951121786189422e-05,
+      "loss": 3.5919,
+      "step": 1975
+    },
+    {
+      "epoch": 0.420537376961958,
+      "grad_norm": 1.5439920088420271,
+      "learning_rate": 9.950948896549202e-05,
+      "loss": 3.6001,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4207501995211492,
+      "grad_norm": 1.375586279507394,
+      "learning_rate": 9.950775703187354e-05,
+      "loss": 3.6195,
+      "step": 1977
+    },
+    {
+      "epoch": 0.42096302208034053,
+      "grad_norm": 1.6432653491877043,
+      "learning_rate": 9.950602206114502e-05,
+      "loss": 3.6382,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4211758446395318,
+      "grad_norm": 1.2666954248802451,
+      "learning_rate": 9.950428405341289e-05,
+      "loss": 3.6324,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4213886671987231,
+      "grad_norm": 2.198329117478154,
+      "learning_rate": 9.950254300878378e-05,
+      "loss": 3.5898,
+      "step": 1980
+    },
+    {
+      "epoch": 0.42160148975791434,
+      "grad_norm": 1.1893535658855243,
+      "learning_rate": 9.950079892736448e-05,
+      "loss": 3.6466,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4218143123171056,
+      "grad_norm": 1.9924351186587208,
+      "learning_rate": 9.9499051809262e-05,
+      "loss": 3.6211,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4220271348762969,
+      "grad_norm": 1.355238355514564,
+      "learning_rate": 9.949730165458351e-05,
+      "loss": 3.6335,
+      "step": 1983
+    },
+    {
+      "epoch": 0.42223995743548814,
+      "grad_norm": 1.85000894170157,
+      "learning_rate": 9.949554846343639e-05,
+      "loss": 3.6235,
+      "step": 1984
+    },
+    {
+      "epoch": 0.42245277999467945,
+      "grad_norm": 1.2499919585947035,
+      "learning_rate": 9.949379223592818e-05,
+      "loss": 3.6285,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4226656025538707,
+      "grad_norm": 1.5475175974893245,
+      "learning_rate": 9.94920329721666e-05,
+      "loss": 3.6295,
+      "step": 1986
+    },
+    {
+      "epoch": 0.422878425113062,
+      "grad_norm": 1.429655675603134,
+      "learning_rate": 9.94902706722596e-05,
+      "loss": 3.6569,
+      "step": 1987
+    },
+    {
+      "epoch": 0.42309124767225326,
+      "grad_norm": 1.6513698329775448,
+      "learning_rate": 9.948850533631528e-05,
+      "loss": 3.5957,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4233040702314445,
+      "grad_norm": 1.4436069642073435,
+      "learning_rate": 9.948673696444195e-05,
+      "loss": 3.6296,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4235168927906358,
+      "grad_norm": 1.5514459513794345,
+      "learning_rate": 9.948496555674807e-05,
+      "loss": 3.6544,
+      "step": 1990
+    },
+    {
+      "epoch": 0.42372971534982706,
+      "grad_norm": 1.4968818791963134,
+      "learning_rate": 9.948319111334232e-05,
+      "loss": 3.5695,
+      "step": 1991
+    },
+    {
+      "epoch": 0.42394253790901837,
+      "grad_norm": 1.4944482277085915,
+      "learning_rate": 9.948141363433355e-05,
+      "loss": 3.6398,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4241553604682096,
+      "grad_norm": 1.7634318132502014,
+      "learning_rate": 9.947963311983081e-05,
+      "loss": 3.642,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4243681830274009,
+      "grad_norm": 1.3318687734594734,
+      "learning_rate": 9.947784956994332e-05,
+      "loss": 3.6428,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4245810055865922,
+      "grad_norm": 1.6030278889182774,
+      "learning_rate": 9.947606298478051e-05,
+      "loss": 3.6583,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4247938281457834,
+      "grad_norm": 1.4153880774182779,
+      "learning_rate": 9.947427336445198e-05,
+      "loss": 3.6062,
+      "step": 1996
+    },
+    {
+      "epoch": 0.42500665070497473,
+      "grad_norm": 1.5029497067889046,
+      "learning_rate": 9.947248070906748e-05,
+      "loss": 3.5815,
+      "step": 1997
+    },
+    {
+      "epoch": 0.425219473264166,
+      "grad_norm": 1.55071733005071,
+      "learning_rate": 9.947068501873701e-05,
+      "loss": 3.6219,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4254322958233573,
+      "grad_norm": 1.3124779995808469,
+      "learning_rate": 9.946888629357074e-05,
+      "loss": 3.6264,
+      "step": 1999
+    },
+    {
+      "epoch": 0.42564511838254854,
+      "grad_norm": 1.7083103126037062,
+      "learning_rate": 9.9467084533679e-05,
+      "loss": 3.6134,
+      "step": 2000
+    },
+    {
+      "epoch": 0.42585794094173984,
+      "grad_norm": 1.3250895994460483,
+      "learning_rate": 9.946527973917231e-05,
+      "loss": 3.6424,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4260707635009311,
+      "grad_norm": 1.8141082441310494,
+      "learning_rate": 9.946347191016141e-05,
+      "loss": 3.6387,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4262835860601224,
+      "grad_norm": 1.1359595136586838,
+      "learning_rate": 9.946166104675718e-05,
+      "loss": 3.6172,
+      "step": 2003
+    },
+    {
+      "epoch": 0.42649640861931365,
+      "grad_norm": 1.7391563351794515,
+      "learning_rate": 9.945984714907073e-05,
+      "loss": 3.6109,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4267092311785049,
+      "grad_norm": 1.3277342630471975,
+      "learning_rate": 9.945803021721332e-05,
+      "loss": 3.5786,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4269220537376962,
+      "grad_norm": 1.8080789224337306,
+      "learning_rate": 9.945621025129644e-05,
+      "loss": 3.6755,
+      "step": 2006
+    },
+    {
+      "epoch": 0.42713487629688746,
+      "grad_norm": 1.2808992511836501,
+      "learning_rate": 9.945438725143169e-05,
+      "loss": 3.567,
+      "step": 2007
+    },
+    {
+      "epoch": 0.42734769885607876,
+      "grad_norm": 1.838552045895947,
+      "learning_rate": 9.945256121773095e-05,
+      "loss": 3.6188,
+      "step": 2008
+    },
+    {
+      "epoch": 0.42756052141527,
+      "grad_norm": 1.2677725552597827,
+      "learning_rate": 9.945073215030621e-05,
+      "loss": 3.6239,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4277733439744613,
+      "grad_norm": 1.7752411469312328,
+      "learning_rate": 9.94489000492697e-05,
+      "loss": 3.6381,
+      "step": 2010
+    },
+    {
+      "epoch": 0.42798616653365257,
+      "grad_norm": 1.5392799669227948,
+      "learning_rate": 9.944706491473378e-05,
+      "loss": 3.6044,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4281989890928438,
+      "grad_norm": 1.6522682758324023,
+      "learning_rate": 9.944522674681107e-05,
+      "loss": 3.6491,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4284118116520351,
+      "grad_norm": 1.4531691121371557,
+      "learning_rate": 9.94433855456143e-05,
+      "loss": 3.6765,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4286246342112264,
+      "grad_norm": 1.7561758696856944,
+      "learning_rate": 9.944154131125642e-05,
+      "loss": 3.6001,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4288374567704177,
+      "grad_norm": 1.0002346877257071,
+      "learning_rate": 9.943969404385061e-05,
+      "loss": 3.5872,
+      "step": 2015
+    },
+    {
+      "epoch": 0.42905027932960893,
+      "grad_norm": 2.028685695298667,
+      "learning_rate": 9.943784374351015e-05,
+      "loss": 3.6094,
+      "step": 2016
+    },
+    {
+      "epoch": 0.42926310188880024,
+      "grad_norm": 1.437604188986267,
+      "learning_rate": 9.943599041034855e-05,
+      "loss": 3.5843,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4294759244479915,
+      "grad_norm": 1.5851821478493326,
+      "learning_rate": 9.943413404447955e-05,
+      "loss": 3.5681,
+      "step": 2018
+    },
+    {
+      "epoch": 0.42968874700718274,
+      "grad_norm": 1.3581788195923215,
+      "learning_rate": 9.943227464601697e-05,
+      "loss": 3.6349,
+      "step": 2019
+    },
+    {
+      "epoch": 0.42990156956637404,
+      "grad_norm": 1.9490435448209413,
+      "learning_rate": 9.94304122150749e-05,
+      "loss": 3.6457,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4301143921255653,
+      "grad_norm": 1.3778062817380097,
+      "learning_rate": 9.942854675176761e-05,
+      "loss": 3.6018,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4303272146847566,
+      "grad_norm": 1.8566441129194595,
+      "learning_rate": 9.942667825620952e-05,
+      "loss": 3.591,
+      "step": 2022
+    },
+    {
+      "epoch": 0.43054003724394785,
+      "grad_norm": 1.4498780715994537,
+      "learning_rate": 9.942480672851526e-05,
+      "loss": 3.6056,
+      "step": 2023
+    },
+    {
+      "epoch": 0.43075285980313915,
+      "grad_norm": 1.4516593891998577,
+      "learning_rate": 9.942293216879965e-05,
+      "loss": 3.5805,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4309656823623304,
+      "grad_norm": 1.4215002296343306,
+      "learning_rate": 9.942105457717768e-05,
+      "loss": 3.6395,
+      "step": 2025
+    },
+    {
+      "epoch": 0.43117850492152165,
+      "grad_norm": 1.4726808781303522,
+      "learning_rate": 9.941917395376452e-05,
+      "loss": 3.632,
+      "step": 2026
+    },
+    {
+      "epoch": 0.43139132748071296,
+      "grad_norm": 1.637871619009159,
+      "learning_rate": 9.941729029867556e-05,
+      "loss": 3.597,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4316041500399042,
+      "grad_norm": 1.3713896505687087,
+      "learning_rate": 9.941540361202635e-05,
+      "loss": 3.6015,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4318169725990955,
+      "grad_norm": 1.9061244587693942,
+      "learning_rate": 9.941351389393261e-05,
+      "loss": 3.5895,
+      "step": 2029
+    },
+    {
+      "epoch": 0.43202979515828677,
+      "grad_norm": 1.366785569250444,
+      "learning_rate": 9.94116211445103e-05,
+      "loss": 3.5737,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4322426177174781,
+      "grad_norm": 1.6783478054391776,
+      "learning_rate": 9.940972536387551e-05,
+      "loss": 3.6506,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4324554402766693,
+      "grad_norm": 1.4073809792880507,
+      "learning_rate": 9.940782655214455e-05,
+      "loss": 3.5973,
+      "step": 2032
+    },
+    {
+      "epoch": 0.43266826283586063,
+      "grad_norm": 1.5629166161579489,
+      "learning_rate": 9.940592470943389e-05,
+      "loss": 3.5871,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4328810853950519,
+      "grad_norm": 1.3081678005152646,
+      "learning_rate": 9.940401983586023e-05,
+      "loss": 3.6214,
+      "step": 2034
+    },
+    {
+      "epoch": 0.43309390795424313,
+      "grad_norm": 1.7998308548940172,
+      "learning_rate": 9.940211193154039e-05,
+      "loss": 3.567,
+      "step": 2035
+    },
+    {
+      "epoch": 0.43330673051343443,
+      "grad_norm": 1.3456071043625346,
+      "learning_rate": 9.940020099659144e-05,
+      "loss": 3.6062,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4335195530726257,
+      "grad_norm": 1.678237995405319,
+      "learning_rate": 9.93982870311306e-05,
+      "loss": 3.6178,
+      "step": 2037
+    },
+    {
+      "epoch": 0.433732375631817,
+      "grad_norm": 1.3284002628803242,
+      "learning_rate": 9.939637003527528e-05,
+      "loss": 3.515,
+      "step": 2038
+    },
+    {
+      "epoch": 0.43394519819100824,
+      "grad_norm": 1.611922312856628,
+      "learning_rate": 9.939445000914309e-05,
+      "loss": 3.6353,
+      "step": 2039
+    },
+    {
+      "epoch": 0.43415802075019955,
+      "grad_norm": 1.8026689735655135,
+      "learning_rate": 9.939252695285181e-05,
+      "loss": 3.5879,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4343708433093908,
+      "grad_norm": 1.3979451526531674,
+      "learning_rate": 9.93906008665194e-05,
+      "loss": 3.594,
+      "step": 2041
+    },
+    {
+      "epoch": 0.43458366586858205,
+      "grad_norm": 1.575584694703196,
+      "learning_rate": 9.938867175026403e-05,
+      "loss": 3.6384,
+      "step": 2042
+    },
+    {
+      "epoch": 0.43479648842777335,
+      "grad_norm": 1.468541310939631,
+      "learning_rate": 9.938673960420407e-05,
+      "loss": 3.5712,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4350093109869646,
+      "grad_norm": 1.654145954615367,
+      "learning_rate": 9.938480442845799e-05,
+      "loss": 3.5809,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4352221335461559,
+      "grad_norm": 1.2673298122679113,
+      "learning_rate": 9.938286622314456e-05,
+      "loss": 3.5527,
+      "step": 2045
+    },
+    {
+      "epoch": 0.43543495610534716,
+      "grad_norm": 1.593536822367752,
+      "learning_rate": 9.938092498838266e-05,
+      "loss": 3.5753,
+      "step": 2046
+    },
+    {
+      "epoch": 0.43564777866453847,
+      "grad_norm": 1.3627542476108963,
+      "learning_rate": 9.937898072429136e-05,
+      "loss": 3.5838,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4358606012237297,
+      "grad_norm": 1.716651090620115,
+      "learning_rate": 9.937703343098996e-05,
+      "loss": 3.5693,
+      "step": 2048
+    },
+    {
+      "epoch": 0.43607342378292097,
+      "grad_norm": 1.457121374910887,
+      "learning_rate": 9.937508310859791e-05,
+      "loss": 3.5549,
+      "step": 2049
+    },
+    {
+      "epoch": 0.43628624634211227,
+      "grad_norm": 1.4691867269356929,
+      "learning_rate": 9.937312975723486e-05,
+      "loss": 3.5461,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4364990689013035,
+      "grad_norm": 1.4323602500033288,
+      "learning_rate": 9.937117337702063e-05,
+      "loss": 3.5502,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4367118914604948,
+      "grad_norm": 1.5772527332227806,
+      "learning_rate": 9.936921396807525e-05,
+      "loss": 3.5689,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4369247140196861,
+      "grad_norm": 1.3714989206801038,
+      "learning_rate": 9.93672515305189e-05,
+      "loss": 3.5729,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4371375365788774,
+      "grad_norm": 1.7574056109852036,
+      "learning_rate": 9.936528606447199e-05,
+      "loss": 3.6079,
+      "step": 2054
+    },
+    {
+      "epoch": 0.43735035913806863,
+      "grad_norm": 1.2186808662803112,
+      "learning_rate": 9.936331757005508e-05,
+      "loss": 3.5164,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4375631816972599,
+      "grad_norm": 1.6188940655261708,
+      "learning_rate": 9.936134604738894e-05,
+      "loss": 3.5511,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4377760042564512,
+      "grad_norm": 1.1789689914400816,
+      "learning_rate": 9.935937149659451e-05,
+      "loss": 3.5963,
+      "step": 2057
+    },
+    {
+      "epoch": 0.43798882681564244,
+      "grad_norm": 2.0816829883406527,
+      "learning_rate": 9.935739391779293e-05,
+      "loss": 3.5993,
+      "step": 2058
+    },
+    {
+      "epoch": 0.43820164937483375,
+      "grad_norm": 1.3599191707597007,
+      "learning_rate": 9.935541331110549e-05,
+      "loss": 3.5897,
+      "step": 2059
+    },
+    {
+      "epoch": 0.438414471934025,
+      "grad_norm": 1.9909840684592053,
+      "learning_rate": 9.935342967665374e-05,
+      "loss": 3.5638,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4386272944932163,
+      "grad_norm": 1.552462712698525,
+      "learning_rate": 9.93514430145593e-05,
+      "loss": 3.5862,
+      "step": 2061
+    },
+    {
+      "epoch": 0.43884011705240755,
+      "grad_norm": 1.6578467879516339,
+      "learning_rate": 9.934945332494411e-05,
+      "loss": 3.6148,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4390529396115988,
+      "grad_norm": 1.6094112094708046,
+      "learning_rate": 9.934746060793018e-05,
+      "loss": 3.5851,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4392657621707901,
+      "grad_norm": 1.5215372741884345,
+      "learning_rate": 9.93454648636398e-05,
+      "loss": 3.5881,
+      "step": 2064
+    },
+    {
+      "epoch": 0.43947858472998136,
+      "grad_norm": 1.3739323783096484,
+      "learning_rate": 9.934346609219538e-05,
+      "loss": 3.5784,
+      "step": 2065
+    },
+    {
+      "epoch": 0.43969140728917266,
+      "grad_norm": 1.5782711713850188,
+      "learning_rate": 9.934146429371953e-05,
+      "loss": 3.5625,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4399042298483639,
+      "grad_norm": 1.5970855347048283,
+      "learning_rate": 9.933945946833506e-05,
+      "loss": 3.6153,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4401170524075552,
+      "grad_norm": 1.4198502226058132,
+      "learning_rate": 9.933745161616497e-05,
+      "loss": 3.5685,
+      "step": 2068
+    },
+    {
+      "epoch": 0.44032987496674647,
+      "grad_norm": 1.4322463072201428,
+      "learning_rate": 9.933544073733241e-05,
+      "loss": 3.5731,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4405426975259378,
+      "grad_norm": 1.5052579623861706,
+      "learning_rate": 9.933342683196075e-05,
+      "loss": 3.5742,
+      "step": 2070
+    },
+    {
+      "epoch": 0.440755520085129,
+      "grad_norm": 1.5317809414216608,
+      "learning_rate": 9.933140990017354e-05,
+      "loss": 3.5745,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4409683426443203,
+      "grad_norm": 1.5739156720142682,
+      "learning_rate": 9.932938994209448e-05,
+      "loss": 3.55,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4411811652035116,
+      "grad_norm": 1.2581138015616207,
+      "learning_rate": 9.932736695784755e-05,
+      "loss": 3.5685,
+      "step": 2073
+    },
+    {
+      "epoch": 0.44139398776270283,
+      "grad_norm": 1.7220031849656583,
+      "learning_rate": 9.932534094755681e-05,
+      "loss": 3.5365,
+      "step": 2074
+    },
+    {
+      "epoch": 0.44160681032189414,
+      "grad_norm": 1.4310081215970378,
+      "learning_rate": 9.932331191134653e-05,
+      "loss": 3.5636,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4418196328810854,
+      "grad_norm": 1.6726346571688107,
+      "learning_rate": 9.932127984934124e-05,
+      "loss": 3.5663,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4420324554402767,
+      "grad_norm": 1.1794156930530983,
+      "learning_rate": 9.931924476166557e-05,
+      "loss": 3.5187,
+      "step": 2077
+    },
+    {
+      "epoch": 0.44224527799946795,
+      "grad_norm": 1.6367078936448667,
+      "learning_rate": 9.931720664844433e-05,
+      "loss": 3.5276,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4424581005586592,
+      "grad_norm": 1.4716758333255193,
+      "learning_rate": 9.93151655098026e-05,
+      "loss": 3.5712,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4426709231178505,
+      "grad_norm": 1.6010076231072505,
+      "learning_rate": 9.931312134586558e-05,
+      "loss": 3.6036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.44288374567704175,
+      "grad_norm": 1.1454427921674506,
+      "learning_rate": 9.931107415675867e-05,
+      "loss": 3.5557,
+      "step": 2081
+    },
+    {
+      "epoch": 0.44309656823623306,
+      "grad_norm": 1.8546108159562895,
+      "learning_rate": 9.930902394260747e-05,
+      "loss": 3.6028,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4433093907954243,
+      "grad_norm": 1.5477469286489445,
+      "learning_rate": 9.93069707035377e-05,
+      "loss": 3.5756,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4435222133546156,
+      "grad_norm": 1.5653075251764414,
+      "learning_rate": 9.930491443967539e-05,
+      "loss": 3.5538,
+      "step": 2084
+    },
+    {
+      "epoch": 0.44373503591380686,
+      "grad_norm": 1.49571340140607,
+      "learning_rate": 9.930285515114664e-05,
+      "loss": 3.5858,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4439478584729981,
+      "grad_norm": 1.3770989793674298,
+      "learning_rate": 9.930079283807781e-05,
+      "loss": 3.5805,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4441606810321894,
+      "grad_norm": 1.6939342462968436,
+      "learning_rate": 9.929872750059537e-05,
+      "loss": 3.5822,
+      "step": 2087
+    },
+    {
+      "epoch": 0.44437350359138067,
+      "grad_norm": 1.455071108842946,
+      "learning_rate": 9.929665913882606e-05,
+      "loss": 3.6022,
+      "step": 2088
+    },
+    {
+      "epoch": 0.444586326150572,
+      "grad_norm": 1.5420133435230672,
+      "learning_rate": 9.929458775289676e-05,
+      "loss": 3.5686,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4447991487097632,
+      "grad_norm": 1.5920138545746938,
+      "learning_rate": 9.929251334293451e-05,
+      "loss": 3.5342,
+      "step": 2090
+    },
+    {
+      "epoch": 0.44501197126895453,
+      "grad_norm": 1.2391159838285968,
+      "learning_rate": 9.929043590906662e-05,
+      "loss": 3.5328,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4452247938281458,
+      "grad_norm": 1.6068385663681752,
+      "learning_rate": 9.928835545142048e-05,
+      "loss": 3.5832,
+      "step": 2092
+    },
+    {
+      "epoch": 0.44543761638733703,
+      "grad_norm": 1.390123093229988,
+      "learning_rate": 9.928627197012376e-05,
+      "loss": 3.5725,
+      "step": 2093
+    },
+    {
+      "epoch": 0.44565043894652834,
+      "grad_norm": 1.6402323880746397,
+      "learning_rate": 9.928418546530424e-05,
+      "loss": 3.5824,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4458632615057196,
+      "grad_norm": 1.3445272051814294,
+      "learning_rate": 9.928209593708995e-05,
+      "loss": 3.6003,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4460760840649109,
+      "grad_norm": 1.6675405835621828,
+      "learning_rate": 9.928000338560905e-05,
+      "loss": 3.5754,
+      "step": 2096
+    },
+    {
+      "epoch": 0.44628890662410214,
+      "grad_norm": 1.5401663959575518,
+      "learning_rate": 9.927790781098992e-05,
+      "loss": 3.5731,
+      "step": 2097
+    },
+    {
+      "epoch": 0.44650172918329345,
+      "grad_norm": 1.4223522169717806,
+      "learning_rate": 9.927580921336111e-05,
+      "loss": 3.5979,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4467145517424847,
+      "grad_norm": 1.4864556037742958,
+      "learning_rate": 9.927370759285137e-05,
+      "loss": 3.5798,
+      "step": 2099
+    },
+    {
+      "epoch": 0.44692737430167595,
+      "grad_norm": 1.1377855341622425,
+      "learning_rate": 9.927160294958964e-05,
+      "loss": 3.5213,
+      "step": 2100
+    },
+    {
+      "epoch": 0.44714019686086726,
+      "grad_norm": 1.7208341858383955,
+      "learning_rate": 9.9269495283705e-05,
+      "loss": 3.585,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4473530194200585,
+      "grad_norm": 1.2946008740690569,
+      "learning_rate": 9.926738459532675e-05,
+      "loss": 3.549,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4475658419792498,
+      "grad_norm": 2.0616275502112846,
+      "learning_rate": 9.926527088458438e-05,
+      "loss": 3.5756,
+      "step": 2103
+    },
+    {
+      "epoch": 0.44777866453844106,
+      "grad_norm": 1.0314180378715812,
+      "learning_rate": 9.926315415160758e-05,
+      "loss": 3.5887,
+      "step": 2104
+    },
+    {
+      "epoch": 0.44799148709763237,
+      "grad_norm": 1.9655833147120116,
+      "learning_rate": 9.926103439652617e-05,
+      "loss": 3.5559,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4482043096568236,
+      "grad_norm": 1.4005689280747742,
+      "learning_rate": 9.925891161947021e-05,
+      "loss": 3.5427,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4484171322160149,
+      "grad_norm": 1.4967061760437232,
+      "learning_rate": 9.925678582056992e-05,
+      "loss": 3.5407,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4486299547752062,
+      "grad_norm": 1.5312668540316237,
+      "learning_rate": 9.925465699995571e-05,
+      "loss": 3.5333,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4488427773343974,
+      "grad_norm": 1.4350629753759288,
+      "learning_rate": 9.925252515775816e-05,
+      "loss": 3.6163,
+      "step": 2109
+    },
+    {
+      "epoch": 0.44905559989358873,
+      "grad_norm": 1.4150619435301122,
+      "learning_rate": 9.925039029410807e-05,
+      "loss": 3.5931,
+      "step": 2110
+    },
+    {
+      "epoch": 0.44926842245278,
+      "grad_norm": 1.5981013517788656,
+      "learning_rate": 9.92482524091364e-05,
+      "loss": 3.4963,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4494812450119713,
+      "grad_norm": 1.3533692694904027,
+      "learning_rate": 9.92461115029743e-05,
+      "loss": 3.534,
+      "step": 2112
+    },
+    {
+      "epoch": 0.44969406757116254,
+      "grad_norm": 1.3515666736040755,
+      "learning_rate": 9.924396757575309e-05,
+      "loss": 3.5841,
+      "step": 2113
+    },
+    {
+      "epoch": 0.44990689013035384,
+      "grad_norm": 1.5622355169689053,
+      "learning_rate": 9.924182062760432e-05,
+      "loss": 3.538,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4501197126895451,
+      "grad_norm": 1.422297190882594,
+      "learning_rate": 9.923967065865969e-05,
+      "loss": 3.5684,
+      "step": 2115
+    },
+    {
+      "epoch": 0.45033253524873634,
+      "grad_norm": 1.5141552615171483,
+      "learning_rate": 9.923751766905107e-05,
+      "loss": 3.6017,
+      "step": 2116
+    },
+    {
+      "epoch": 0.45054535780792765,
+      "grad_norm": 1.4480949102606304,
+      "learning_rate": 9.923536165891056e-05,
+      "loss": 3.5058,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4507581803671189,
+      "grad_norm": 1.4105470431572353,
+      "learning_rate": 9.923320262837041e-05,
+      "loss": 3.5691,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4509710029263102,
+      "grad_norm": 1.4309885977304848,
+      "learning_rate": 9.923104057756309e-05,
+      "loss": 3.5244,
+      "step": 2119
+    },
+    {
+      "epoch": 0.45118382548550146,
+      "grad_norm": 1.4794128634733723,
+      "learning_rate": 9.922887550662123e-05,
+      "loss": 3.5592,
+      "step": 2120
+    },
+    {
+      "epoch": 0.45139664804469276,
+      "grad_norm": 1.5297138055459476,
+      "learning_rate": 9.92267074156776e-05,
+      "loss": 3.5552,
+      "step": 2121
+    },
+    {
+      "epoch": 0.451609470603884,
+      "grad_norm": 1.5527372562534003,
+      "learning_rate": 9.922453630486527e-05,
+      "loss": 3.5626,
+      "step": 2122
+    },
+    {
+      "epoch": 0.45182229316307526,
+      "grad_norm": 1.3975025843143296,
+      "learning_rate": 9.92223621743174e-05,
+      "loss": 3.5323,
+      "step": 2123
+    },
+    {
+      "epoch": 0.45203511572226657,
+      "grad_norm": 1.3401775473808495,
+      "learning_rate": 9.922018502416736e-05,
+      "loss": 3.5675,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4522479382814578,
+      "grad_norm": 1.7243107956919674,
+      "learning_rate": 9.921800485454871e-05,
+      "loss": 3.5656,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4524607608406491,
+      "grad_norm": 1.052492010305671,
+      "learning_rate": 9.92158216655952e-05,
+      "loss": 3.5448,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4526735833998404,
+      "grad_norm": 2.016637921890496,
+      "learning_rate": 9.921363545744075e-05,
+      "loss": 3.519,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4528864059590317,
+      "grad_norm": 1.1366000590927086,
+      "learning_rate": 9.921144623021952e-05,
+      "loss": 3.527,
+      "step": 2128
+    },
+    {
+      "epoch": 0.45309922851822293,
+      "grad_norm": 1.7708855512503887,
+      "learning_rate": 9.920925398406576e-05,
+      "loss": 3.563,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4533120510774142,
+      "grad_norm": 1.3809475249178051,
+      "learning_rate": 9.920705871911395e-05,
+      "loss": 3.5425,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4535248736366055,
+      "grad_norm": 1.51862057207627,
+      "learning_rate": 9.92048604354988e-05,
+      "loss": 3.5201,
+      "step": 2131
+    },
+    {
+      "epoch": 0.45373769619579674,
+      "grad_norm": 1.600458637488564,
+      "learning_rate": 9.920265913335516e-05,
+      "loss": 3.5376,
+      "step": 2132
+    },
+    {
+      "epoch": 0.45395051875498804,
+      "grad_norm": 1.4613960558927672,
+      "learning_rate": 9.920045481281803e-05,
+      "loss": 3.4765,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4541633413141793,
+      "grad_norm": 1.2411085216186681,
+      "learning_rate": 9.91982474740227e-05,
+      "loss": 3.5112,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4543761638733706,
+      "grad_norm": 1.946529834016049,
+      "learning_rate": 9.919603711710453e-05,
+      "loss": 3.5693,
+      "step": 2135
+    },
+    {
+      "epoch": 0.45458898643256185,
+      "grad_norm": 1.139813530422803,
+      "learning_rate": 9.919382374219914e-05,
+      "loss": 3.5762,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4548018089917531,
+      "grad_norm": 1.728485802839492,
+      "learning_rate": 9.919160734944231e-05,
+      "loss": 3.4984,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4550146315509444,
+      "grad_norm": 1.3644685674575892,
+      "learning_rate": 9.918938793897001e-05,
+      "loss": 3.5656,
+      "step": 2138
+    },
+    {
+      "epoch": 0.45522745411013565,
+      "grad_norm": 1.2656409411966387,
+      "learning_rate": 9.918716551091838e-05,
+      "loss": 3.5106,
+      "step": 2139
+    },
+    {
+      "epoch": 0.45544027666932696,
+      "grad_norm": 1.7254607645673194,
+      "learning_rate": 9.918494006542378e-05,
+      "loss": 3.5182,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4556530992285182,
+      "grad_norm": 1.2026683738070685,
+      "learning_rate": 9.91827116026227e-05,
+      "loss": 3.5908,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4558659217877095,
+      "grad_norm": 1.7400582902763753,
+      "learning_rate": 9.918048012265186e-05,
+      "loss": 3.5426,
+      "step": 2142
+    },
+    {
+      "epoch": 0.45607874434690077,
+      "grad_norm": 1.3365804116827418,
+      "learning_rate": 9.917824562564818e-05,
+      "loss": 3.5482,
+      "step": 2143
+    },
+    {
+      "epoch": 0.45629156690609207,
+      "grad_norm": 1.499807281184078,
+      "learning_rate": 9.91760081117487e-05,
+      "loss": 3.5786,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4565043894652833,
+      "grad_norm": 1.5595899103592974,
+      "learning_rate": 9.91737675810907e-05,
+      "loss": 3.4768,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4567172120244746,
+      "grad_norm": 1.2921523344004398,
+      "learning_rate": 9.917152403381163e-05,
+      "loss": 3.5642,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4569300345836659,
+      "grad_norm": 1.4126143318832198,
+      "learning_rate": 9.916927747004912e-05,
+      "loss": 3.5533,
+      "step": 2147
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 1.5143746659840962,
+      "learning_rate": 9.916702788994097e-05,
+      "loss": 3.5644,
+      "step": 2148
+    },
+    {
+      "epoch": 0.45735567970204843,
+      "grad_norm": 1.4841094049799333,
+      "learning_rate": 9.916477529362523e-05,
+      "loss": 3.552,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4575685022612397,
+      "grad_norm": 1.2772734010830509,
+      "learning_rate": 9.916251968124004e-05,
+      "loss": 3.5333,
+      "step": 2150
+    },
+    {
+      "epoch": 0.457781324820431,
+      "grad_norm": 1.5490379710692772,
+      "learning_rate": 9.916026105292379e-05,
+      "loss": 3.5768,
+      "step": 2151
+    },
+    {
+      "epoch": 0.45799414737962224,
+      "grad_norm": 1.1708796718983918,
+      "learning_rate": 9.915799940881504e-05,
+      "loss": 3.5883,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4582069699388135,
+      "grad_norm": 1.668523914752012,
+      "learning_rate": 9.915573474905253e-05,
+      "loss": 3.5325,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4584197924980048,
+      "grad_norm": 1.4061954548417679,
+      "learning_rate": 9.91534670737752e-05,
+      "loss": 3.5279,
+      "step": 2154
+    },
+    {
+      "epoch": 0.45863261505719605,
+      "grad_norm": 1.4830873634290507,
+      "learning_rate": 9.915119638312215e-05,
+      "loss": 3.4659,
+      "step": 2155
+    },
+    {
+      "epoch": 0.45884543761638735,
+      "grad_norm": 1.172265927928014,
+      "learning_rate": 9.914892267723268e-05,
+      "loss": 3.4837,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4590582601755786,
+      "grad_norm": 1.6385738168174928,
+      "learning_rate": 9.914664595624627e-05,
+      "loss": 3.572,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4592710827347699,
+      "grad_norm": 1.459129374345066,
+      "learning_rate": 9.91443662203026e-05,
+      "loss": 3.5589,
+      "step": 2158
+    },
+    {
+      "epoch": 0.45948390529396116,
+      "grad_norm": 1.3042347135730117,
+      "learning_rate": 9.914208346954152e-05,
+      "loss": 3.4832,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4596967278531524,
+      "grad_norm": 1.4612176599696391,
+      "learning_rate": 9.913979770410306e-05,
+      "loss": 3.5351,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4599095504123437,
+      "grad_norm": 1.491833145083417,
+      "learning_rate": 9.913750892412746e-05,
+      "loss": 3.5439,
+      "step": 2161
+    },
+    {
+      "epoch": 0.46012237297153497,
+      "grad_norm": 1.4493856261070361,
+      "learning_rate": 9.913521712975511e-05,
+      "loss": 3.5321,
+      "step": 2162
+    },
+    {
+      "epoch": 0.46033519553072627,
+      "grad_norm": 1.7361200685043532,
+      "learning_rate": 9.91329223211266e-05,
+      "loss": 3.5644,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4605480180899175,
+      "grad_norm": 1.1427138023105252,
+      "learning_rate": 9.91306244983827e-05,
+      "loss": 3.5133,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4607608406491088,
+      "grad_norm": 1.7573893896165782,
+      "learning_rate": 9.912832366166442e-05,
+      "loss": 3.5239,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4609736632083001,
+      "grad_norm": 1.2629987416640327,
+      "learning_rate": 9.912601981111286e-05,
+      "loss": 3.5061,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4611864857674913,
+      "grad_norm": 1.663078811333668,
+      "learning_rate": 9.912371294686937e-05,
+      "loss": 3.4866,
+      "step": 2167
+    },
+    {
+      "epoch": 0.46139930832668263,
+      "grad_norm": 1.5704476302933625,
+      "learning_rate": 9.912140306907546e-05,
+      "loss": 3.4958,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4616121308858739,
+      "grad_norm": 1.437958163286061,
+      "learning_rate": 9.911909017787284e-05,
+      "loss": 3.5298,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4618249534450652,
+      "grad_norm": 1.454885428369392,
+      "learning_rate": 9.911677427340339e-05,
+      "loss": 3.5271,
+      "step": 2170
+    },
+    {
+      "epoch": 0.46203777600425644,
+      "grad_norm": 1.597143520941883,
+      "learning_rate": 9.911445535580918e-05,
+      "loss": 3.5112,
+      "step": 2171
+    },
+    {
+      "epoch": 0.46225059856344775,
+      "grad_norm": 1.2153024257479805,
+      "learning_rate": 9.911213342523249e-05,
+      "loss": 3.5027,
+      "step": 2172
+    },
+    {
+      "epoch": 0.462463421122639,
+      "grad_norm": 1.6431062271774264,
+      "learning_rate": 9.910980848181573e-05,
+      "loss": 3.5256,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4626762436818303,
+      "grad_norm": 1.389626346228638,
+      "learning_rate": 9.910748052570156e-05,
+      "loss": 3.5528,
+      "step": 2174
+    },
+    {
+      "epoch": 0.46288906624102155,
+      "grad_norm": 1.386909903525284,
+      "learning_rate": 9.910514955703275e-05,
+      "loss": 3.5427,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4631018888002128,
+      "grad_norm": 1.3327530832366214,
+      "learning_rate": 9.910281557595233e-05,
+      "loss": 3.5409,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4633147113594041,
+      "grad_norm": 1.3586388608990667,
+      "learning_rate": 9.910047858260346e-05,
+      "loss": 3.5044,
+      "step": 2177
+    },
+    {
+      "epoch": 0.46352753391859536,
+      "grad_norm": 1.6474008617855278,
+      "learning_rate": 9.909813857712951e-05,
+      "loss": 3.5884,
+      "step": 2178
+    },
+    {
+      "epoch": 0.46374035647778666,
+      "grad_norm": 1.1503580767217836,
+      "learning_rate": 9.909579555967406e-05,
+      "loss": 3.5251,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4639531790369779,
+      "grad_norm": 1.7205002981550022,
+      "learning_rate": 9.909344953038078e-05,
+      "loss": 3.4912,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4641660015961692,
+      "grad_norm": 1.224223840375713,
+      "learning_rate": 9.909110048939365e-05,
+      "loss": 3.5526,
+      "step": 2181
+    },
+    {
+      "epoch": 0.46437882415536047,
+      "grad_norm": 1.624234540684325,
+      "learning_rate": 9.908874843685676e-05,
+      "loss": 3.5001,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4645916467145517,
+      "grad_norm": 1.290219872998848,
+      "learning_rate": 9.908639337291438e-05,
+      "loss": 3.5227,
+      "step": 2183
+    },
+    {
+      "epoch": 0.464804469273743,
+      "grad_norm": 1.7169119677562932,
+      "learning_rate": 9.9084035297711e-05,
+      "loss": 3.4852,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4650172918329343,
+      "grad_norm": 1.4338217617761055,
+      "learning_rate": 9.908167421139129e-05,
+      "loss": 3.5177,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4652301143921256,
+      "grad_norm": 1.3293093509346523,
+      "learning_rate": 9.907931011410007e-05,
+      "loss": 3.5157,
+      "step": 2186
+    },
+    {
+      "epoch": 0.46544293695131683,
+      "grad_norm": 1.610437432538521,
+      "learning_rate": 9.907694300598237e-05,
+      "loss": 3.5296,
+      "step": 2187
+    },
+    {
+      "epoch": 0.46565575951050814,
+      "grad_norm": 1.329761051614342,
+      "learning_rate": 9.907457288718342e-05,
+      "loss": 3.5244,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4658685820696994,
+      "grad_norm": 1.5959758361030016,
+      "learning_rate": 9.907219975784861e-05,
+      "loss": 3.5029,
+      "step": 2189
+    },
+    {
+      "epoch": 0.46608140462889064,
+      "grad_norm": 1.3852079541699007,
+      "learning_rate": 9.90698236181235e-05,
+      "loss": 3.5259,
+      "step": 2190
+    },
+    {
+      "epoch": 0.46629422718808194,
+      "grad_norm": 1.5537991329805998,
+      "learning_rate": 9.90674444681539e-05,
+      "loss": 3.5528,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4665070497472732,
+      "grad_norm": 1.5988015965046738,
+      "learning_rate": 9.906506230808572e-05,
+      "loss": 3.5214,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4667198723064645,
+      "grad_norm": 1.3006595678528203,
+      "learning_rate": 9.906267713806514e-05,
+      "loss": 3.5138,
+      "step": 2193
+    },
+    {
+      "epoch": 0.46693269486565575,
+      "grad_norm": 1.658709311821262,
+      "learning_rate": 9.906028895823843e-05,
+      "loss": 3.5002,
+      "step": 2194
+    },
+    {
+      "epoch": 0.46714551742484706,
+      "grad_norm": 1.1306705106951145,
+      "learning_rate": 9.905789776875215e-05,
+      "loss": 3.4892,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4673583399840383,
+      "grad_norm": 1.623230084465071,
+      "learning_rate": 9.905550356975293e-05,
+      "loss": 3.4555,
+      "step": 2196
+    },
+    {
+      "epoch": 0.46757116254322956,
+      "grad_norm": 1.402842272142635,
+      "learning_rate": 9.90531063613877e-05,
+      "loss": 3.4732,
+      "step": 2197
+    },
+    {
+      "epoch": 0.46778398510242086,
+      "grad_norm": 1.3211084169674565,
+      "learning_rate": 9.90507061438035e-05,
+      "loss": 3.4831,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4679968076616121,
+      "grad_norm": 1.5465346205629644,
+      "learning_rate": 9.904830291714755e-05,
+      "loss": 3.502,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4682096302208034,
+      "grad_norm": 1.410498804625109,
+      "learning_rate": 9.904589668156732e-05,
+      "loss": 3.4643,
+      "step": 2200
+    },
+    {
+      "epoch": 0.46842245277999467,
+      "grad_norm": 1.6442855862492634,
+      "learning_rate": 9.90434874372104e-05,
+      "loss": 3.4595,
+      "step": 2201
+    },
+    {
+      "epoch": 0.468635275339186,
+      "grad_norm": 1.2654446061814026,
+      "learning_rate": 9.904107518422458e-05,
+      "loss": 3.5158,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4688480978983772,
+      "grad_norm": 1.750401461792842,
+      "learning_rate": 9.903865992275786e-05,
+      "loss": 3.5547,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4690609204575685,
+      "grad_norm": 1.041087903743483,
+      "learning_rate": 9.903624165295839e-05,
+      "loss": 3.4825,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4692737430167598,
+      "grad_norm": 1.9032070510216155,
+      "learning_rate": 9.903382037497455e-05,
+      "loss": 3.5484,
+      "step": 2205
+    },
+    {
+      "epoch": 0.46948656557595103,
+      "grad_norm": 1.3120681386453816,
+      "learning_rate": 9.903139608895486e-05,
+      "loss": 3.4777,
+      "step": 2206
+    },
+    {
+      "epoch": 0.46969938813514234,
+      "grad_norm": 1.645811493601363,
+      "learning_rate": 9.902896879504801e-05,
+      "loss": 3.5288,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4699122106943336,
+      "grad_norm": 1.3651818141905978,
+      "learning_rate": 9.902653849340295e-05,
+      "loss": 3.4914,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4701250332535249,
+      "grad_norm": 1.4136947830582516,
+      "learning_rate": 9.902410518416875e-05,
+      "loss": 3.5077,
+      "step": 2209
+    },
+    {
+      "epoch": 0.47033785581271614,
+      "grad_norm": 1.5157463954427894,
+      "learning_rate": 9.90216688674947e-05,
+      "loss": 3.55,
+      "step": 2210
+    },
+    {
+      "epoch": 0.47055067837190745,
+      "grad_norm": 1.3573981207354713,
+      "learning_rate": 9.901922954353024e-05,
+      "loss": 3.5051,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4707635009310987,
+      "grad_norm": 1.650294599737523,
+      "learning_rate": 9.901678721242502e-05,
+      "loss": 3.459,
+      "step": 2212
+    },
+    {
+      "epoch": 0.47097632349028995,
+      "grad_norm": 1.173211176314109,
+      "learning_rate": 9.901434187432885e-05,
+      "loss": 3.4775,
+      "step": 2213
+    },
+    {
+      "epoch": 0.47118914604948126,
+      "grad_norm": 1.6979844625398113,
+      "learning_rate": 9.901189352939177e-05,
+      "loss": 3.5033,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4714019686086725,
+      "grad_norm": 1.436195304013237,
+      "learning_rate": 9.900944217776398e-05,
+      "loss": 3.494,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4716147911678638,
+      "grad_norm": 1.3973768118361627,
+      "learning_rate": 9.900698781959584e-05,
+      "loss": 3.4963,
+      "step": 2216
+    },
+    {
+      "epoch": 0.47182761372705506,
+      "grad_norm": 1.5233664954013513,
+      "learning_rate": 9.90045304550379e-05,
+      "loss": 3.4782,
+      "step": 2217
+    },
+    {
+      "epoch": 0.47204043628624637,
+      "grad_norm": 1.3781756497071995,
+      "learning_rate": 9.900207008424094e-05,
+      "loss": 3.5237,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4722532588454376,
+      "grad_norm": 1.5500677841441521,
+      "learning_rate": 9.89996067073559e-05,
+      "loss": 3.4868,
+      "step": 2219
+    },
+    {
+      "epoch": 0.47246608140462887,
+      "grad_norm": 1.1751765875526496,
+      "learning_rate": 9.899714032453387e-05,
+      "loss": 3.4977,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4726789039638202,
+      "grad_norm": 1.542185045866411,
+      "learning_rate": 9.899467093592617e-05,
+      "loss": 3.5372,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4728917265230114,
+      "grad_norm": 1.3613486653443005,
+      "learning_rate": 9.899219854168429e-05,
+      "loss": 3.4635,
+      "step": 2222
+    },
+    {
+      "epoch": 0.47310454908220273,
+      "grad_norm": 1.4438189642469614,
+      "learning_rate": 9.898972314195988e-05,
+      "loss": 3.5295,
+      "step": 2223
+    },
+    {
+      "epoch": 0.473317371641394,
+      "grad_norm": 1.2601373735947272,
+      "learning_rate": 9.898724473690483e-05,
+      "loss": 3.4692,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4735301942005853,
+      "grad_norm": 1.5550229411741507,
+      "learning_rate": 9.898476332667117e-05,
+      "loss": 3.518,
+      "step": 2225
+    },
+    {
+      "epoch": 0.47374301675977654,
+      "grad_norm": 1.3995988543570403,
+      "learning_rate": 9.89822789114111e-05,
+      "loss": 3.4972,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4739558393189678,
+      "grad_norm": 1.410213620822416,
+      "learning_rate": 9.897979149127708e-05,
+      "loss": 3.5151,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4741686618781591,
+      "grad_norm": 1.5969112804425825,
+      "learning_rate": 9.897730106642164e-05,
+      "loss": 3.4996,
+      "step": 2228
+    },
+    {
+      "epoch": 0.47438148443735034,
+      "grad_norm": 1.2572716858891702,
+      "learning_rate": 9.89748076369976e-05,
+      "loss": 3.4895,
+      "step": 2229
+    },
+    {
+      "epoch": 0.47459430699654165,
+      "grad_norm": 1.6849539735777548,
+      "learning_rate": 9.89723112031579e-05,
+      "loss": 3.4965,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4748071295557329,
+      "grad_norm": 1.1697119352037433,
+      "learning_rate": 9.896981176505571e-05,
+      "loss": 3.5042,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4750199521149242,
+      "grad_norm": 2.084961463815915,
+      "learning_rate": 9.896730932284435e-05,
+      "loss": 3.4924,
+      "step": 2232
+    },
+    {
+      "epoch": 0.47523277467411545,
+      "grad_norm": 0.9931669942483252,
+      "learning_rate": 9.896480387667732e-05,
+      "loss": 3.5061,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4754455972333067,
+      "grad_norm": 2.2119957722226804,
+      "learning_rate": 9.896229542670836e-05,
+      "loss": 3.526,
+      "step": 2234
+    },
+    {
+      "epoch": 0.475658419792498,
+      "grad_norm": 1.4305900126267883,
+      "learning_rate": 9.895978397309132e-05,
+      "loss": 3.4713,
+      "step": 2235
+    },
+    {
+      "epoch": 0.47587124235168926,
+      "grad_norm": 1.4375721451442487,
+      "learning_rate": 9.895726951598026e-05,
+      "loss": 3.4656,
+      "step": 2236
+    },
+    {
+      "epoch": 0.47608406491088057,
+      "grad_norm": 1.6933973558648934,
+      "learning_rate": 9.895475205552944e-05,
+      "loss": 3.4897,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4762968874700718,
+      "grad_norm": 1.1453708165630467,
+      "learning_rate": 9.895223159189332e-05,
+      "loss": 3.4941,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4765097100292631,
+      "grad_norm": 1.561857457307003,
+      "learning_rate": 9.894970812522651e-05,
+      "loss": 3.4956,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4767225325884544,
+      "grad_norm": 1.2779827054290291,
+      "learning_rate": 9.894718165568382e-05,
+      "loss": 3.5329,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4769353551476456,
+      "grad_norm": 1.3930458745233003,
+      "learning_rate": 9.89446521834202e-05,
+      "loss": 3.5032,
+      "step": 2241
+    },
+    {
+      "epoch": 0.47714817770683693,
+      "grad_norm": 1.4230699591787508,
+      "learning_rate": 9.894211970859087e-05,
+      "loss": 3.4988,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4773610002660282,
+      "grad_norm": 1.4839020850281261,
+      "learning_rate": 9.893958423135117e-05,
+      "loss": 3.506,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4775738228252195,
+      "grad_norm": 1.2327079270960026,
+      "learning_rate": 9.893704575185664e-05,
+      "loss": 3.5197,
+      "step": 2244
+    },
+    {
+      "epoch": 0.47778664538441074,
+      "grad_norm": 1.5149636258923271,
+      "learning_rate": 9.893450427026301e-05,
+      "loss": 3.5014,
+      "step": 2245
+    },
+    {
+      "epoch": 0.47799946794360204,
+      "grad_norm": 1.6055186422947803,
+      "learning_rate": 9.893195978672618e-05,
+      "loss": 3.5024,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4782122905027933,
+      "grad_norm": 1.0668735713870299,
+      "learning_rate": 9.892941230140226e-05,
+      "loss": 3.4766,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4784251130619846,
+      "grad_norm": 1.4213777064384745,
+      "learning_rate": 9.892686181444752e-05,
+      "loss": 3.5182,
+      "step": 2248
+    },
+    {
+      "epoch": 0.47863793562117585,
+      "grad_norm": 1.5018966772829745,
+      "learning_rate": 9.892430832601842e-05,
+      "loss": 3.4773,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4788507581803671,
+      "grad_norm": 1.4119563377293094,
+      "learning_rate": 9.892175183627161e-05,
+      "loss": 3.4722,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4790635807395584,
+      "grad_norm": 1.2591895031007152,
+      "learning_rate": 9.891919234536391e-05,
+      "loss": 3.5247,
+      "step": 2251
+    },
+    {
+      "epoch": 0.47927640329874965,
+      "grad_norm": 1.4657163349688525,
+      "learning_rate": 9.891662985345235e-05,
+      "loss": 3.5339,
+      "step": 2252
+    },
+    {
+      "epoch": 0.47948922585794096,
+      "grad_norm": 1.267488846247363,
+      "learning_rate": 9.891406436069415e-05,
+      "loss": 3.5241,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4797020484171322,
+      "grad_norm": 1.6617994055845435,
+      "learning_rate": 9.891149586724664e-05,
+      "loss": 3.4839,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4799148709763235,
+      "grad_norm": 1.1889741017447135,
+      "learning_rate": 9.890892437326742e-05,
+      "loss": 3.4584,
+      "step": 2255
+    },
+    {
+      "epoch": 0.48012769353551477,
+      "grad_norm": 1.5353130032376425,
+      "learning_rate": 9.890634987891424e-05,
+      "loss": 3.4895,
+      "step": 2256
+    },
+    {
+      "epoch": 0.480340516094706,
+      "grad_norm": 1.3195436667837388,
+      "learning_rate": 9.890377238434503e-05,
+      "loss": 3.5265,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4805533386538973,
+      "grad_norm": 1.5887444870361198,
+      "learning_rate": 9.890119188971791e-05,
+      "loss": 3.4872,
+      "step": 2258
+    },
+    {
+      "epoch": 0.48076616121308857,
+      "grad_norm": 1.3161848124360696,
+      "learning_rate": 9.889860839519119e-05,
+      "loss": 3.5209,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4809789837722799,
+      "grad_norm": 1.7323791397628734,
+      "learning_rate": 9.889602190092335e-05,
+      "loss": 3.4633,
+      "step": 2260
+    },
+    {
+      "epoch": 0.48119180633147113,
+      "grad_norm": 1.2949616582775418,
+      "learning_rate": 9.889343240707306e-05,
+      "loss": 3.4579,
+      "step": 2261
+    },
+    {
+      "epoch": 0.48140462889066243,
+      "grad_norm": 1.3949816597044886,
+      "learning_rate": 9.889083991379917e-05,
+      "loss": 3.464,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4816174514498537,
+      "grad_norm": 1.3339577149326662,
+      "learning_rate": 9.888824442126073e-05,
+      "loss": 3.4108,
+      "step": 2263
+    },
+    {
+      "epoch": 0.48183027400904493,
+      "grad_norm": 1.3336225147407736,
+      "learning_rate": 9.888564592961698e-05,
+      "loss": 3.4713,
+      "step": 2264
+    },
+    {
+      "epoch": 0.48204309656823624,
+      "grad_norm": 1.5068421176618434,
+      "learning_rate": 9.888304443902728e-05,
+      "loss": 3.4876,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4822559191274275,
+      "grad_norm": 1.3020004792551547,
+      "learning_rate": 9.888043994965127e-05,
+      "loss": 3.4559,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4824687416866188,
+      "grad_norm": 1.4809715517185205,
+      "learning_rate": 9.88778324616487e-05,
+      "loss": 3.4741,
+      "step": 2267
+    },
+    {
+      "epoch": 0.48268156424581005,
+      "grad_norm": 1.1668972209475534,
+      "learning_rate": 9.887522197517953e-05,
+      "loss": 3.4638,
+      "step": 2268
+    },
+    {
+      "epoch": 0.48289438680500135,
+      "grad_norm": 1.4265335194888333,
+      "learning_rate": 9.887260849040392e-05,
+      "loss": 3.4848,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4831072093641926,
+      "grad_norm": 1.1972215591776953,
+      "learning_rate": 9.886999200748219e-05,
+      "loss": 3.4619,
+      "step": 2270
+    },
+    {
+      "epoch": 0.48332003192338385,
+      "grad_norm": 1.5609173550287385,
+      "learning_rate": 9.886737252657483e-05,
+      "loss": 3.4198,
+      "step": 2271
+    },
+    {
+      "epoch": 0.48353285448257516,
+      "grad_norm": 1.2624031091446617,
+      "learning_rate": 9.886475004784254e-05,
+      "loss": 3.4942,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4837456770417664,
+      "grad_norm": 1.714207524208865,
+      "learning_rate": 9.886212457144623e-05,
+      "loss": 3.497,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4839584996009577,
+      "grad_norm": 1.1547520868510024,
+      "learning_rate": 9.885949609754695e-05,
+      "loss": 3.4977,
+      "step": 2274
+    },
+    {
+      "epoch": 0.48417132216014896,
+      "grad_norm": 1.4522457467003578,
+      "learning_rate": 9.885686462630594e-05,
+      "loss": 3.4645,
+      "step": 2275
+    },
+    {
+      "epoch": 0.48438414471934027,
+      "grad_norm": 1.2241379865211595,
+      "learning_rate": 9.885423015788463e-05,
+      "loss": 3.4987,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4845969672785315,
+      "grad_norm": 1.5245620152976451,
+      "learning_rate": 9.885159269244462e-05,
+      "loss": 3.4546,
+      "step": 2277
+    },
+    {
+      "epoch": 0.48480978983772277,
+      "grad_norm": 1.190829776337086,
+      "learning_rate": 9.884895223014772e-05,
+      "loss": 3.4495,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4850226123969141,
+      "grad_norm": 1.6884283923433017,
+      "learning_rate": 9.884630877115592e-05,
+      "loss": 3.5189,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4852354349561053,
+      "grad_norm": 1.4303126952199472,
+      "learning_rate": 9.88436623156314e-05,
+      "loss": 3.4303,
+      "step": 2280
+    },
+    {
+      "epoch": 0.48544825751529663,
+      "grad_norm": 1.4636434854610252,
+      "learning_rate": 9.884101286373648e-05,
+      "loss": 3.4491,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4856610800744879,
+      "grad_norm": 1.413404172523139,
+      "learning_rate": 9.88383604156337e-05,
+      "loss": 3.5005,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4858739026336792,
+      "grad_norm": 1.3011349725406,
+      "learning_rate": 9.883570497148578e-05,
+      "loss": 3.4533,
+      "step": 2283
+    },
+    {
+      "epoch": 0.48608672519287044,
+      "grad_norm": 1.532012438042983,
+      "learning_rate": 9.883304653145564e-05,
+      "loss": 3.4727,
+      "step": 2284
+    },
+    {
+      "epoch": 0.48629954775206174,
+      "grad_norm": 1.3205197831795208,
+      "learning_rate": 9.883038509570634e-05,
+      "loss": 3.4511,
+      "step": 2285
+    },
+    {
+      "epoch": 0.486512370311253,
+      "grad_norm": 1.51837885604267,
+      "learning_rate": 9.882772066440116e-05,
+      "loss": 3.4477,
+      "step": 2286
+    },
+    {
+      "epoch": 0.48672519287044425,
+      "grad_norm": 1.641904944728538,
+      "learning_rate": 9.882505323770354e-05,
+      "loss": 3.484,
+      "step": 2287
+    },
+    {
+      "epoch": 0.48693801542963555,
+      "grad_norm": 1.1627210476897698,
+      "learning_rate": 9.882238281577715e-05,
+      "loss": 3.4642,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4871508379888268,
+      "grad_norm": 1.7182864582021387,
+      "learning_rate": 9.881970939878577e-05,
+      "loss": 3.463,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4873636605480181,
+      "grad_norm": 1.1205321111232456,
+      "learning_rate": 9.881703298689342e-05,
+      "loss": 3.4554,
+      "step": 2290
+    },
+    {
+      "epoch": 0.48757648310720936,
+      "grad_norm": 1.7243239423987518,
+      "learning_rate": 9.881435358026429e-05,
+      "loss": 3.48,
+      "step": 2291
+    },
+    {
+      "epoch": 0.48778930566640066,
+      "grad_norm": 1.2539867632795485,
+      "learning_rate": 9.881167117906275e-05,
+      "loss": 3.446,
+      "step": 2292
+    },
+    {
+      "epoch": 0.4880021282255919,
+      "grad_norm": 1.6968612435179429,
+      "learning_rate": 9.880898578345337e-05,
+      "loss": 3.5551,
+      "step": 2293
+    },
+    {
+      "epoch": 0.48821495078478316,
+      "grad_norm": 1.191648685336079,
+      "learning_rate": 9.880629739360085e-05,
+      "loss": 3.4824,
+      "step": 2294
+    },
+    {
+      "epoch": 0.48842777334397447,
+      "grad_norm": 1.6026599405216098,
+      "learning_rate": 9.880360600967013e-05,
+      "loss": 3.5142,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4886405959031657,
+      "grad_norm": 1.1735003133743724,
+      "learning_rate": 9.880091163182635e-05,
+      "loss": 3.4787,
+      "step": 2296
+    },
+    {
+      "epoch": 0.488853418462357,
+      "grad_norm": 1.5767575171435997,
+      "learning_rate": 9.879821426023476e-05,
+      "loss": 3.4589,
+      "step": 2297
+    },
+    {
+      "epoch": 0.4890662410215483,
+      "grad_norm": 1.299235189458737,
+      "learning_rate": 9.879551389506083e-05,
+      "loss": 3.4152,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4892790635807396,
+      "grad_norm": 1.3598125449751792,
+      "learning_rate": 9.879281053647024e-05,
+      "loss": 3.467,
+      "step": 2299
+    },
+    {
+      "epoch": 0.48949188613993083,
+      "grad_norm": 1.4470238028049107,
+      "learning_rate": 9.879010418462884e-05,
+      "loss": 3.499,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4897047086991221,
+      "grad_norm": 1.3670519196447966,
+      "learning_rate": 9.878739483970261e-05,
+      "loss": 3.4312,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4899175312583134,
+      "grad_norm": 1.4477622493073226,
+      "learning_rate": 9.87846825018578e-05,
+      "loss": 3.4433,
+      "step": 2302
+    },
+    {
+      "epoch": 0.49013035381750464,
+      "grad_norm": 1.3032105510211927,
+      "learning_rate": 9.878196717126078e-05,
+      "loss": 3.4725,
+      "step": 2303
+    },
+    {
+      "epoch": 0.49034317637669594,
+      "grad_norm": 1.344261737214939,
+      "learning_rate": 9.877924884807814e-05,
+      "loss": 3.482,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4905559989358872,
+      "grad_norm": 1.294993807561026,
+      "learning_rate": 9.877652753247661e-05,
+      "loss": 3.4611,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4907688214950785,
+      "grad_norm": 1.4636002394414207,
+      "learning_rate": 9.877380322462318e-05,
+      "loss": 3.5168,
+      "step": 2306
+    },
+    {
+      "epoch": 0.49098164405426975,
+      "grad_norm": 1.2405934425810328,
+      "learning_rate": 9.877107592468494e-05,
+      "loss": 3.4439,
+      "step": 2307
+    },
+    {
+      "epoch": 0.491194466613461,
+      "grad_norm": 1.7938516813109702,
+      "learning_rate": 9.876834563282919e-05,
+      "loss": 3.4678,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4914072891726523,
+      "grad_norm": 1.154675879546492,
+      "learning_rate": 9.876561234922345e-05,
+      "loss": 3.4697,
+      "step": 2309
+    },
+    {
+      "epoch": 0.49162011173184356,
+      "grad_norm": 1.5868196999046984,
+      "learning_rate": 9.876287607403538e-05,
+      "loss": 3.4472,
+      "step": 2310
+    },
+    {
+      "epoch": 0.49183293429103486,
+      "grad_norm": 1.2472679076615234,
+      "learning_rate": 9.876013680743285e-05,
+      "loss": 3.4547,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4920457568502261,
+      "grad_norm": 1.6083209486502539,
+      "learning_rate": 9.87573945495839e-05,
+      "loss": 3.4784,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4922585794094174,
+      "grad_norm": 1.3268907311173506,
+      "learning_rate": 9.875464930065675e-05,
+      "loss": 3.4585,
+      "step": 2313
+    },
+    {
+      "epoch": 0.49247140196860867,
+      "grad_norm": 1.4094320256438793,
+      "learning_rate": 9.875190106081983e-05,
+      "loss": 3.4338,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4926842245278,
+      "grad_norm": 1.2516256035237883,
+      "learning_rate": 9.87491498302417e-05,
+      "loss": 3.4521,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4928970470869912,
+      "grad_norm": 1.5149459733045971,
+      "learning_rate": 9.874639560909117e-05,
+      "loss": 3.4587,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4931098696461825,
+      "grad_norm": 1.2091354820742484,
+      "learning_rate": 9.874363839753721e-05,
+      "loss": 3.4853,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4933226922053738,
+      "grad_norm": 1.3888162473319148,
+      "learning_rate": 9.874087819574892e-05,
+      "loss": 3.418,
+      "step": 2318
+    },
+    {
+      "epoch": 0.49353551476456503,
+      "grad_norm": 1.3932767725206394,
+      "learning_rate": 9.873811500389566e-05,
+      "loss": 3.4509,
+      "step": 2319
+    },
+    {
+      "epoch": 0.49374833732375634,
+      "grad_norm": 1.18982463342346,
+      "learning_rate": 9.873534882214692e-05,
+      "loss": 3.4718,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4939611598829476,
+      "grad_norm": 1.6435092648348224,
+      "learning_rate": 9.873257965067242e-05,
+      "loss": 3.4518,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4941739824421389,
+      "grad_norm": 1.0227571449015767,
+      "learning_rate": 9.872980748964203e-05,
+      "loss": 3.4193,
+      "step": 2322
+    },
+    {
+      "epoch": 0.49438680500133014,
+      "grad_norm": 1.6423004116119695,
+      "learning_rate": 9.872703233922578e-05,
+      "loss": 3.4681,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4945996275605214,
+      "grad_norm": 1.3196292786981363,
+      "learning_rate": 9.872425419959396e-05,
+      "loss": 3.4571,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4948124501197127,
+      "grad_norm": 1.337122680793686,
+      "learning_rate": 9.872147307091699e-05,
+      "loss": 3.468,
+      "step": 2325
+    },
+    {
+      "epoch": 0.49502527267890395,
+      "grad_norm": 1.3895836728499142,
+      "learning_rate": 9.871868895336544e-05,
+      "loss": 3.4361,
+      "step": 2326
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 1.407339677424214,
+      "learning_rate": 9.871590184711015e-05,
+      "loss": 3.4851,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4954509177972865,
+      "grad_norm": 1.41007039053998,
+      "learning_rate": 9.871311175232209e-05,
+      "loss": 3.4388,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4956637403564778,
+      "grad_norm": 1.3708238273558078,
+      "learning_rate": 9.871031866917243e-05,
+      "loss": 3.4555,
+      "step": 2329
+    },
+    {
+      "epoch": 0.49587656291566906,
+      "grad_norm": 1.231271639087127,
+      "learning_rate": 9.870752259783247e-05,
+      "loss": 3.4626,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4960893854748603,
+      "grad_norm": 1.4029115092909639,
+      "learning_rate": 9.870472353847377e-05,
+      "loss": 3.4244,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4963022080340516,
+      "grad_norm": 1.3816319723106627,
+      "learning_rate": 9.870192149126807e-05,
+      "loss": 3.4245,
+      "step": 2332
+    },
+    {
+      "epoch": 0.49651503059324287,
+      "grad_norm": 1.4464826292493427,
+      "learning_rate": 9.86991164563872e-05,
+      "loss": 3.4827,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4967278531524342,
+      "grad_norm": 1.17381499569248,
+      "learning_rate": 9.86963084340033e-05,
+      "loss": 3.4497,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4969406757116254,
+      "grad_norm": 1.6763585796548683,
+      "learning_rate": 9.869349742428859e-05,
+      "loss": 3.4795,
+      "step": 2335
+    },
+    {
+      "epoch": 0.49715349827081673,
+      "grad_norm": 1.2781455979100163,
+      "learning_rate": 9.869068342741554e-05,
+      "loss": 3.4179,
+      "step": 2336
+    },
+    {
+      "epoch": 0.497366320830008,
+      "grad_norm": 1.713712666318683,
+      "learning_rate": 9.868786644355674e-05,
+      "loss": 3.4813,
+      "step": 2337
+    },
+    {
+      "epoch": 0.49757914338919923,
+      "grad_norm": 1.213844362402591,
+      "learning_rate": 9.868504647288505e-05,
+      "loss": 3.4513,
+      "step": 2338
+    },
+    {
+      "epoch": 0.49779196594839054,
+      "grad_norm": 1.456526269659767,
+      "learning_rate": 9.868222351557343e-05,
+      "loss": 3.4633,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4980047885075818,
+      "grad_norm": 1.2736619665482471,
+      "learning_rate": 9.867939757179509e-05,
+      "loss": 3.4894,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4982176110667731,
+      "grad_norm": 1.439220502185124,
+      "learning_rate": 9.867656864172335e-05,
+      "loss": 3.4712,
+      "step": 2341
+    },
+    {
+      "epoch": 0.49843043362596434,
+      "grad_norm": 1.497949773590217,
+      "learning_rate": 9.867373672553179e-05,
+      "loss": 3.4373,
+      "step": 2342
+    },
+    {
+      "epoch": 0.49864325618515565,
+      "grad_norm": 1.280903041207505,
+      "learning_rate": 9.867090182339411e-05,
+      "loss": 3.4224,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4988560787443469,
+      "grad_norm": 1.3922744618918028,
+      "learning_rate": 9.866806393548424e-05,
+      "loss": 3.4222,
+      "step": 2344
+    },
+    {
+      "epoch": 0.49906890130353815,
+      "grad_norm": 1.4332703897203134,
+      "learning_rate": 9.866522306197626e-05,
+      "loss": 3.4267,
+      "step": 2345
+    },
+    {
+      "epoch": 0.49928172386272945,
+      "grad_norm": 1.5244203088241544,
+      "learning_rate": 9.866237920304444e-05,
+      "loss": 3.4408,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4994945464219207,
+      "grad_norm": 1.1721918116610495,
+      "learning_rate": 9.865953235886327e-05,
+      "loss": 3.4888,
+      "step": 2347
+    },
+    {
+      "epoch": 0.499707368981112,
+      "grad_norm": 1.6070350731477332,
+      "learning_rate": 9.865668252960737e-05,
+      "loss": 3.4529,
+      "step": 2348
+    },
+    {
+      "epoch": 0.49992019154030326,
+      "grad_norm": 1.2197966691846769,
+      "learning_rate": 9.865382971545156e-05,
+      "loss": 3.427,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5001330140994945,
+      "grad_norm": 1.4573917716182918,
+      "learning_rate": 9.865097391657088e-05,
+      "loss": 3.4286,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5003458366586858,
+      "grad_norm": 1.381458325365222,
+      "learning_rate": 9.86481151331405e-05,
+      "loss": 3.444,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5005586592178771,
+      "grad_norm": 1.3907110532066624,
+      "learning_rate": 9.864525336533578e-05,
+      "loss": 3.4486,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5007714817770683,
+      "grad_norm": 1.2882607816650173,
+      "learning_rate": 9.86423886133323e-05,
+      "loss": 3.4692,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5009843043362596,
+      "grad_norm": 1.4498096562529945,
+      "learning_rate": 9.863952087730581e-05,
+      "loss": 3.4229,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5011971268954509,
+      "grad_norm": 1.3509768725193922,
+      "learning_rate": 9.863665015743221e-05,
+      "loss": 3.4513,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5014099494546422,
+      "grad_norm": 1.2792583163458309,
+      "learning_rate": 9.863377645388762e-05,
+      "loss": 3.4841,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5016227720138334,
+      "grad_norm": 1.6891527459919151,
+      "learning_rate": 9.863089976684834e-05,
+      "loss": 3.4671,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5018355945730247,
+      "grad_norm": 1.1694539130512303,
+      "learning_rate": 9.862802009649081e-05,
+      "loss": 3.407,
+      "step": 2358
+    },
+    {
+      "epoch": 0.502048417132216,
+      "grad_norm": 1.5051816161446714,
+      "learning_rate": 9.862513744299173e-05,
+      "loss": 3.4525,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5022612396914072,
+      "grad_norm": 1.1866189943767789,
+      "learning_rate": 9.862225180652793e-05,
+      "loss": 3.4564,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5024740622505985,
+      "grad_norm": 1.7514549527534002,
+      "learning_rate": 9.861936318727641e-05,
+      "loss": 3.4331,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5026868848097898,
+      "grad_norm": 1.1601715344580035,
+      "learning_rate": 9.861647158541439e-05,
+      "loss": 3.447,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5028997073689812,
+      "grad_norm": 1.6592482129459756,
+      "learning_rate": 9.861357700111925e-05,
+      "loss": 3.4353,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5031125299281723,
+      "grad_norm": 1.1356327351108828,
+      "learning_rate": 9.861067943456858e-05,
+      "loss": 3.4019,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5033253524873637,
+      "grad_norm": 1.3866919160102433,
+      "learning_rate": 9.86077788859401e-05,
+      "loss": 3.4103,
+      "step": 2365
+    },
+    {
+      "epoch": 0.503538175046555,
+      "grad_norm": 1.4666434796564831,
+      "learning_rate": 9.86048753554118e-05,
+      "loss": 3.4327,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5037509976057462,
+      "grad_norm": 1.1383995322691138,
+      "learning_rate": 9.860196884316176e-05,
+      "loss": 3.443,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5039638201649375,
+      "grad_norm": 1.9079067207515361,
+      "learning_rate": 9.85990593493683e-05,
+      "loss": 3.4635,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5041766427241288,
+      "grad_norm": 1.2830253169107728,
+      "learning_rate": 9.859614687420989e-05,
+      "loss": 3.4467,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5043894652833201,
+      "grad_norm": 1.5118030965325777,
+      "learning_rate": 9.85932314178652e-05,
+      "loss": 3.4045,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5046022878425113,
+      "grad_norm": 1.321205863155472,
+      "learning_rate": 9.85903129805131e-05,
+      "loss": 3.4151,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5048151104017026,
+      "grad_norm": 1.4511675990463453,
+      "learning_rate": 9.858739156233263e-05,
+      "loss": 3.4419,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5050279329608939,
+      "grad_norm": 1.3393057050824666,
+      "learning_rate": 9.858446716350301e-05,
+      "loss": 3.4071,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5052407555200852,
+      "grad_norm": 1.263643214103833,
+      "learning_rate": 9.858153978420359e-05,
+      "loss": 3.4372,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5054535780792764,
+      "grad_norm": 1.3227182872425536,
+      "learning_rate": 9.8578609424614e-05,
+      "loss": 3.4362,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5056664006384677,
+      "grad_norm": 1.54937024909769,
+      "learning_rate": 9.857567608491399e-05,
+      "loss": 3.5017,
+      "step": 2376
+    },
+    {
+      "epoch": 0.505879223197659,
+      "grad_norm": 1.3312675138665926,
+      "learning_rate": 9.857273976528351e-05,
+      "loss": 3.4197,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5060920457568502,
+      "grad_norm": 1.5834101614670604,
+      "learning_rate": 9.856980046590272e-05,
+      "loss": 3.5284,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5063048683160415,
+      "grad_norm": 1.1687733806814853,
+      "learning_rate": 9.856685818695189e-05,
+      "loss": 3.4653,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5065176908752328,
+      "grad_norm": 1.5070232902038743,
+      "learning_rate": 9.856391292861153e-05,
+      "loss": 3.4645,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5067305134344241,
+      "grad_norm": 1.2533106224725876,
+      "learning_rate": 9.856096469106235e-05,
+      "loss": 3.4374,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5069433359936153,
+      "grad_norm": 1.481645691349127,
+      "learning_rate": 9.855801347448519e-05,
+      "loss": 3.4408,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5071561585528066,
+      "grad_norm": 1.3141840479828282,
+      "learning_rate": 9.855505927906108e-05,
+      "loss": 3.4637,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5073689811119979,
+      "grad_norm": 1.3441125767321385,
+      "learning_rate": 9.855210210497128e-05,
+      "loss": 3.4115,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5075818036711891,
+      "grad_norm": 1.2828005966898042,
+      "learning_rate": 9.854914195239718e-05,
+      "loss": 3.442,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5077946262303804,
+      "grad_norm": 1.3680080763556735,
+      "learning_rate": 9.854617882152038e-05,
+      "loss": 3.4676,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5080074487895717,
+      "grad_norm": 1.2430581873148392,
+      "learning_rate": 9.854321271252265e-05,
+      "loss": 3.4327,
+      "step": 2387
+    },
+    {
+      "epoch": 0.508220271348763,
+      "grad_norm": 1.569080499333323,
+      "learning_rate": 9.854024362558596e-05,
+      "loss": 3.4172,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5084330939079542,
+      "grad_norm": 1.2450835912385911,
+      "learning_rate": 9.853727156089245e-05,
+      "loss": 3.4079,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5086459164671455,
+      "grad_norm": 1.409729787073755,
+      "learning_rate": 9.853429651862445e-05,
+      "loss": 3.4088,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5088587390263368,
+      "grad_norm": 1.213129865081596,
+      "learning_rate": 9.853131849896446e-05,
+      "loss": 3.388,
+      "step": 2391
+    },
+    {
+      "epoch": 0.509071561585528,
+      "grad_norm": 1.5181365707012615,
+      "learning_rate": 9.852833750209518e-05,
+      "loss": 3.4084,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5092843841447193,
+      "grad_norm": 1.3308940636118767,
+      "learning_rate": 9.852535352819945e-05,
+      "loss": 3.4451,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5094972067039106,
+      "grad_norm": 1.5085112822834172,
+      "learning_rate": 9.852236657746035e-05,
+      "loss": 3.4431,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5097100292631019,
+      "grad_norm": 1.109113025501849,
+      "learning_rate": 9.851937665006114e-05,
+      "loss": 3.3929,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5099228518222931,
+      "grad_norm": 1.4845151169468582,
+      "learning_rate": 9.85163837461852e-05,
+      "loss": 3.3795,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5101356743814844,
+      "grad_norm": 1.288334403170574,
+      "learning_rate": 9.851338786601614e-05,
+      "loss": 3.4693,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5103484969406757,
+      "grad_norm": 1.5677354895882338,
+      "learning_rate": 9.851038900973778e-05,
+      "loss": 3.4318,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5105613194998669,
+      "grad_norm": 1.2023458314530722,
+      "learning_rate": 9.850738717753405e-05,
+      "loss": 3.4432,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5107741420590582,
+      "grad_norm": 1.4135923983938856,
+      "learning_rate": 9.850438236958912e-05,
+      "loss": 3.426,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5109869646182496,
+      "grad_norm": 1.28764854253577,
+      "learning_rate": 9.850137458608731e-05,
+      "loss": 3.3987,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5111997871774409,
+      "grad_norm": 1.3848395229257733,
+      "learning_rate": 9.849836382721315e-05,
+      "loss": 3.3978,
+      "step": 2402
+    },
+    {
+      "epoch": 0.511412609736632,
+      "grad_norm": 1.3752879251862429,
+      "learning_rate": 9.849535009315134e-05,
+      "loss": 3.4384,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5116254322958234,
+      "grad_norm": 1.3485229711478988,
+      "learning_rate": 9.849233338408674e-05,
+      "loss": 3.432,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5118382548550147,
+      "grad_norm": 1.278015199491982,
+      "learning_rate": 9.848931370020444e-05,
+      "loss": 3.3858,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5120510774142059,
+      "grad_norm": 1.7037323573347423,
+      "learning_rate": 9.848629104168967e-05,
+      "loss": 3.4429,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5122638999733972,
+      "grad_norm": 1.1171766241654155,
+      "learning_rate": 9.848326540872786e-05,
+      "loss": 3.4049,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5124767225325885,
+      "grad_norm": 1.810999453038205,
+      "learning_rate": 9.848023680150463e-05,
+      "loss": 3.4225,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5126895450917798,
+      "grad_norm": 1.0414136108747296,
+      "learning_rate": 9.847720522020574e-05,
+      "loss": 3.4392,
+      "step": 2409
+    },
+    {
+      "epoch": 0.512902367650971,
+      "grad_norm": 1.6215664229501634,
+      "learning_rate": 9.847417066501721e-05,
+      "loss": 3.3999,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5131151902101623,
+      "grad_norm": 1.2392286839028628,
+      "learning_rate": 9.847113313612517e-05,
+      "loss": 3.4471,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5133280127693536,
+      "grad_norm": 1.3498503921180232,
+      "learning_rate": 9.846809263371598e-05,
+      "loss": 3.425,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5135408353285448,
+      "grad_norm": 1.4622471054446626,
+      "learning_rate": 9.846504915797617e-05,
+      "loss": 3.4052,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5137536578877361,
+      "grad_norm": 1.271573662215953,
+      "learning_rate": 9.846200270909241e-05,
+      "loss": 3.4451,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5139664804469274,
+      "grad_norm": 1.4836015359515282,
+      "learning_rate": 9.845895328725161e-05,
+      "loss": 3.3849,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5141793030061187,
+      "grad_norm": 1.29366308441987,
+      "learning_rate": 9.845590089264083e-05,
+      "loss": 3.3914,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5143921255653099,
+      "grad_norm": 1.3699433948039315,
+      "learning_rate": 9.845284552544733e-05,
+      "loss": 3.4512,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5146049481245012,
+      "grad_norm": 1.6172201843344534,
+      "learning_rate": 9.844978718585855e-05,
+      "loss": 3.4623,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5148177706836925,
+      "grad_norm": 1.1742713757283771,
+      "learning_rate": 9.844672587406212e-05,
+      "loss": 3.4093,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5150305932428837,
+      "grad_norm": 1.527573597124004,
+      "learning_rate": 9.844366159024581e-05,
+      "loss": 3.412,
+      "step": 2420
+    },
+    {
+      "epoch": 0.515243415802075,
+      "grad_norm": 1.1528614806082935,
+      "learning_rate": 9.844059433459761e-05,
+      "loss": 3.4171,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5154562383612663,
+      "grad_norm": 1.4852594103849595,
+      "learning_rate": 9.843752410730568e-05,
+      "loss": 3.4631,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5156690609204576,
+      "grad_norm": 1.3633376299224103,
+      "learning_rate": 9.843445090855839e-05,
+      "loss": 3.3913,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5158818834796488,
+      "grad_norm": 1.3859858943303138,
+      "learning_rate": 9.843137473854424e-05,
+      "loss": 3.4347,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5160947060388401,
+      "grad_norm": 1.598438473407897,
+      "learning_rate": 9.842829559745195e-05,
+      "loss": 3.418,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5163075285980314,
+      "grad_norm": 1.3836248338003385,
+      "learning_rate": 9.842521348547045e-05,
+      "loss": 3.3985,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5165203511572226,
+      "grad_norm": 1.3511920385845975,
+      "learning_rate": 9.842212840278875e-05,
+      "loss": 3.4043,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5167331737164139,
+      "grad_norm": 1.312118008411586,
+      "learning_rate": 9.841904034959618e-05,
+      "loss": 3.4033,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5169459962756052,
+      "grad_norm": 1.4779035433114387,
+      "learning_rate": 9.84159493260821e-05,
+      "loss": 3.4258,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5171588188347965,
+      "grad_norm": 1.2459696782766567,
+      "learning_rate": 9.84128553324362e-05,
+      "loss": 3.3865,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5173716413939877,
+      "grad_norm": 1.4338138640882576,
+      "learning_rate": 9.840975836884824e-05,
+      "loss": 3.4153,
+      "step": 2431
+    },
+    {
+      "epoch": 0.517584463953179,
+      "grad_norm": 1.1805593302848003,
+      "learning_rate": 9.840665843550824e-05,
+      "loss": 3.3886,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5177972865123703,
+      "grad_norm": 1.4980898238226918,
+      "learning_rate": 9.840355553260637e-05,
+      "loss": 3.4002,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5180101090715615,
+      "grad_norm": 1.215511129783755,
+      "learning_rate": 9.840044966033296e-05,
+      "loss": 3.4086,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5182229316307528,
+      "grad_norm": 1.4786930150896407,
+      "learning_rate": 9.839734081887853e-05,
+      "loss": 3.4494,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5184357541899441,
+      "grad_norm": 1.3111301048677606,
+      "learning_rate": 9.839422900843382e-05,
+      "loss": 3.358,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5186485767491354,
+      "grad_norm": 1.3553394016845186,
+      "learning_rate": 9.839111422918973e-05,
+      "loss": 3.4436,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5188613993083266,
+      "grad_norm": 1.1804591138295313,
+      "learning_rate": 9.838799648133732e-05,
+      "loss": 3.3742,
+      "step": 2438
+    },
+    {
+      "epoch": 0.519074221867518,
+      "grad_norm": 1.3333206314714663,
+      "learning_rate": 9.838487576506787e-05,
+      "loss": 3.4259,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5192870444267093,
+      "grad_norm": 1.5653112043059563,
+      "learning_rate": 9.838175208057283e-05,
+      "loss": 3.4239,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5194998669859006,
+      "grad_norm": 1.36142449513553,
+      "learning_rate": 9.837862542804381e-05,
+      "loss": 3.4275,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5197126895450918,
+      "grad_norm": 1.1632705655205846,
+      "learning_rate": 9.837549580767261e-05,
+      "loss": 3.3907,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5199255121042831,
+      "grad_norm": 1.3126244186046048,
+      "learning_rate": 9.837236321965124e-05,
+      "loss": 3.4287,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5201383346634744,
+      "grad_norm": 1.5448252581680633,
+      "learning_rate": 9.836922766417188e-05,
+      "loss": 3.4075,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5203511572226656,
+      "grad_norm": 1.2493464274694668,
+      "learning_rate": 9.836608914142684e-05,
+      "loss": 3.4141,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5205639797818569,
+      "grad_norm": 1.5344992839418574,
+      "learning_rate": 9.836294765160869e-05,
+      "loss": 3.3877,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5207768023410482,
+      "grad_norm": 1.0516915758682954,
+      "learning_rate": 9.835980319491016e-05,
+      "loss": 3.4269,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5209896249002395,
+      "grad_norm": 1.5932520066590938,
+      "learning_rate": 9.835665577152411e-05,
+      "loss": 3.4214,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5212024474594307,
+      "grad_norm": 1.198444161708372,
+      "learning_rate": 9.835350538164366e-05,
+      "loss": 3.4153,
+      "step": 2449
+    },
+    {
+      "epoch": 0.521415270018622,
+      "grad_norm": 1.7942515692531533,
+      "learning_rate": 9.835035202546205e-05,
+      "loss": 3.4464,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5216280925778133,
+      "grad_norm": 1.0691070760511139,
+      "learning_rate": 9.834719570317276e-05,
+      "loss": 3.3802,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5218409151370045,
+      "grad_norm": 1.723437195246177,
+      "learning_rate": 9.834403641496938e-05,
+      "loss": 3.4154,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5220537376961958,
+      "grad_norm": 1.1836724759081008,
+      "learning_rate": 9.834087416104574e-05,
+      "loss": 3.3675,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5222665602553871,
+      "grad_norm": 1.3522231017541464,
+      "learning_rate": 9.833770894159581e-05,
+      "loss": 3.4205,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5224793828145784,
+      "grad_norm": 1.3113771721338616,
+      "learning_rate": 9.833454075681378e-05,
+      "loss": 3.4052,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5226922053737696,
+      "grad_norm": 1.2965087343034938,
+      "learning_rate": 9.833136960689402e-05,
+      "loss": 3.4028,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5229050279329609,
+      "grad_norm": 1.4879090286853571,
+      "learning_rate": 9.832819549203106e-05,
+      "loss": 3.4484,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5231178504921522,
+      "grad_norm": 1.3741605689717158,
+      "learning_rate": 9.832501841241961e-05,
+      "loss": 3.3896,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5233306730513434,
+      "grad_norm": 1.26228272347239,
+      "learning_rate": 9.832183836825458e-05,
+      "loss": 3.4321,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5235434956105347,
+      "grad_norm": 1.7293494339557018,
+      "learning_rate": 9.831865535973103e-05,
+      "loss": 3.388,
+      "step": 2460
+    },
+    {
+      "epoch": 0.523756318169726,
+      "grad_norm": 1.27967758598576,
+      "learning_rate": 9.831546938704426e-05,
+      "loss": 3.4096,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5239691407289173,
+      "grad_norm": 1.6697398640180436,
+      "learning_rate": 9.83122804503897e-05,
+      "loss": 3.4208,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5241819632881085,
+      "grad_norm": 1.272778493139224,
+      "learning_rate": 9.830908854996297e-05,
+      "loss": 3.4416,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5243947858472998,
+      "grad_norm": 1.3332817883908923,
+      "learning_rate": 9.83058936859599e-05,
+      "loss": 3.3708,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5246076084064911,
+      "grad_norm": 1.3873577926481608,
+      "learning_rate": 9.830269585857647e-05,
+      "loss": 3.3925,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5248204309656823,
+      "grad_norm": 1.2696594608104668,
+      "learning_rate": 9.829949506800886e-05,
+      "loss": 3.4462,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5250332535248736,
+      "grad_norm": 1.4307273824876419,
+      "learning_rate": 9.829629131445342e-05,
+      "loss": 3.3946,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5252460760840649,
+      "grad_norm": 1.4278297835435199,
+      "learning_rate": 9.82930845981067e-05,
+      "loss": 3.4853,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5254588986432562,
+      "grad_norm": 1.2151303920055805,
+      "learning_rate": 9.82898749191654e-05,
+      "loss": 3.3753,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5256717212024474,
+      "grad_norm": 1.2223932695702668,
+      "learning_rate": 9.828666227782645e-05,
+      "loss": 3.3744,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5258845437616387,
+      "grad_norm": 1.379375817644663,
+      "learning_rate": 9.828344667428691e-05,
+      "loss": 3.3791,
+      "step": 2471
+    },
+    {
+      "epoch": 0.52609736632083,
+      "grad_norm": 1.3958214088708105,
+      "learning_rate": 9.828022810874404e-05,
+      "loss": 3.4191,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5263101888800212,
+      "grad_norm": 1.3770739532241698,
+      "learning_rate": 9.827700658139531e-05,
+      "loss": 3.3806,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5265230114392125,
+      "grad_norm": 1.289826472098686,
+      "learning_rate": 9.827378209243834e-05,
+      "loss": 3.4092,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5267358339984038,
+      "grad_norm": 1.2167373083068433,
+      "learning_rate": 9.827055464207094e-05,
+      "loss": 3.4486,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5269486565575952,
+      "grad_norm": 1.3975186440025607,
+      "learning_rate": 9.826732423049107e-05,
+      "loss": 3.4296,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5271614791167863,
+      "grad_norm": 1.3145220278759306,
+      "learning_rate": 9.826409085789696e-05,
+      "loss": 3.3938,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5273743016759777,
+      "grad_norm": 1.4009398433479554,
+      "learning_rate": 9.826085452448694e-05,
+      "loss": 3.4069,
+      "step": 2478
+    },
+    {
+      "epoch": 0.527587124235169,
+      "grad_norm": 1.3645966457623604,
+      "learning_rate": 9.825761523045952e-05,
+      "loss": 3.4208,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5277999467943602,
+      "grad_norm": 1.1501489658620363,
+      "learning_rate": 9.825437297601345e-05,
+      "loss": 3.3757,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5280127693535515,
+      "grad_norm": 1.6757335066662749,
+      "learning_rate": 9.825112776134762e-05,
+      "loss": 3.4434,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5282255919127428,
+      "grad_norm": 1.1177927082436356,
+      "learning_rate": 9.824787958666111e-05,
+      "loss": 3.3944,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5284384144719341,
+      "grad_norm": 1.534459162406453,
+      "learning_rate": 9.82446284521532e-05,
+      "loss": 3.3486,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5286512370311253,
+      "grad_norm": 1.0727173220108859,
+      "learning_rate": 9.824137435802331e-05,
+      "loss": 3.4057,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5288640595903166,
+      "grad_norm": 1.5939621998489022,
+      "learning_rate": 9.823811730447107e-05,
+      "loss": 3.4091,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5290768821495079,
+      "grad_norm": 1.2441929948196988,
+      "learning_rate": 9.82348572916963e-05,
+      "loss": 3.3693,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5292897047086991,
+      "grad_norm": 1.5446085193481276,
+      "learning_rate": 9.823159431989899e-05,
+      "loss": 3.438,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5295025272678904,
+      "grad_norm": 1.245116593095709,
+      "learning_rate": 9.822832838927929e-05,
+      "loss": 3.3705,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5297153498270817,
+      "grad_norm": 1.3860326340971743,
+      "learning_rate": 9.822505950003758e-05,
+      "loss": 3.3801,
+      "step": 2489
+    },
+    {
+      "epoch": 0.529928172386273,
+      "grad_norm": 1.3172390152556002,
+      "learning_rate": 9.822178765237436e-05,
+      "loss": 3.3945,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5301409949454642,
+      "grad_norm": 1.369447051113877,
+      "learning_rate": 9.821851284649037e-05,
+      "loss": 3.3789,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5303538175046555,
+      "grad_norm": 1.3050350194212381,
+      "learning_rate": 9.821523508258652e-05,
+      "loss": 3.359,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5305666400638468,
+      "grad_norm": 1.3406407606677098,
+      "learning_rate": 9.821195436086385e-05,
+      "loss": 3.3872,
+      "step": 2493
+    },
+    {
+      "epoch": 0.530779462623038,
+      "grad_norm": 1.3673940728080418,
+      "learning_rate": 9.820867068152364e-05,
+      "loss": 3.3554,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5309922851822293,
+      "grad_norm": 1.3293626522068251,
+      "learning_rate": 9.820538404476734e-05,
+      "loss": 3.3998,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5312051077414206,
+      "grad_norm": 1.1872270654465744,
+      "learning_rate": 9.820209445079656e-05,
+      "loss": 3.3392,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5314179303006119,
+      "grad_norm": 1.4115822270791332,
+      "learning_rate": 9.819880189981311e-05,
+      "loss": 3.3883,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5316307528598031,
+      "grad_norm": 1.3133699016814608,
+      "learning_rate": 9.819550639201897e-05,
+      "loss": 3.3734,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5318435754189944,
+      "grad_norm": 1.3932972589194061,
+      "learning_rate": 9.81922079276163e-05,
+      "loss": 3.3443,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5320563979781857,
+      "grad_norm": 1.4637193423240864,
+      "learning_rate": 9.818890650680747e-05,
+      "loss": 3.3965,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5322692205373769,
+      "grad_norm": 1.247091758774057,
+      "learning_rate": 9.8185602129795e-05,
+      "loss": 3.3725,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5324820430965682,
+      "grad_norm": 1.6914493882196224,
+      "learning_rate": 9.818229479678158e-05,
+      "loss": 3.3255,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5326948656557595,
+      "grad_norm": 0.9236357274963063,
+      "learning_rate": 9.817898450797012e-05,
+      "loss": 3.3721,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5329076882149508,
+      "grad_norm": 1.3665109259671229,
+      "learning_rate": 9.817567126356371e-05,
+      "loss": 3.4046,
+      "step": 2504
+    },
+    {
+      "epoch": 0.533120510774142,
+      "grad_norm": 1.1416725668999776,
+      "learning_rate": 9.817235506376556e-05,
+      "loss": 3.3993,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.7500800697117542,
+      "learning_rate": 9.816903590877917e-05,
+      "loss": 3.4024,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5335461558925246,
+      "grad_norm": 1.159254587263203,
+      "learning_rate": 9.81657137988081e-05,
+      "loss": 3.3815,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5337589784517158,
+      "grad_norm": 1.693183954465906,
+      "learning_rate": 9.816238873405616e-05,
+      "loss": 3.4009,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5339718010109071,
+      "grad_norm": 1.1521913254735743,
+      "learning_rate": 9.815906071472736e-05,
+      "loss": 3.3815,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5341846235700984,
+      "grad_norm": 1.4617589465630798,
+      "learning_rate": 9.815572974102582e-05,
+      "loss": 3.3874,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5343974461292897,
+      "grad_norm": 1.2998763625090792,
+      "learning_rate": 9.815239581315592e-05,
+      "loss": 3.3958,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5346102686884809,
+      "grad_norm": 1.1586256933836259,
+      "learning_rate": 9.814905893132215e-05,
+      "loss": 3.4094,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5348230912476722,
+      "grad_norm": 1.3620912367446554,
+      "learning_rate": 9.814571909572925e-05,
+      "loss": 3.4135,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5350359138068636,
+      "grad_norm": 1.4728911341690418,
+      "learning_rate": 9.814237630658208e-05,
+      "loss": 3.3655,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5352487363660549,
+      "grad_norm": 1.2341175163337488,
+      "learning_rate": 9.813903056408572e-05,
+      "loss": 3.4063,
+      "step": 2515
+    },
+    {
+      "epoch": 0.535461558925246,
+      "grad_norm": 1.3080908704635257,
+      "learning_rate": 9.813568186844541e-05,
+      "loss": 3.4134,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5356743814844374,
+      "grad_norm": 1.3654367439332793,
+      "learning_rate": 9.813233021986659e-05,
+      "loss": 3.4273,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5358872040436287,
+      "grad_norm": 1.3464108608587555,
+      "learning_rate": 9.812897561855484e-05,
+      "loss": 3.3822,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5361000266028199,
+      "grad_norm": 1.3146315707540226,
+      "learning_rate": 9.812561806471598e-05,
+      "loss": 3.3781,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5363128491620112,
+      "grad_norm": 1.3610778813632436,
+      "learning_rate": 9.812225755855599e-05,
+      "loss": 3.3649,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5365256717212025,
+      "grad_norm": 1.4320834043411204,
+      "learning_rate": 9.811889410028102e-05,
+      "loss": 3.3584,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5367384942803938,
+      "grad_norm": 1.2693149934471275,
+      "learning_rate": 9.811552769009738e-05,
+      "loss": 3.3505,
+      "step": 2522
+    },
+    {
+      "epoch": 0.536951316839585,
+      "grad_norm": 1.3328836894499827,
+      "learning_rate": 9.811215832821161e-05,
+      "loss": 3.3767,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5371641393987763,
+      "grad_norm": 1.2869315643474561,
+      "learning_rate": 9.810878601483042e-05,
+      "loss": 3.3773,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5373769619579676,
+      "grad_norm": 1.4061741274299069,
+      "learning_rate": 9.810541075016065e-05,
+      "loss": 3.3798,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5375897845171588,
+      "grad_norm": 1.111385102886947,
+      "learning_rate": 9.810203253440937e-05,
+      "loss": 3.3632,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5378026070763501,
+      "grad_norm": 1.6878435977378299,
+      "learning_rate": 9.809865136778384e-05,
+      "loss": 3.3715,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5380154296355414,
+      "grad_norm": 1.1021747177020618,
+      "learning_rate": 9.809526725049147e-05,
+      "loss": 3.3464,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5382282521947327,
+      "grad_norm": 1.747900332140184,
+      "learning_rate": 9.809188018273986e-05,
+      "loss": 3.4199,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5384410747539239,
+      "grad_norm": 1.232508454198229,
+      "learning_rate": 9.808849016473683e-05,
+      "loss": 3.3701,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5386538973131152,
+      "grad_norm": 1.4166213553975826,
+      "learning_rate": 9.808509719669028e-05,
+      "loss": 3.3574,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5388667198723065,
+      "grad_norm": 1.0406672662476069,
+      "learning_rate": 9.808170127880838e-05,
+      "loss": 3.3675,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5390795424314977,
+      "grad_norm": 1.749803965115892,
+      "learning_rate": 9.807830241129948e-05,
+      "loss": 3.372,
+      "step": 2533
+    },
+    {
+      "epoch": 0.539292364990689,
+      "grad_norm": 1.1335866759788453,
+      "learning_rate": 9.807490059437208e-05,
+      "loss": 3.3946,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5395051875498803,
+      "grad_norm": 1.4516343242006606,
+      "learning_rate": 9.807149582823484e-05,
+      "loss": 3.3737,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5397180101090716,
+      "grad_norm": 1.5236547172699237,
+      "learning_rate": 9.806808811309666e-05,
+      "loss": 3.3795,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5399308326682628,
+      "grad_norm": 1.012352303715033,
+      "learning_rate": 9.806467744916656e-05,
+      "loss": 3.3533,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5401436552274541,
+      "grad_norm": 1.8099668600005703,
+      "learning_rate": 9.806126383665381e-05,
+      "loss": 3.3496,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5403564777866454,
+      "grad_norm": 1.2850453030391131,
+      "learning_rate": 9.805784727576782e-05,
+      "loss": 3.3502,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5405693003458366,
+      "grad_norm": 1.3701044309949444,
+      "learning_rate": 9.805442776671813e-05,
+      "loss": 3.366,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5407821229050279,
+      "grad_norm": 1.3219946758697574,
+      "learning_rate": 9.805100530971456e-05,
+      "loss": 3.3835,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5409949454642192,
+      "grad_norm": 1.4914846736796492,
+      "learning_rate": 9.804757990496704e-05,
+      "loss": 3.3904,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5412077680234105,
+      "grad_norm": 1.2529980341406826,
+      "learning_rate": 9.804415155268574e-05,
+      "loss": 3.3341,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5414205905826017,
+      "grad_norm": 1.2759400316560756,
+      "learning_rate": 9.804072025308095e-05,
+      "loss": 3.3516,
+      "step": 2544
+    },
+    {
+      "epoch": 0.541633413141793,
+      "grad_norm": 1.4906595954105673,
+      "learning_rate": 9.803728600636317e-05,
+      "loss": 3.3783,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5418462357009843,
+      "grad_norm": 1.1358784784736942,
+      "learning_rate": 9.803384881274308e-05,
+      "loss": 3.4276,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5420590582601755,
+      "grad_norm": 1.5880954908413716,
+      "learning_rate": 9.803040867243154e-05,
+      "loss": 3.3701,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5422718808193668,
+      "grad_norm": 1.2237210107428516,
+      "learning_rate": 9.802696558563958e-05,
+      "loss": 3.3675,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5424847033785581,
+      "grad_norm": 1.3651657857670687,
+      "learning_rate": 9.802351955257843e-05,
+      "loss": 3.3524,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5426975259377494,
+      "grad_norm": 1.1865271513886584,
+      "learning_rate": 9.80200705734595e-05,
+      "loss": 3.411,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5429103484969406,
+      "grad_norm": 1.483942471577189,
+      "learning_rate": 9.801661864849434e-05,
+      "loss": 3.398,
+      "step": 2551
+    },
+    {
+      "epoch": 0.543123171056132,
+      "grad_norm": 1.1809673271993129,
+      "learning_rate": 9.801316377789474e-05,
+      "loss": 3.4042,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5433359936153233,
+      "grad_norm": 1.5329177143788963,
+      "learning_rate": 9.800970596187265e-05,
+      "loss": 3.3942,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5435488161745144,
+      "grad_norm": 1.1289266350581626,
+      "learning_rate": 9.800624520064017e-05,
+      "loss": 3.3764,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5437616387337058,
+      "grad_norm": 1.3975348754506256,
+      "learning_rate": 9.80027814944096e-05,
+      "loss": 3.3847,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5439744612928971,
+      "grad_norm": 1.3061763965234459,
+      "learning_rate": 9.799931484339344e-05,
+      "loss": 3.355,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5441872838520884,
+      "grad_norm": 1.513171596755743,
+      "learning_rate": 9.799584524780437e-05,
+      "loss": 3.3974,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5444001064112796,
+      "grad_norm": 1.1777423195113792,
+      "learning_rate": 9.799237270785521e-05,
+      "loss": 3.3318,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5446129289704709,
+      "grad_norm": 1.4252645044550802,
+      "learning_rate": 9.7988897223759e-05,
+      "loss": 3.4246,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5448257515296622,
+      "grad_norm": 1.2340104132695153,
+      "learning_rate": 9.798541879572896e-05,
+      "loss": 3.3616,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5450385740888534,
+      "grad_norm": 1.4009081986617649,
+      "learning_rate": 9.798193742397843e-05,
+      "loss": 3.3386,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5452513966480447,
+      "grad_norm": 1.3516853568264882,
+      "learning_rate": 9.797845310872104e-05,
+      "loss": 3.3636,
+      "step": 2562
+    },
+    {
+      "epoch": 0.545464219207236,
+      "grad_norm": 1.2099091442024579,
+      "learning_rate": 9.79749658501705e-05,
+      "loss": 3.3712,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5456770417664273,
+      "grad_norm": 1.4090665474225643,
+      "learning_rate": 9.797147564854075e-05,
+      "loss": 3.3713,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5458898643256185,
+      "grad_norm": 1.1990235201813808,
+      "learning_rate": 9.796798250404591e-05,
+      "loss": 3.4026,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5461026868848098,
+      "grad_norm": 1.4238892152845337,
+      "learning_rate": 9.796448641690025e-05,
+      "loss": 3.3518,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5463155094440011,
+      "grad_norm": 1.191460778971336,
+      "learning_rate": 9.796098738731825e-05,
+      "loss": 3.3542,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5465283320031923,
+      "grad_norm": 1.4842068263167285,
+      "learning_rate": 9.795748541551457e-05,
+      "loss": 3.3663,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5467411545623836,
+      "grad_norm": 1.1399589242234505,
+      "learning_rate": 9.795398050170406e-05,
+      "loss": 3.3043,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5469539771215749,
+      "grad_norm": 1.474495493329563,
+      "learning_rate": 9.795047264610168e-05,
+      "loss": 3.3744,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5471667996807662,
+      "grad_norm": 0.997080943976407,
+      "learning_rate": 9.794696184892267e-05,
+      "loss": 3.3625,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5473796222399574,
+      "grad_norm": 1.413269177390686,
+      "learning_rate": 9.794344811038239e-05,
+      "loss": 3.3683,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5475924447991487,
+      "grad_norm": 1.3911471943625724,
+      "learning_rate": 9.793993143069639e-05,
+      "loss": 3.3775,
+      "step": 2573
+    },
+    {
+      "epoch": 0.54780526735834,
+      "grad_norm": 1.3220547672210228,
+      "learning_rate": 9.793641181008041e-05,
+      "loss": 3.3558,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5480180899175312,
+      "grad_norm": 1.250392603875186,
+      "learning_rate": 9.793288924875034e-05,
+      "loss": 3.3738,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5482309124767225,
+      "grad_norm": 1.3390602525161825,
+      "learning_rate": 9.792936374692233e-05,
+      "loss": 3.402,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5484437350359138,
+      "grad_norm": 1.2484806908843016,
+      "learning_rate": 9.79258353048126e-05,
+      "loss": 3.3177,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5486565575951051,
+      "grad_norm": 1.1909448752691594,
+      "learning_rate": 9.792230392263766e-05,
+      "loss": 3.3364,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5488693801542963,
+      "grad_norm": 1.3736111681246632,
+      "learning_rate": 9.79187696006141e-05,
+      "loss": 3.3216,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5490822027134876,
+      "grad_norm": 1.3237029728446124,
+      "learning_rate": 9.791523233895875e-05,
+      "loss": 3.3662,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5492950252726789,
+      "grad_norm": 1.3369737937072672,
+      "learning_rate": 9.791169213788862e-05,
+      "loss": 3.3876,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5495078478318702,
+      "grad_norm": 1.2358548455389335,
+      "learning_rate": 9.790814899762087e-05,
+      "loss": 3.3888,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5497206703910614,
+      "grad_norm": 1.3301261415744245,
+      "learning_rate": 9.790460291837286e-05,
+      "loss": 3.3406,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5499334929502527,
+      "grad_norm": 1.091014225699254,
+      "learning_rate": 9.790105390036214e-05,
+      "loss": 3.3793,
+      "step": 2584
+    },
+    {
+      "epoch": 0.550146315509444,
+      "grad_norm": 1.5581852923650905,
+      "learning_rate": 9.789750194380645e-05,
+      "loss": 3.3411,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5503591380686352,
+      "grad_norm": 1.1552634390175092,
+      "learning_rate": 9.789394704892364e-05,
+      "loss": 3.3481,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5505719606278265,
+      "grad_norm": 1.5589651400957603,
+      "learning_rate": 9.789038921593181e-05,
+      "loss": 3.3914,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5507847831870178,
+      "grad_norm": 1.2323403744674888,
+      "learning_rate": 9.788682844504923e-05,
+      "loss": 3.3526,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5509976057462092,
+      "grad_norm": 1.549736983634238,
+      "learning_rate": 9.788326473649433e-05,
+      "loss": 3.4023,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5512104283054003,
+      "grad_norm": 1.195008931209447,
+      "learning_rate": 9.787969809048574e-05,
+      "loss": 3.3481,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5514232508645917,
+      "grad_norm": 1.706491315216526,
+      "learning_rate": 9.787612850724224e-05,
+      "loss": 3.3709,
+      "step": 2591
+    },
+    {
+      "epoch": 0.551636073423783,
+      "grad_norm": 1.161975654131822,
+      "learning_rate": 9.787255598698281e-05,
+      "loss": 3.3667,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5518488959829742,
+      "grad_norm": 1.5800131184659176,
+      "learning_rate": 9.786898052992664e-05,
+      "loss": 3.36,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5520617185421655,
+      "grad_norm": 1.197133206901303,
+      "learning_rate": 9.786540213629304e-05,
+      "loss": 3.3588,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5522745411013568,
+      "grad_norm": 1.434032670010433,
+      "learning_rate": 9.786182080630154e-05,
+      "loss": 3.3768,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5524873636605481,
+      "grad_norm": 1.3473293191494093,
+      "learning_rate": 9.785823654017186e-05,
+      "loss": 3.3808,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5527001862197393,
+      "grad_norm": 1.1075281362207647,
+      "learning_rate": 9.785464933812385e-05,
+      "loss": 3.3905,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5529130087789306,
+      "grad_norm": 1.5721818493087028,
+      "learning_rate": 9.785105920037759e-05,
+      "loss": 3.3611,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5531258313381219,
+      "grad_norm": 1.2291147483042557,
+      "learning_rate": 9.78474661271533e-05,
+      "loss": 3.3684,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5533386538973131,
+      "grad_norm": 1.4331852748905587,
+      "learning_rate": 9.784387011867144e-05,
+      "loss": 3.3336,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5535514764565044,
+      "grad_norm": 1.3470255152671542,
+      "learning_rate": 9.784027117515258e-05,
+      "loss": 3.359,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5537642990156957,
+      "grad_norm": 1.3418601865623743,
+      "learning_rate": 9.78366692968175e-05,
+      "loss": 3.3426,
+      "step": 2602
+    },
+    {
+      "epoch": 0.553977121574887,
+      "grad_norm": 1.1418285388045852,
+      "learning_rate": 9.783306448388718e-05,
+      "loss": 3.3243,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5541899441340782,
+      "grad_norm": 1.311186023367553,
+      "learning_rate": 9.782945673658276e-05,
+      "loss": 3.329,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5544027666932695,
+      "grad_norm": 1.2362661876191015,
+      "learning_rate": 9.782584605512554e-05,
+      "loss": 3.3526,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5546155892524608,
+      "grad_norm": 1.505096025118102,
+      "learning_rate": 9.782223243973705e-05,
+      "loss": 3.3955,
+      "step": 2606
+    },
+    {
+      "epoch": 0.554828411811652,
+      "grad_norm": 1.1587118078781893,
+      "learning_rate": 9.781861589063895e-05,
+      "loss": 3.3047,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5550412343708433,
+      "grad_norm": 1.2467430468257825,
+      "learning_rate": 9.781499640805312e-05,
+      "loss": 3.3667,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5552540569300346,
+      "grad_norm": 1.4238843329184376,
+      "learning_rate": 9.781137399220158e-05,
+      "loss": 3.3612,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5554668794892259,
+      "grad_norm": 1.2518727547233652,
+      "learning_rate": 9.780774864330655e-05,
+      "loss": 3.3303,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5556797020484171,
+      "grad_norm": 1.3646496926342706,
+      "learning_rate": 9.780412036159045e-05,
+      "loss": 3.4023,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5558925246076084,
+      "grad_norm": 1.1966682582509316,
+      "learning_rate": 9.780048914727586e-05,
+      "loss": 3.3119,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5561053471667997,
+      "grad_norm": 1.4892955616228916,
+      "learning_rate": 9.779685500058552e-05,
+      "loss": 3.3468,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5563181697259909,
+      "grad_norm": 1.0094518500672705,
+      "learning_rate": 9.779321792174238e-05,
+      "loss": 3.3365,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5565309922851822,
+      "grad_norm": 1.6716299633086809,
+      "learning_rate": 9.778957791096959e-05,
+      "loss": 3.3785,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5567438148443735,
+      "grad_norm": 1.026914416041159,
+      "learning_rate": 9.77859349684904e-05,
+      "loss": 3.369,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5569566374035648,
+      "grad_norm": 1.5959559308739557,
+      "learning_rate": 9.778228909452834e-05,
+      "loss": 3.3262,
+      "step": 2617
+    },
+    {
+      "epoch": 0.557169459962756,
+      "grad_norm": 1.051994341539747,
+      "learning_rate": 9.777864028930705e-05,
+      "loss": 3.36,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5573822825219473,
+      "grad_norm": 1.3720510740664986,
+      "learning_rate": 9.777498855305033e-05,
+      "loss": 3.3172,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5575951050811386,
+      "grad_norm": 1.1465874699702439,
+      "learning_rate": 9.777133388598228e-05,
+      "loss": 3.3389,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5578079276403298,
+      "grad_norm": 1.3036200108468314,
+      "learning_rate": 9.776767628832703e-05,
+      "loss": 3.3461,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5580207501995211,
+      "grad_norm": 1.4375770390089653,
+      "learning_rate": 9.7764015760309e-05,
+      "loss": 3.3606,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5582335727587124,
+      "grad_norm": 1.0752162347609548,
+      "learning_rate": 9.776035230215274e-05,
+      "loss": 3.3098,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5584463953179037,
+      "grad_norm": 1.6101533130014802,
+      "learning_rate": 9.775668591408298e-05,
+      "loss": 3.3384,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5586592178770949,
+      "grad_norm": 1.4902748402792567,
+      "learning_rate": 9.775301659632463e-05,
+      "loss": 3.3266,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5588720404362862,
+      "grad_norm": 1.1652012939091514,
+      "learning_rate": 9.774934434910281e-05,
+      "loss": 3.322,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5590848629954775,
+      "grad_norm": 1.347963004806917,
+      "learning_rate": 9.77456691726428e-05,
+      "loss": 3.3596,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5592976855546687,
+      "grad_norm": 1.3319232293336343,
+      "learning_rate": 9.774199106717003e-05,
+      "loss": 3.3726,
+      "step": 2628
+    },
+    {
+      "epoch": 0.55951050811386,
+      "grad_norm": 1.1059679327474303,
+      "learning_rate": 9.773831003291016e-05,
+      "loss": 3.3279,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5597233306730514,
+      "grad_norm": 1.2751672073009692,
+      "learning_rate": 9.773462607008902e-05,
+      "loss": 3.3387,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5599361532322427,
+      "grad_norm": 1.3153517335389122,
+      "learning_rate": 9.773093917893257e-05,
+      "loss": 3.3286,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5601489757914339,
+      "grad_norm": 1.3405443352437296,
+      "learning_rate": 9.772724935966701e-05,
+      "loss": 3.347,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5603617983506252,
+      "grad_norm": 1.3127398599292608,
+      "learning_rate": 9.77235566125187e-05,
+      "loss": 3.3186,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5605746209098165,
+      "grad_norm": 1.1475536637620432,
+      "learning_rate": 9.771986093771416e-05,
+      "loss": 3.3515,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5607874434690077,
+      "grad_norm": 1.4472751998447502,
+      "learning_rate": 9.771616233548014e-05,
+      "loss": 3.3693,
+      "step": 2635
+    },
+    {
+      "epoch": 0.561000266028199,
+      "grad_norm": 1.2341797454888386,
+      "learning_rate": 9.771246080604349e-05,
+      "loss": 3.32,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5612130885873903,
+      "grad_norm": 1.6253321362983213,
+      "learning_rate": 9.770875634963129e-05,
+      "loss": 3.3335,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5614259111465816,
+      "grad_norm": 1.0176458196817895,
+      "learning_rate": 9.770504896647081e-05,
+      "loss": 3.3135,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5616387337057728,
+      "grad_norm": 1.5028131768334632,
+      "learning_rate": 9.770133865678949e-05,
+      "loss": 3.3615,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5618515562649641,
+      "grad_norm": 1.0395934780698228,
+      "learning_rate": 9.769762542081496e-05,
+      "loss": 3.3694,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5620643788241554,
+      "grad_norm": 1.619852889953387,
+      "learning_rate": 9.769390925877495e-05,
+      "loss": 3.3143,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5622772013833466,
+      "grad_norm": 1.1376237005922658,
+      "learning_rate": 9.769019017089748e-05,
+      "loss": 3.3279,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5624900239425379,
+      "grad_norm": 1.384773220732218,
+      "learning_rate": 9.768646815741069e-05,
+      "loss": 3.357,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5627028465017292,
+      "grad_norm": 1.4710952757214941,
+      "learning_rate": 9.768274321854292e-05,
+      "loss": 3.3292,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5629156690609205,
+      "grad_norm": 1.2778967370907435,
+      "learning_rate": 9.767901535452266e-05,
+      "loss": 3.3233,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5631284916201117,
+      "grad_norm": 1.3196815887213982,
+      "learning_rate": 9.767528456557862e-05,
+      "loss": 3.3309,
+      "step": 2646
+    },
+    {
+      "epoch": 0.563341314179303,
+      "grad_norm": 1.405456388242113,
+      "learning_rate": 9.767155085193966e-05,
+      "loss": 3.3005,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5635541367384943,
+      "grad_norm": 1.0910169129874685,
+      "learning_rate": 9.766781421383483e-05,
+      "loss": 3.341,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5637669592976855,
+      "grad_norm": 1.4816398399655601,
+      "learning_rate": 9.766407465149337e-05,
+      "loss": 3.3266,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5639797818568768,
+      "grad_norm": 1.4020013875919493,
+      "learning_rate": 9.766033216514467e-05,
+      "loss": 3.357,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5641926044160681,
+      "grad_norm": 1.229321244881879,
+      "learning_rate": 9.765658675501833e-05,
+      "loss": 3.3654,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5644054269752594,
+      "grad_norm": 1.3377077200447423,
+      "learning_rate": 9.76528384213441e-05,
+      "loss": 3.3455,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5646182495344506,
+      "grad_norm": 1.298706236292527,
+      "learning_rate": 9.764908716435195e-05,
+      "loss": 3.3518,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5648310720936419,
+      "grad_norm": 1.3405249720628005,
+      "learning_rate": 9.764533298427198e-05,
+      "loss": 3.3334,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5650438946528332,
+      "grad_norm": 1.3045556008176291,
+      "learning_rate": 9.764157588133452e-05,
+      "loss": 3.337,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5652567172120245,
+      "grad_norm": 1.4572589864676295,
+      "learning_rate": 9.763781585577003e-05,
+      "loss": 3.3545,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5654695397712157,
+      "grad_norm": 1.1025584781847309,
+      "learning_rate": 9.763405290780918e-05,
+      "loss": 3.317,
+      "step": 2657
+    },
+    {
+      "epoch": 0.565682362330407,
+      "grad_norm": 1.5372109131647462,
+      "learning_rate": 9.763028703768282e-05,
+      "loss": 3.3201,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5658951848895983,
+      "grad_norm": 1.1305584236168322,
+      "learning_rate": 9.762651824562197e-05,
+      "loss": 3.33,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5661080074487895,
+      "grad_norm": 1.5473805043919036,
+      "learning_rate": 9.762274653185784e-05,
+      "loss": 3.3992,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5663208300079808,
+      "grad_norm": 1.0481776763631259,
+      "learning_rate": 9.761897189662178e-05,
+      "loss": 3.3344,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5665336525671721,
+      "grad_norm": 1.4912281615811134,
+      "learning_rate": 9.761519434014537e-05,
+      "loss": 3.2816,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5667464751263634,
+      "grad_norm": 1.009082052791196,
+      "learning_rate": 9.761141386266034e-05,
+      "loss": 3.3361,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5669592976855546,
+      "grad_norm": 1.497059291250652,
+      "learning_rate": 9.760763046439863e-05,
+      "loss": 3.314,
+      "step": 2664
+    },
+    {
+      "epoch": 0.567172120244746,
+      "grad_norm": 1.1002556757755746,
+      "learning_rate": 9.76038441455923e-05,
+      "loss": 3.351,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5673849428039373,
+      "grad_norm": 1.532702986253714,
+      "learning_rate": 9.760005490647366e-05,
+      "loss": 3.3857,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5675977653631284,
+      "grad_norm": 1.2814206546562117,
+      "learning_rate": 9.759626274727516e-05,
+      "loss": 3.3313,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5678105879223198,
+      "grad_norm": 1.056950532183542,
+      "learning_rate": 9.759246766822942e-05,
+      "loss": 3.2824,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5680234104815111,
+      "grad_norm": 1.3760733644657173,
+      "learning_rate": 9.758866966956926e-05,
+      "loss": 3.3106,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5682362330407024,
+      "grad_norm": 1.2060677700760818,
+      "learning_rate": 9.758486875152766e-05,
+      "loss": 3.3325,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5684490555998936,
+      "grad_norm": 1.27126057751213,
+      "learning_rate": 9.758106491433781e-05,
+      "loss": 3.3168,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5686618781590849,
+      "grad_norm": 1.4429447235754722,
+      "learning_rate": 9.757725815823306e-05,
+      "loss": 3.3732,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5688747007182762,
+      "grad_norm": 1.1814303520317468,
+      "learning_rate": 9.757344848344694e-05,
+      "loss": 3.3258,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5690875232774674,
+      "grad_norm": 1.4557927339567385,
+      "learning_rate": 9.756963589021314e-05,
+      "loss": 3.3499,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5693003458366587,
+      "grad_norm": 1.1573054735946582,
+      "learning_rate": 9.756582037876558e-05,
+      "loss": 3.3119,
+      "step": 2675
+    },
+    {
+      "epoch": 0.56951316839585,
+      "grad_norm": 1.3659437579153164,
+      "learning_rate": 9.756200194933829e-05,
+      "loss": 3.3342,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5697259909550413,
+      "grad_norm": 1.1648019147234303,
+      "learning_rate": 9.755818060216554e-05,
+      "loss": 3.3205,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5699388135142325,
+      "grad_norm": 1.4523232492331137,
+      "learning_rate": 9.755435633748175e-05,
+      "loss": 3.3561,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5701516360734238,
+      "grad_norm": 1.226913514846268,
+      "learning_rate": 9.755052915552152e-05,
+      "loss": 3.3572,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5703644586326151,
+      "grad_norm": 1.272151470633851,
+      "learning_rate": 9.754669905651962e-05,
+      "loss": 3.3498,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5705772811918063,
+      "grad_norm": 1.2804334018302206,
+      "learning_rate": 9.754286604071105e-05,
+      "loss": 3.2964,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5707901037509976,
+      "grad_norm": 1.279110839174267,
+      "learning_rate": 9.753903010833094e-05,
+      "loss": 3.294,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5710029263101889,
+      "grad_norm": 1.3358141454726498,
+      "learning_rate": 9.753519125961458e-05,
+      "loss": 3.3438,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5712157488693802,
+      "grad_norm": 1.2073721069005876,
+      "learning_rate": 9.75313494947975e-05,
+      "loss": 3.3539,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.2351129143894577,
+      "learning_rate": 9.752750481411534e-05,
+      "loss": 3.3498,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5716413939877627,
+      "grad_norm": 1.3647266541221887,
+      "learning_rate": 9.7523657217804e-05,
+      "loss": 3.388,
+      "step": 2686
+    },
+    {
+      "epoch": 0.571854216546954,
+      "grad_norm": 1.2636649786270606,
+      "learning_rate": 9.751980670609948e-05,
+      "loss": 3.3516,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5720670391061452,
+      "grad_norm": 1.3944530000513384,
+      "learning_rate": 9.751595327923802e-05,
+      "loss": 3.3041,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5722798616653365,
+      "grad_norm": 1.103846258175167,
+      "learning_rate": 9.751209693745601e-05,
+      "loss": 3.3538,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5724926842245278,
+      "grad_norm": 1.471698609889461,
+      "learning_rate": 9.750823768099e-05,
+      "loss": 3.3216,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5727055067837191,
+      "grad_norm": 1.1770049401640155,
+      "learning_rate": 9.750437551007676e-05,
+      "loss": 3.3232,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5729183293429103,
+      "grad_norm": 1.4818073104008807,
+      "learning_rate": 9.750051042495321e-05,
+      "loss": 3.4063,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5731311519021016,
+      "grad_norm": 1.1713519373028276,
+      "learning_rate": 9.749664242585648e-05,
+      "loss": 3.3217,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5733439744612929,
+      "grad_norm": 1.3100852178776679,
+      "learning_rate": 9.749277151302383e-05,
+      "loss": 3.3274,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5735567970204841,
+      "grad_norm": 1.3381474544891272,
+      "learning_rate": 9.748889768669272e-05,
+      "loss": 3.3176,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5737696195796754,
+      "grad_norm": 1.1609061018598328,
+      "learning_rate": 9.748502094710083e-05,
+      "loss": 3.3022,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5739824421388667,
+      "grad_norm": 1.438066406024806,
+      "learning_rate": 9.748114129448594e-05,
+      "loss": 3.3538,
+      "step": 2697
+    },
+    {
+      "epoch": 0.574195264698058,
+      "grad_norm": 1.1940404419442219,
+      "learning_rate": 9.747725872908609e-05,
+      "loss": 3.3362,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5744080872572492,
+      "grad_norm": 1.5955845936315096,
+      "learning_rate": 9.747337325113945e-05,
+      "loss": 3.3235,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5746209098164405,
+      "grad_norm": 1.04932901222465,
+      "learning_rate": 9.746948486088435e-05,
+      "loss": 3.3024,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5748337323756318,
+      "grad_norm": 1.5256364877687412,
+      "learning_rate": 9.746559355855936e-05,
+      "loss": 3.361,
+      "step": 2701
+    },
+    {
+      "epoch": 0.575046554934823,
+      "grad_norm": 1.237494925108093,
+      "learning_rate": 9.74616993444032e-05,
+      "loss": 3.3104,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5752593774940143,
+      "grad_norm": 1.3785754614089598,
+      "learning_rate": 9.745780221865474e-05,
+      "loss": 3.3457,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5754722000532057,
+      "grad_norm": 1.1497297551215957,
+      "learning_rate": 9.745390218155307e-05,
+      "loss": 3.3033,
+      "step": 2704
+    },
+    {
+      "epoch": 0.575685022612397,
+      "grad_norm": 1.3274481296669083,
+      "learning_rate": 9.744999923333744e-05,
+      "loss": 3.2759,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5758978451715882,
+      "grad_norm": 1.2376340974615407,
+      "learning_rate": 9.744609337424727e-05,
+      "loss": 3.3494,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5761106677307795,
+      "grad_norm": 1.209517175364465,
+      "learning_rate": 9.744218460452219e-05,
+      "loss": 3.2984,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5763234902899708,
+      "grad_norm": 1.3443548360042041,
+      "learning_rate": 9.743827292440198e-05,
+      "loss": 3.3051,
+      "step": 2708
+    },
+    {
+      "epoch": 0.576536312849162,
+      "grad_norm": 1.1567369314049705,
+      "learning_rate": 9.743435833412659e-05,
+      "loss": 3.3143,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5767491354083533,
+      "grad_norm": 1.3964232572304063,
+      "learning_rate": 9.743044083393618e-05,
+      "loss": 3.3219,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5769619579675446,
+      "grad_norm": 1.058415359974026,
+      "learning_rate": 9.742652042407108e-05,
+      "loss": 3.2989,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5771747805267359,
+      "grad_norm": 1.328750100810777,
+      "learning_rate": 9.742259710477177e-05,
+      "loss": 3.3538,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5773876030859271,
+      "grad_norm": 1.2947829694967063,
+      "learning_rate": 9.741867087627895e-05,
+      "loss": 3.2943,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5776004256451184,
+      "grad_norm": 1.3775349729622817,
+      "learning_rate": 9.741474173883347e-05,
+      "loss": 3.3048,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5778132482043097,
+      "grad_norm": 1.2866940689635682,
+      "learning_rate": 9.741080969267637e-05,
+      "loss": 3.3198,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5780260707635009,
+      "grad_norm": 1.2264558128851306,
+      "learning_rate": 9.740687473804888e-05,
+      "loss": 3.3294,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5782388933226922,
+      "grad_norm": 1.5521772048409994,
+      "learning_rate": 9.740293687519238e-05,
+      "loss": 3.331,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5784517158818835,
+      "grad_norm": 1.0309009044462807,
+      "learning_rate": 9.739899610434841e-05,
+      "loss": 3.3459,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5786645384410748,
+      "grad_norm": 1.5851787727472273,
+      "learning_rate": 9.739505242575879e-05,
+      "loss": 3.2913,
+      "step": 2719
+    },
+    {
+      "epoch": 0.578877361000266,
+      "grad_norm": 1.0121693660518165,
+      "learning_rate": 9.73911058396654e-05,
+      "loss": 3.367,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5790901835594573,
+      "grad_norm": 1.4926001653573142,
+      "learning_rate": 9.738715634631038e-05,
+      "loss": 3.2959,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5793030061186486,
+      "grad_norm": 1.2496018948012433,
+      "learning_rate": 9.738320394593596e-05,
+      "loss": 3.2985,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5795158286778399,
+      "grad_norm": 1.1707127311224295,
+      "learning_rate": 9.737924863878467e-05,
+      "loss": 3.2929,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5797286512370311,
+      "grad_norm": 1.1785210336049252,
+      "learning_rate": 9.737529042509913e-05,
+      "loss": 3.3463,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5799414737962224,
+      "grad_norm": 1.2905333158060193,
+      "learning_rate": 9.737132930512214e-05,
+      "loss": 3.2953,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5801542963554137,
+      "grad_norm": 1.4090816950843292,
+      "learning_rate": 9.736736527909673e-05,
+      "loss": 3.3413,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5803671189146049,
+      "grad_norm": 1.1441311193933996,
+      "learning_rate": 9.736339834726604e-05,
+      "loss": 3.3202,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5805799414737962,
+      "grad_norm": 1.3714563428722255,
+      "learning_rate": 9.735942850987347e-05,
+      "loss": 3.3263,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5807927640329875,
+      "grad_norm": 1.1555650738309957,
+      "learning_rate": 9.735545576716254e-05,
+      "loss": 3.3049,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5810055865921788,
+      "grad_norm": 1.4577870717608326,
+      "learning_rate": 9.735148011937694e-05,
+      "loss": 3.2919,
+      "step": 2730
+    },
+    {
+      "epoch": 0.58121840915137,
+      "grad_norm": 1.1461181877300672,
+      "learning_rate": 9.734750156676058e-05,
+      "loss": 3.3316,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5814312317105613,
+      "grad_norm": 1.4759932010981662,
+      "learning_rate": 9.734352010955753e-05,
+      "loss": 3.2742,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5816440542697526,
+      "grad_norm": 1.1586380007737767,
+      "learning_rate": 9.733953574801203e-05,
+      "loss": 3.3374,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5818568768289438,
+      "grad_norm": 1.2165022084744532,
+      "learning_rate": 9.733554848236852e-05,
+      "loss": 3.3534,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5820696993881351,
+      "grad_norm": 1.3075415296372201,
+      "learning_rate": 9.733155831287157e-05,
+      "loss": 3.3571,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5822825219473264,
+      "grad_norm": 1.4482440599629365,
+      "learning_rate": 9.7327565239766e-05,
+      "loss": 3.353,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5824953445065177,
+      "grad_norm": 1.2893977328272108,
+      "learning_rate": 9.732356926329674e-05,
+      "loss": 3.3123,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5827081670657089,
+      "grad_norm": 1.2328499275213556,
+      "learning_rate": 9.731957038370895e-05,
+      "loss": 3.2831,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5829209896249002,
+      "grad_norm": 1.2890750061177723,
+      "learning_rate": 9.731556860124793e-05,
+      "loss": 3.3094,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5831338121840915,
+      "grad_norm": 1.2762413711294158,
+      "learning_rate": 9.731156391615919e-05,
+      "loss": 3.3166,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5833466347432827,
+      "grad_norm": 1.3037725096233237,
+      "learning_rate": 9.730755632868838e-05,
+      "loss": 3.2961,
+      "step": 2741
+    },
+    {
+      "epoch": 0.583559457302474,
+      "grad_norm": 1.267303339027181,
+      "learning_rate": 9.730354583908137e-05,
+      "loss": 3.3299,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5837722798616654,
+      "grad_norm": 1.1712812098399616,
+      "learning_rate": 9.729953244758417e-05,
+      "loss": 3.3337,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5839851024208567,
+      "grad_norm": 1.2669103857771027,
+      "learning_rate": 9.729551615444301e-05,
+      "loss": 3.3193,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5841979249800479,
+      "grad_norm": 1.2654056386561516,
+      "learning_rate": 9.729149695990425e-05,
+      "loss": 3.3122,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5844107475392392,
+      "grad_norm": 1.1904075726824674,
+      "learning_rate": 9.728747486421447e-05,
+      "loss": 3.3786,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5846235700984305,
+      "grad_norm": 1.4040758515345306,
+      "learning_rate": 9.72834498676204e-05,
+      "loss": 3.305,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5848363926576217,
+      "grad_norm": 1.2642460181274235,
+      "learning_rate": 9.727942197036895e-05,
+      "loss": 3.2927,
+      "step": 2748
+    },
+    {
+      "epoch": 0.585049215216813,
+      "grad_norm": 1.2462427122986766,
+      "learning_rate": 9.727539117270724e-05,
+      "loss": 3.3192,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5852620377760043,
+      "grad_norm": 1.292192761541366,
+      "learning_rate": 9.727135747488252e-05,
+      "loss": 3.3248,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5854748603351956,
+      "grad_norm": 1.2972205009796784,
+      "learning_rate": 9.726732087714228e-05,
+      "loss": 3.3427,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5856876828943868,
+      "grad_norm": 1.2241045056359883,
+      "learning_rate": 9.726328137973408e-05,
+      "loss": 3.2919,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5859005054535781,
+      "grad_norm": 1.4006946004180176,
+      "learning_rate": 9.72592389829058e-05,
+      "loss": 3.3064,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5861133280127694,
+      "grad_norm": 1.1396434296655715,
+      "learning_rate": 9.725519368690538e-05,
+      "loss": 3.2778,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5863261505719606,
+      "grad_norm": 1.4227667047932335,
+      "learning_rate": 9.7251145491981e-05,
+      "loss": 3.3638,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5865389731311519,
+      "grad_norm": 1.0201798524036818,
+      "learning_rate": 9.724709439838101e-05,
+      "loss": 3.2761,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5867517956903432,
+      "grad_norm": 1.3817839514889934,
+      "learning_rate": 9.724304040635391e-05,
+      "loss": 3.2859,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5869646182495345,
+      "grad_norm": 1.1596055116263035,
+      "learning_rate": 9.723898351614841e-05,
+      "loss": 3.3085,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5871774408087257,
+      "grad_norm": 1.3931684268333335,
+      "learning_rate": 9.723492372801338e-05,
+      "loss": 3.2911,
+      "step": 2759
+    },
+    {
+      "epoch": 0.587390263367917,
+      "grad_norm": 1.3559318724429998,
+      "learning_rate": 9.723086104219787e-05,
+      "loss": 3.3244,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5876030859271083,
+      "grad_norm": 1.1791948614085956,
+      "learning_rate": 9.72267954589511e-05,
+      "loss": 3.3094,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5878159084862995,
+      "grad_norm": 1.3780719060923297,
+      "learning_rate": 9.72227269785225e-05,
+      "loss": 3.2795,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5880287310454908,
+      "grad_norm": 1.2744115034087986,
+      "learning_rate": 9.721865560116166e-05,
+      "loss": 3.3365,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5882415536046821,
+      "grad_norm": 1.4878718465093637,
+      "learning_rate": 9.721458132711831e-05,
+      "loss": 3.3268,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5884543761638734,
+      "grad_norm": 1.2576679479851984,
+      "learning_rate": 9.721050415664242e-05,
+      "loss": 3.2936,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5886671987230646,
+      "grad_norm": 1.3520809404095748,
+      "learning_rate": 9.72064240899841e-05,
+      "loss": 3.3212,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5888800212822559,
+      "grad_norm": 1.3001264673059092,
+      "learning_rate": 9.720234112739363e-05,
+      "loss": 3.3184,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5890928438414472,
+      "grad_norm": 1.4689257189979403,
+      "learning_rate": 9.719825526912151e-05,
+      "loss": 3.3234,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5893056664006384,
+      "grad_norm": 1.1849053374678618,
+      "learning_rate": 9.719416651541839e-05,
+      "loss": 3.3048,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5895184889598297,
+      "grad_norm": 1.2322591299820023,
+      "learning_rate": 9.719007486653507e-05,
+      "loss": 3.3081,
+      "step": 2770
+    },
+    {
+      "epoch": 0.589731311519021,
+      "grad_norm": 1.2929983303716668,
+      "learning_rate": 9.718598032272259e-05,
+      "loss": 3.2888,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5899441340782123,
+      "grad_norm": 1.488988192857633,
+      "learning_rate": 9.718188288423212e-05,
+      "loss": 3.2763,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5901569566374035,
+      "grad_norm": 1.2571262125205782,
+      "learning_rate": 9.717778255131502e-05,
+      "loss": 3.3105,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5903697791965948,
+      "grad_norm": 1.163242341184613,
+      "learning_rate": 9.717367932422284e-05,
+      "loss": 3.3089,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5905826017557861,
+      "grad_norm": 1.3518792997243338,
+      "learning_rate": 9.716957320320727e-05,
+      "loss": 3.3163,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5907954243149773,
+      "grad_norm": 1.048652293988682,
+      "learning_rate": 9.716546418852025e-05,
+      "loss": 3.2895,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5910082468741686,
+      "grad_norm": 1.544623994552474,
+      "learning_rate": 9.716135228041382e-05,
+      "loss": 3.3214,
+      "step": 2777
+    },
+    {
+      "epoch": 0.59122106943336,
+      "grad_norm": 0.9767190988470699,
+      "learning_rate": 9.715723747914023e-05,
+      "loss": 3.2697,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5914338919925513,
+      "grad_norm": 1.4012386988203958,
+      "learning_rate": 9.715311978495192e-05,
+      "loss": 3.3465,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5916467145517424,
+      "grad_norm": 1.2410990542125873,
+      "learning_rate": 9.714899919810148e-05,
+      "loss": 3.2856,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5918595371109338,
+      "grad_norm": 1.3937488636468889,
+      "learning_rate": 9.714487571884172e-05,
+      "loss": 3.2984,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5920723596701251,
+      "grad_norm": 1.0996828254866602,
+      "learning_rate": 9.714074934742556e-05,
+      "loss": 3.3294,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5922851822293163,
+      "grad_norm": 1.4370727005279142,
+      "learning_rate": 9.713662008410617e-05,
+      "loss": 3.308,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5924980047885076,
+      "grad_norm": 1.2064310100398832,
+      "learning_rate": 9.713248792913685e-05,
+      "loss": 3.3071,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5927108273476989,
+      "grad_norm": 1.3743412831429378,
+      "learning_rate": 9.71283528827711e-05,
+      "loss": 3.3381,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5929236499068902,
+      "grad_norm": 1.307302177334519,
+      "learning_rate": 9.712421494526256e-05,
+      "loss": 3.2685,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5931364724660814,
+      "grad_norm": 1.3609912176959815,
+      "learning_rate": 9.712007411686512e-05,
+      "loss": 3.2703,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5933492950252727,
+      "grad_norm": 1.188220670576602,
+      "learning_rate": 9.711593039783277e-05,
+      "loss": 3.2689,
+      "step": 2788
+    },
+    {
+      "epoch": 0.593562117584464,
+      "grad_norm": 1.3420322813690249,
+      "learning_rate": 9.711178378841973e-05,
+      "loss": 3.3245,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5937749401436552,
+      "grad_norm": 1.2866907423978184,
+      "learning_rate": 9.710763428888036e-05,
+      "loss": 3.2615,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5939877627028465,
+      "grad_norm": 1.0135734362041744,
+      "learning_rate": 9.710348189946924e-05,
+      "loss": 3.2995,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5942005852620378,
+      "grad_norm": 1.5677910116281049,
+      "learning_rate": 9.70993266204411e-05,
+      "loss": 3.3284,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5944134078212291,
+      "grad_norm": 1.0471746883487618,
+      "learning_rate": 9.709516845205084e-05,
+      "loss": 3.2732,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5946262303804203,
+      "grad_norm": 1.5668232646990348,
+      "learning_rate": 9.709100739455355e-05,
+      "loss": 3.2795,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5948390529396116,
+      "grad_norm": 0.9925622857280059,
+      "learning_rate": 9.70868434482045e-05,
+      "loss": 3.2759,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5950518754988029,
+      "grad_norm": 1.3563266013117585,
+      "learning_rate": 9.708267661325909e-05,
+      "loss": 3.3255,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5952646980579942,
+      "grad_norm": 1.3610067486334974,
+      "learning_rate": 9.707850688997301e-05,
+      "loss": 3.3177,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5954775206171854,
+      "grad_norm": 1.1138363620891463,
+      "learning_rate": 9.707433427860202e-05,
+      "loss": 3.2928,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5956903431763767,
+      "grad_norm": 1.2679045773522784,
+      "learning_rate": 9.707015877940209e-05,
+      "loss": 3.3291,
+      "step": 2799
+    },
+    {
+      "epoch": 0.595903165735568,
+      "grad_norm": 1.2976349448629891,
+      "learning_rate": 9.706598039262938e-05,
+      "loss": 3.291,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5961159882947592,
+      "grad_norm": 1.1084431045565688,
+      "learning_rate": 9.706179911854019e-05,
+      "loss": 3.2805,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5963288108539505,
+      "grad_norm": 1.4145457884076456,
+      "learning_rate": 9.705761495739107e-05,
+      "loss": 3.3191,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5965416334131418,
+      "grad_norm": 1.1919341092324285,
+      "learning_rate": 9.70534279094387e-05,
+      "loss": 3.2971,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5967544559723331,
+      "grad_norm": 1.3268664014662748,
+      "learning_rate": 9.70492379749399e-05,
+      "loss": 3.3089,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5969672785315243,
+      "grad_norm": 1.2730941998852532,
+      "learning_rate": 9.704504515415173e-05,
+      "loss": 3.2553,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5971801010907156,
+      "grad_norm": 1.300442746760671,
+      "learning_rate": 9.70408494473314e-05,
+      "loss": 3.2561,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5973929236499069,
+      "grad_norm": 1.0447255188705677,
+      "learning_rate": 9.70366508547363e-05,
+      "loss": 3.2859,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5976057462090981,
+      "grad_norm": 1.2666116550000022,
+      "learning_rate": 9.7032449376624e-05,
+      "loss": 3.2463,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5978185687682894,
+      "grad_norm": 1.5490953644350063,
+      "learning_rate": 9.702824501325223e-05,
+      "loss": 3.3245,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5980313913274807,
+      "grad_norm": 1.0290957468152133,
+      "learning_rate": 9.702403776487895e-05,
+      "loss": 3.2644,
+      "step": 2810
+    },
+    {
+      "epoch": 0.598244213886672,
+      "grad_norm": 1.5396265421343114,
+      "learning_rate": 9.701982763176222e-05,
+      "loss": 3.3409,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5984570364458632,
+      "grad_norm": 1.072078123480471,
+      "learning_rate": 9.701561461416034e-05,
+      "loss": 3.3129,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5986698590050545,
+      "grad_norm": 1.5263298369421288,
+      "learning_rate": 9.701139871233173e-05,
+      "loss": 3.2738,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5988826815642458,
+      "grad_norm": 1.049199019810681,
+      "learning_rate": 9.700717992653506e-05,
+      "loss": 3.2658,
+      "step": 2814
+    },
+    {
+      "epoch": 0.599095504123437,
+      "grad_norm": 1.6373340732011137,
+      "learning_rate": 9.70029582570291e-05,
+      "loss": 3.2847,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5993083266826283,
+      "grad_norm": 1.260463106031657,
+      "learning_rate": 9.699873370407285e-05,
+      "loss": 3.2702,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5995211492418197,
+      "grad_norm": 1.1632065919006886,
+      "learning_rate": 9.699450626792548e-05,
+      "loss": 3.2982,
+      "step": 2817
+    },
+    {
+      "epoch": 0.599733971801011,
+      "grad_norm": 1.405904310400256,
+      "learning_rate": 9.699027594884631e-05,
+      "loss": 3.2742,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5999467943602022,
+      "grad_norm": 1.1321853623895546,
+      "learning_rate": 9.698604274709487e-05,
+      "loss": 3.3277,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6001596169193935,
+      "grad_norm": 1.3618217923561233,
+      "learning_rate": 9.698180666293083e-05,
+      "loss": 3.3013,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6003724394785848,
+      "grad_norm": 1.2350271264263257,
+      "learning_rate": 9.697756769661409e-05,
+      "loss": 3.2819,
+      "step": 2821
+    },
+    {
+      "epoch": 0.600585262037776,
+      "grad_norm": 1.115349668348199,
+      "learning_rate": 9.697332584840464e-05,
+      "loss": 3.305,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6007980845969673,
+      "grad_norm": 1.3467241652146893,
+      "learning_rate": 9.696908111856275e-05,
+      "loss": 3.2691,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6010109071561586,
+      "grad_norm": 1.1987184187718107,
+      "learning_rate": 9.69648335073488e-05,
+      "loss": 3.261,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6012237297153499,
+      "grad_norm": 1.2069159606241695,
+      "learning_rate": 9.696058301502335e-05,
+      "loss": 3.2717,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6014365522745411,
+      "grad_norm": 1.3102960656990328,
+      "learning_rate": 9.69563296418472e-05,
+      "loss": 3.311,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6016493748337324,
+      "grad_norm": 1.3213831125677296,
+      "learning_rate": 9.695207338808123e-05,
+      "loss": 3.2775,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6018621973929237,
+      "grad_norm": 1.1430943242181135,
+      "learning_rate": 9.694781425398656e-05,
+      "loss": 3.286,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6020750199521149,
+      "grad_norm": 1.3777910284655743,
+      "learning_rate": 9.694355223982446e-05,
+      "loss": 3.2946,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6022878425113062,
+      "grad_norm": 1.0987543437603886,
+      "learning_rate": 9.69392873458564e-05,
+      "loss": 3.2776,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6025006650704975,
+      "grad_norm": 1.2953843585462737,
+      "learning_rate": 9.693501957234404e-05,
+      "loss": 3.2792,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6027134876296888,
+      "grad_norm": 1.1099482026960896,
+      "learning_rate": 9.693074891954915e-05,
+      "loss": 3.2758,
+      "step": 2832
+    },
+    {
+      "epoch": 0.60292631018888,
+      "grad_norm": 1.533575500444798,
+      "learning_rate": 9.692647538773374e-05,
+      "loss": 3.3069,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6031391327480713,
+      "grad_norm": 0.8701264786687234,
+      "learning_rate": 9.692219897715995e-05,
+      "loss": 3.2731,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6033519553072626,
+      "grad_norm": 1.4105937164804092,
+      "learning_rate": 9.691791968809016e-05,
+      "loss": 3.2701,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6035647778664538,
+      "grad_norm": 1.1910352396010373,
+      "learning_rate": 9.691363752078687e-05,
+      "loss": 3.293,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6037776004256451,
+      "grad_norm": 1.2165611990875083,
+      "learning_rate": 9.690935247551276e-05,
+      "loss": 3.2899,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6039904229848364,
+      "grad_norm": 1.2295487930126834,
+      "learning_rate": 9.690506455253073e-05,
+      "loss": 3.2679,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6042032455440277,
+      "grad_norm": 1.3658179546083165,
+      "learning_rate": 9.690077375210378e-05,
+      "loss": 3.2857,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6044160681032189,
+      "grad_norm": 1.1634297983802309,
+      "learning_rate": 9.689648007449519e-05,
+      "loss": 3.2919,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6046288906624102,
+      "grad_norm": 1.313584144397816,
+      "learning_rate": 9.689218351996833e-05,
+      "loss": 3.2644,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6048417132216015,
+      "grad_norm": 1.2035339674674133,
+      "learning_rate": 9.688788408878677e-05,
+      "loss": 3.2879,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6050545357807927,
+      "grad_norm": 1.3072350923036788,
+      "learning_rate": 9.68835817812143e-05,
+      "loss": 3.3189,
+      "step": 2843
+    },
+    {
+      "epoch": 0.605267358339984,
+      "grad_norm": 1.1495977156876342,
+      "learning_rate": 9.687927659751481e-05,
+      "loss": 3.2105,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6054801808991753,
+      "grad_norm": 1.4073164474933295,
+      "learning_rate": 9.687496853795243e-05,
+      "loss": 3.3109,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6056930034583666,
+      "grad_norm": 0.949878364472448,
+      "learning_rate": 9.687065760279142e-05,
+      "loss": 3.2871,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6059058260175578,
+      "grad_norm": 1.3492459124631564,
+      "learning_rate": 9.686634379229627e-05,
+      "loss": 3.2306,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6061186485767491,
+      "grad_norm": 1.1388574668165725,
+      "learning_rate": 9.68620271067316e-05,
+      "loss": 3.3252,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6063314711359404,
+      "grad_norm": 1.4810889164839467,
+      "learning_rate": 9.685770754636222e-05,
+      "loss": 3.2886,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6065442936951316,
+      "grad_norm": 1.103488246440217,
+      "learning_rate": 9.685338511145312e-05,
+      "loss": 3.2532,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6067571162543229,
+      "grad_norm": 1.2655698202374694,
+      "learning_rate": 9.684905980226945e-05,
+      "loss": 3.3028,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6069699388135142,
+      "grad_norm": 1.3409971386566881,
+      "learning_rate": 9.68447316190766e-05,
+      "loss": 3.2779,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6071827613727055,
+      "grad_norm": 1.1680997342888684,
+      "learning_rate": 9.684040056214003e-05,
+      "loss": 3.2365,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6073955839318967,
+      "grad_norm": 1.3438169414835446,
+      "learning_rate": 9.683606663172547e-05,
+      "loss": 3.2511,
+      "step": 2854
+    },
+    {
+      "epoch": 0.607608406491088,
+      "grad_norm": 1.2616218227865459,
+      "learning_rate": 9.683172982809878e-05,
+      "loss": 3.2768,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6078212290502794,
+      "grad_norm": 1.3698804865166676,
+      "learning_rate": 9.682739015152598e-05,
+      "loss": 3.2937,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6080340516094705,
+      "grad_norm": 1.1421169936558837,
+      "learning_rate": 9.682304760227334e-05,
+      "loss": 3.2922,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6082468741686619,
+      "grad_norm": 1.474044631937744,
+      "learning_rate": 9.681870218060722e-05,
+      "loss": 3.3063,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6084596967278532,
+      "grad_norm": 1.030026202769578,
+      "learning_rate": 9.68143538867942e-05,
+      "loss": 3.2921,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6086725192870445,
+      "grad_norm": 1.529645637270696,
+      "learning_rate": 9.681000272110107e-05,
+      "loss": 3.2685,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6088853418462357,
+      "grad_norm": 1.2268437554539893,
+      "learning_rate": 9.680564868379471e-05,
+      "loss": 3.269,
+      "step": 2861
+    },
+    {
+      "epoch": 0.609098164405427,
+      "grad_norm": 1.2015747510339958,
+      "learning_rate": 9.680129177514225e-05,
+      "loss": 3.2629,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6093109869646183,
+      "grad_norm": 1.2685174308371503,
+      "learning_rate": 9.679693199541095e-05,
+      "loss": 3.2546,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 1.1496711907969077,
+      "learning_rate": 9.679256934486829e-05,
+      "loss": 3.2866,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6097366320830008,
+      "grad_norm": 1.3959080330277183,
+      "learning_rate": 9.678820382378188e-05,
+      "loss": 3.2414,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6099494546421921,
+      "grad_norm": 0.9947410541789244,
+      "learning_rate": 9.678383543241953e-05,
+      "loss": 3.3004,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6101622772013834,
+      "grad_norm": 1.3792824694895467,
+      "learning_rate": 9.677946417104924e-05,
+      "loss": 3.2382,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6103750997605746,
+      "grad_norm": 1.2544603357022954,
+      "learning_rate": 9.677509003993915e-05,
+      "loss": 3.2851,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6105879223197659,
+      "grad_norm": 1.2908152933223664,
+      "learning_rate": 9.677071303935762e-05,
+      "loss": 3.2818,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6108007448789572,
+      "grad_norm": 1.1112943832334852,
+      "learning_rate": 9.676633316957314e-05,
+      "loss": 3.2111,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6110135674381485,
+      "grad_norm": 1.1601735186083713,
+      "learning_rate": 9.676195043085441e-05,
+      "loss": 3.3,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6112263899973397,
+      "grad_norm": 1.2773758466864935,
+      "learning_rate": 9.67575648234703e-05,
+      "loss": 3.2648,
+      "step": 2872
+    },
+    {
+      "epoch": 0.611439212556531,
+      "grad_norm": 1.2807428892006547,
+      "learning_rate": 9.675317634768981e-05,
+      "loss": 3.2937,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6116520351157223,
+      "grad_norm": 1.185384045043331,
+      "learning_rate": 9.674878500378222e-05,
+      "loss": 3.2681,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6118648576749135,
+      "grad_norm": 1.4459300882899668,
+      "learning_rate": 9.674439079201686e-05,
+      "loss": 3.2978,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6120776802341048,
+      "grad_norm": 1.1829080139499237,
+      "learning_rate": 9.673999371266334e-05,
+      "loss": 3.2962,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6122905027932961,
+      "grad_norm": 1.3026143526817717,
+      "learning_rate": 9.67355937659914e-05,
+      "loss": 3.2654,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6125033253524874,
+      "grad_norm": 1.011304740493569,
+      "learning_rate": 9.673119095227094e-05,
+      "loss": 3.2452,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6127161479116786,
+      "grad_norm": 1.4233783538552618,
+      "learning_rate": 9.672678527177207e-05,
+      "loss": 3.2454,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6129289704708699,
+      "grad_norm": 1.110979312047089,
+      "learning_rate": 9.672237672476506e-05,
+      "loss": 3.2772,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6131417930300612,
+      "grad_norm": 1.4637766979951554,
+      "learning_rate": 9.671796531152035e-05,
+      "loss": 3.2643,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6133546155892524,
+      "grad_norm": 1.1052530039818893,
+      "learning_rate": 9.671355103230858e-05,
+      "loss": 3.2622,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6135674381484437,
+      "grad_norm": 1.4101053165869724,
+      "learning_rate": 9.670913388740051e-05,
+      "loss": 3.274,
+      "step": 2883
+    },
+    {
+      "epoch": 0.613780260707635,
+      "grad_norm": 1.1880236611696087,
+      "learning_rate": 9.670471387706716e-05,
+      "loss": 3.2941,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6139930832668263,
+      "grad_norm": 1.3039721933826196,
+      "learning_rate": 9.670029100157966e-05,
+      "loss": 3.2545,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6142059058260175,
+      "grad_norm": 1.1273996908350197,
+      "learning_rate": 9.669586526120936e-05,
+      "loss": 3.2611,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6144187283852088,
+      "grad_norm": 1.3480614691903388,
+      "learning_rate": 9.669143665622771e-05,
+      "loss": 3.245,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6146315509444001,
+      "grad_norm": 1.217545204019635,
+      "learning_rate": 9.668700518690644e-05,
+      "loss": 3.2785,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6148443735035913,
+      "grad_norm": 1.0837411686400509,
+      "learning_rate": 9.668257085351735e-05,
+      "loss": 3.2546,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6150571960627826,
+      "grad_norm": 1.5064200142291606,
+      "learning_rate": 9.667813365633254e-05,
+      "loss": 3.2794,
+      "step": 2890
+    },
+    {
+      "epoch": 0.615270018621974,
+      "grad_norm": 1.0645415330497414,
+      "learning_rate": 9.667369359562415e-05,
+      "loss": 3.2177,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6154828411811653,
+      "grad_norm": 1.3522018350190164,
+      "learning_rate": 9.66692506716646e-05,
+      "loss": 3.2332,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6156956637403564,
+      "grad_norm": 1.1270703637080335,
+      "learning_rate": 9.666480488472642e-05,
+      "loss": 3.2619,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6159084862995478,
+      "grad_norm": 1.2084449048097519,
+      "learning_rate": 9.666035623508238e-05,
+      "loss": 3.2686,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6161213088587391,
+      "grad_norm": 1.4840692987474533,
+      "learning_rate": 9.665590472300534e-05,
+      "loss": 3.2525,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6163341314179303,
+      "grad_norm": 1.0652135313815956,
+      "learning_rate": 9.665145034876842e-05,
+      "loss": 3.2763,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6165469539771216,
+      "grad_norm": 1.6440622492737216,
+      "learning_rate": 9.664699311264486e-05,
+      "loss": 3.2872,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6167597765363129,
+      "grad_norm": 0.9512539360243785,
+      "learning_rate": 9.66425330149081e-05,
+      "loss": 3.2819,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6169725990955042,
+      "grad_norm": 1.5572019650502087,
+      "learning_rate": 9.663807005583175e-05,
+      "loss": 3.2825,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6171854216546954,
+      "grad_norm": 1.0827122627701184,
+      "learning_rate": 9.66336042356896e-05,
+      "loss": 3.2616,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6173982442138867,
+      "grad_norm": 1.330063102080298,
+      "learning_rate": 9.662913555475558e-05,
+      "loss": 3.2804,
+      "step": 2901
+    },
+    {
+      "epoch": 0.617611066773078,
+      "grad_norm": 1.1450922390312097,
+      "learning_rate": 9.662466401330388e-05,
+      "loss": 3.2716,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6178238893322692,
+      "grad_norm": 1.3308970233312634,
+      "learning_rate": 9.662018961160877e-05,
+      "loss": 3.2585,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6180367118914605,
+      "grad_norm": 1.2604786913709765,
+      "learning_rate": 9.661571234994475e-05,
+      "loss": 3.2956,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6182495344506518,
+      "grad_norm": 1.2445883158283428,
+      "learning_rate": 9.661123222858647e-05,
+      "loss": 3.2876,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6184623570098431,
+      "grad_norm": 1.377308503126674,
+      "learning_rate": 9.660674924780882e-05,
+      "loss": 3.3122,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6186751795690343,
+      "grad_norm": 1.1442840119461122,
+      "learning_rate": 9.660226340788674e-05,
+      "loss": 3.2791,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6188880021282256,
+      "grad_norm": 1.2190247711326363,
+      "learning_rate": 9.659777470909546e-05,
+      "loss": 3.2746,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6191008246874169,
+      "grad_norm": 1.3528377532268014,
+      "learning_rate": 9.659328315171035e-05,
+      "loss": 3.2566,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6193136472466081,
+      "grad_norm": 1.1628491320701475,
+      "learning_rate": 9.658878873600692e-05,
+      "loss": 3.2247,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6195264698057994,
+      "grad_norm": 1.3031103378942632,
+      "learning_rate": 9.658429146226092e-05,
+      "loss": 3.2812,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6197392923649907,
+      "grad_norm": 1.351061321703247,
+      "learning_rate": 9.657979133074821e-05,
+      "loss": 3.2704,
+      "step": 2912
+    },
+    {
+      "epoch": 0.619952114924182,
+      "grad_norm": 1.1320175052981598,
+      "learning_rate": 9.657528834174488e-05,
+      "loss": 3.2765,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6201649374833732,
+      "grad_norm": 1.2295160929511915,
+      "learning_rate": 9.657078249552716e-05,
+      "loss": 3.2549,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6203777600425645,
+      "grad_norm": 1.2676511784486362,
+      "learning_rate": 9.656627379237146e-05,
+      "loss": 3.2445,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6205905826017558,
+      "grad_norm": 1.1767756249316326,
+      "learning_rate": 9.656176223255439e-05,
+      "loss": 3.254,
+      "step": 2916
+    },
+    {
+      "epoch": 0.620803405160947,
+      "grad_norm": 1.1242589723117096,
+      "learning_rate": 9.655724781635269e-05,
+      "loss": 3.2759,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6210162277201383,
+      "grad_norm": 1.46399723944273,
+      "learning_rate": 9.655273054404333e-05,
+      "loss": 3.2495,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6212290502793296,
+      "grad_norm": 1.0064166546493978,
+      "learning_rate": 9.654821041590341e-05,
+      "loss": 3.2799,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6214418728385209,
+      "grad_norm": 1.6355917806945537,
+      "learning_rate": 9.654368743221022e-05,
+      "loss": 3.309,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6216546953977121,
+      "grad_norm": 0.914957828459611,
+      "learning_rate": 9.653916159324123e-05,
+      "loss": 3.2574,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6218675179569034,
+      "grad_norm": 1.6068275172121216,
+      "learning_rate": 9.653463289927411e-05,
+      "loss": 3.2487,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6220803405160947,
+      "grad_norm": 1.039235655834765,
+      "learning_rate": 9.653010135058663e-05,
+      "loss": 3.2737,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6222931630752859,
+      "grad_norm": 1.3372384455629935,
+      "learning_rate": 9.652556694745682e-05,
+      "loss": 3.2543,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6225059856344772,
+      "grad_norm": 1.3584165328781375,
+      "learning_rate": 9.652102969016284e-05,
+      "loss": 3.2669,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6227188081936685,
+      "grad_norm": 1.2931871712175163,
+      "learning_rate": 9.651648957898302e-05,
+      "loss": 3.269,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6229316307528598,
+      "grad_norm": 1.1813095300263832,
+      "learning_rate": 9.65119466141959e-05,
+      "loss": 3.2815,
+      "step": 2927
+    },
+    {
+      "epoch": 0.623144453312051,
+      "grad_norm": 1.2441436745377974,
+      "learning_rate": 9.650740079608014e-05,
+      "loss": 3.2495,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6233572758712423,
+      "grad_norm": 1.2629310764084718,
+      "learning_rate": 9.650285212491463e-05,
+      "loss": 3.2803,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6235700984304337,
+      "grad_norm": 1.2362351011084474,
+      "learning_rate": 9.649830060097843e-05,
+      "loss": 3.288,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6237829209896248,
+      "grad_norm": 1.3688212282938768,
+      "learning_rate": 9.649374622455074e-05,
+      "loss": 3.2813,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6239957435488162,
+      "grad_norm": 1.1078533536219968,
+      "learning_rate": 9.648918899591095e-05,
+      "loss": 3.2646,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6242085661080075,
+      "grad_norm": 1.2649489258978175,
+      "learning_rate": 9.648462891533862e-05,
+      "loss": 3.2392,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6244213886671988,
+      "grad_norm": 1.1689169310119931,
+      "learning_rate": 9.648006598311353e-05,
+      "loss": 3.2534,
+      "step": 2934
+    },
+    {
+      "epoch": 0.62463421122639,
+      "grad_norm": 1.4155113143782032,
+      "learning_rate": 9.647550019951557e-05,
+      "loss": 3.2629,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6248470337855813,
+      "grad_norm": 1.1421190716064005,
+      "learning_rate": 9.647093156482484e-05,
+      "loss": 3.2096,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6250598563447726,
+      "grad_norm": 1.2879211867287041,
+      "learning_rate": 9.646636007932159e-05,
+      "loss": 3.2575,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6252726789039639,
+      "grad_norm": 1.1865567184237638,
+      "learning_rate": 9.646178574328628e-05,
+      "loss": 3.2124,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6254855014631551,
+      "grad_norm": 1.207588338031219,
+      "learning_rate": 9.645720855699954e-05,
+      "loss": 3.2369,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6256983240223464,
+      "grad_norm": 1.251031270923989,
+      "learning_rate": 9.645262852074216e-05,
+      "loss": 3.2629,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6259111465815377,
+      "grad_norm": 1.270705299080731,
+      "learning_rate": 9.644804563479506e-05,
+      "loss": 3.2637,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6261239691407289,
+      "grad_norm": 1.1430628656086528,
+      "learning_rate": 9.644345989943945e-05,
+      "loss": 3.2664,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6263367916999202,
+      "grad_norm": 1.0997138174998289,
+      "learning_rate": 9.643887131495661e-05,
+      "loss": 3.2529,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6265496142591115,
+      "grad_norm": 1.3415083240981684,
+      "learning_rate": 9.643427988162804e-05,
+      "loss": 3.2703,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6267624368183028,
+      "grad_norm": 1.2467626594618302,
+      "learning_rate": 9.64296855997354e-05,
+      "loss": 3.2838,
+      "step": 2945
+    },
+    {
+      "epoch": 0.626975259377494,
+      "grad_norm": 1.2052965165629392,
+      "learning_rate": 9.642508846956053e-05,
+      "loss": 3.2275,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6271880819366853,
+      "grad_norm": 1.1953178691924222,
+      "learning_rate": 9.642048849138547e-05,
+      "loss": 3.2348,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6274009044958766,
+      "grad_norm": 1.2741155561345745,
+      "learning_rate": 9.641588566549239e-05,
+      "loss": 3.2662,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6276137270550678,
+      "grad_norm": 1.3583476328034656,
+      "learning_rate": 9.641127999216366e-05,
+      "loss": 3.2858,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6278265496142591,
+      "grad_norm": 1.082568081981582,
+      "learning_rate": 9.640667147168182e-05,
+      "loss": 3.2476,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6280393721734504,
+      "grad_norm": 1.3072859507920656,
+      "learning_rate": 9.640206010432957e-05,
+      "loss": 3.2348,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6282521947326417,
+      "grad_norm": 1.0944059439680893,
+      "learning_rate": 9.639744589038984e-05,
+      "loss": 3.2655,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6284650172918329,
+      "grad_norm": 1.3311461379339866,
+      "learning_rate": 9.639282883014565e-05,
+      "loss": 3.2738,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6286778398510242,
+      "grad_norm": 1.0502579264479808,
+      "learning_rate": 9.638820892388027e-05,
+      "loss": 3.2729,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6288906624102155,
+      "grad_norm": 1.4638722450085462,
+      "learning_rate": 9.63835861718771e-05,
+      "loss": 3.2501,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6291034849694067,
+      "grad_norm": 1.0570907189978906,
+      "learning_rate": 9.637896057441973e-05,
+      "loss": 3.2582,
+      "step": 2956
+    },
+    {
+      "epoch": 0.629316307528598,
+      "grad_norm": 1.570594808203431,
+      "learning_rate": 9.637433213179191e-05,
+      "loss": 3.3166,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6295291300877893,
+      "grad_norm": 1.1505597645178938,
+      "learning_rate": 9.636970084427759e-05,
+      "loss": 3.2576,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6297419526469806,
+      "grad_norm": 1.3533368511209478,
+      "learning_rate": 9.636506671216089e-05,
+      "loss": 3.2673,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6299547752061718,
+      "grad_norm": 1.229320695681734,
+      "learning_rate": 9.636042973572608e-05,
+      "loss": 3.2297,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6301675977653631,
+      "grad_norm": 1.3525045674562362,
+      "learning_rate": 9.635578991525764e-05,
+      "loss": 3.2293,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6303804203245544,
+      "grad_norm": 1.2493538358942147,
+      "learning_rate": 9.635114725104018e-05,
+      "loss": 3.2448,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6305932428837456,
+      "grad_norm": 1.2662537460731809,
+      "learning_rate": 9.634650174335853e-05,
+      "loss": 3.278,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6308060654429369,
+      "grad_norm": 1.3034636576835774,
+      "learning_rate": 9.634185339249766e-05,
+      "loss": 3.2296,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6310188880021282,
+      "grad_norm": 1.1632260015941085,
+      "learning_rate": 9.633720219874274e-05,
+      "loss": 3.2669,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6312317105613195,
+      "grad_norm": 1.3214276138745376,
+      "learning_rate": 9.63325481623791e-05,
+      "loss": 3.2279,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6314445331205107,
+      "grad_norm": 1.2103048233593237,
+      "learning_rate": 9.632789128369222e-05,
+      "loss": 3.2487,
+      "step": 2967
+    },
+    {
+      "epoch": 0.631657355679702,
+      "grad_norm": 1.2270499612353944,
+      "learning_rate": 9.632323156296784e-05,
+      "loss": 3.2248,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6318701782388934,
+      "grad_norm": 1.1244397926632725,
+      "learning_rate": 9.631856900049175e-05,
+      "loss": 3.26,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6320830007980845,
+      "grad_norm": 1.3306172445717572,
+      "learning_rate": 9.631390359655004e-05,
+      "loss": 3.2112,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6322958233572759,
+      "grad_norm": 1.1641026395244136,
+      "learning_rate": 9.630923535142886e-05,
+      "loss": 3.2329,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6325086459164672,
+      "grad_norm": 1.1291151514059399,
+      "learning_rate": 9.630456426541463e-05,
+      "loss": 3.2707,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6327214684756585,
+      "grad_norm": 1.2089287065506464,
+      "learning_rate": 9.62998903387939e-05,
+      "loss": 3.2219,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6329342910348497,
+      "grad_norm": 1.2424644341146773,
+      "learning_rate": 9.629521357185337e-05,
+      "loss": 3.2495,
+      "step": 2974
+    },
+    {
+      "epoch": 0.633147113594041,
+      "grad_norm": 1.1809873460565534,
+      "learning_rate": 9.629053396487998e-05,
+      "loss": 3.2743,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6333599361532323,
+      "grad_norm": 1.3704264637256645,
+      "learning_rate": 9.628585151816074e-05,
+      "loss": 3.2386,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6335727587124235,
+      "grad_norm": 0.9168202948202893,
+      "learning_rate": 9.628116623198298e-05,
+      "loss": 3.2408,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6337855812716148,
+      "grad_norm": 1.2802220458011015,
+      "learning_rate": 9.627647810663407e-05,
+      "loss": 3.2714,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6339984038308061,
+      "grad_norm": 1.30059360010152,
+      "learning_rate": 9.627178714240163e-05,
+      "loss": 3.2268,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6342112263899974,
+      "grad_norm": 1.024852762851626,
+      "learning_rate": 9.626709333957345e-05,
+      "loss": 3.2448,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6344240489491886,
+      "grad_norm": 1.394179663667279,
+      "learning_rate": 9.626239669843741e-05,
+      "loss": 3.2445,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6346368715083799,
+      "grad_norm": 1.157969271894071,
+      "learning_rate": 9.625769721928172e-05,
+      "loss": 3.1961,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6348496940675712,
+      "grad_norm": 1.4112690862562458,
+      "learning_rate": 9.625299490239461e-05,
+      "loss": 3.2686,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6350625166267624,
+      "grad_norm": 1.0740425034888195,
+      "learning_rate": 9.624828974806458e-05,
+      "loss": 3.2336,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6352753391859537,
+      "grad_norm": 1.3411877702430421,
+      "learning_rate": 9.624358175658025e-05,
+      "loss": 3.2681,
+      "step": 2985
+    },
+    {
+      "epoch": 0.635488161745145,
+      "grad_norm": 1.138384368408514,
+      "learning_rate": 9.623887092823048e-05,
+      "loss": 3.2605,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6357009843043363,
+      "grad_norm": 1.260137514704961,
+      "learning_rate": 9.62341572633042e-05,
+      "loss": 3.2679,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6359138068635275,
+      "grad_norm": 1.1186500182723826,
+      "learning_rate": 9.622944076209062e-05,
+      "loss": 3.2473,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6361266294227188,
+      "grad_norm": 1.3473600636328966,
+      "learning_rate": 9.622472142487906e-05,
+      "loss": 3.2449,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6363394519819101,
+      "grad_norm": 1.268488505827981,
+      "learning_rate": 9.621999925195903e-05,
+      "loss": 3.237,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6365522745411013,
+      "grad_norm": 1.1820197202470126,
+      "learning_rate": 9.621527424362024e-05,
+      "loss": 3.2168,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6367650971002926,
+      "grad_norm": 1.3780190308007216,
+      "learning_rate": 9.621054640015254e-05,
+      "loss": 3.3105,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6369779196594839,
+      "grad_norm": 1.2654276760822423,
+      "learning_rate": 9.620581572184596e-05,
+      "loss": 3.2455,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6371907422186752,
+      "grad_norm": 1.1991775753164793,
+      "learning_rate": 9.620108220899071e-05,
+      "loss": 3.2195,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6374035647778664,
+      "grad_norm": 1.2790691773921752,
+      "learning_rate": 9.619634586187718e-05,
+      "loss": 3.2207,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6376163873370577,
+      "grad_norm": 1.1242791540481991,
+      "learning_rate": 9.61916066807959e-05,
+      "loss": 3.251,
+      "step": 2996
+    },
+    {
+      "epoch": 0.637829209896249,
+      "grad_norm": 1.102039548978879,
+      "learning_rate": 9.618686466603764e-05,
+      "loss": 3.2525,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6380420324554402,
+      "grad_norm": 1.3338006357373098,
+      "learning_rate": 9.618211981789329e-05,
+      "loss": 3.2369,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6382548550146315,
+      "grad_norm": 1.1785289064652822,
+      "learning_rate": 9.617737213665391e-05,
+      "loss": 3.2182,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6384676775738228,
+      "grad_norm": 1.2786972631666436,
+      "learning_rate": 9.617262162261075e-05,
+      "loss": 3.2521,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6386805001330141,
+      "grad_norm": 1.3027367970552457,
+      "learning_rate": 9.616786827605528e-05,
+      "loss": 3.2581,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6388933226922053,
+      "grad_norm": 1.3058157889804112,
+      "learning_rate": 9.616311209727906e-05,
+      "loss": 3.2975,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6391061452513966,
+      "grad_norm": 1.4803238410618136,
+      "learning_rate": 9.615835308657388e-05,
+      "loss": 3.2407,
+      "step": 3003
+    },
+    {
+      "epoch": 0.639318967810588,
+      "grad_norm": 1.0459708768027982,
+      "learning_rate": 9.615359124423167e-05,
+      "loss": 3.2537,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6395317903697793,
+      "grad_norm": 1.6392234100507312,
+      "learning_rate": 9.614882657054457e-05,
+      "loss": 3.3026,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6397446129289704,
+      "grad_norm": 1.112652672779339,
+      "learning_rate": 9.614405906580485e-05,
+      "loss": 3.2147,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6399574354881618,
+      "grad_norm": 1.6673856291585434,
+      "learning_rate": 9.613928873030503e-05,
+      "loss": 3.2648,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6401702580473531,
+      "grad_norm": 1.1238129265059793,
+      "learning_rate": 9.613451556433768e-05,
+      "loss": 3.2618,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6403830806065443,
+      "grad_norm": 1.4990058883012314,
+      "learning_rate": 9.612973956819567e-05,
+      "loss": 3.2649,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6405959031657356,
+      "grad_norm": 1.2458274259402855,
+      "learning_rate": 9.612496074217197e-05,
+      "loss": 3.2442,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6408087257249269,
+      "grad_norm": 1.253190409463169,
+      "learning_rate": 9.612017908655973e-05,
+      "loss": 3.2257,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6410215482841182,
+      "grad_norm": 1.227086474333175,
+      "learning_rate": 9.61153946016523e-05,
+      "loss": 3.2697,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6412343708433094,
+      "grad_norm": 1.2263651285682797,
+      "learning_rate": 9.611060728774319e-05,
+      "loss": 3.2487,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6414471934025007,
+      "grad_norm": 1.1358333635351057,
+      "learning_rate": 9.610581714512608e-05,
+      "loss": 3.2275,
+      "step": 3014
+    },
+    {
+      "epoch": 0.641660015961692,
+      "grad_norm": 1.270111171532051,
+      "learning_rate": 9.610102417409483e-05,
+      "loss": 3.2611,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6418728385208832,
+      "grad_norm": 1.126718502148881,
+      "learning_rate": 9.609622837494348e-05,
+      "loss": 3.2576,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6420856610800745,
+      "grad_norm": 1.4578697974853176,
+      "learning_rate": 9.609142974796622e-05,
+      "loss": 3.2751,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6422984836392658,
+      "grad_norm": 1.1620831998024312,
+      "learning_rate": 9.608662829345741e-05,
+      "loss": 3.2517,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6425113061984571,
+      "grad_norm": 1.2452439778267939,
+      "learning_rate": 9.608182401171162e-05,
+      "loss": 3.2,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6427241287576483,
+      "grad_norm": 1.257938651923558,
+      "learning_rate": 9.60770169030236e-05,
+      "loss": 3.2134,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6429369513168396,
+      "grad_norm": 1.0083286758643688,
+      "learning_rate": 9.607220696768821e-05,
+      "loss": 3.2156,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6431497738760309,
+      "grad_norm": 1.5052914076999562,
+      "learning_rate": 9.606739420600052e-05,
+      "loss": 3.2451,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6433625964352221,
+      "grad_norm": 1.1411744063153546,
+      "learning_rate": 9.606257861825581e-05,
+      "loss": 3.2279,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6435754189944134,
+      "grad_norm": 1.3244271677163049,
+      "learning_rate": 9.605776020474946e-05,
+      "loss": 3.24,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6437882415536047,
+      "grad_norm": 1.3122048430764341,
+      "learning_rate": 9.605293896577709e-05,
+      "loss": 3.2018,
+      "step": 3025
+    },
+    {
+      "epoch": 0.644001064112796,
+      "grad_norm": 1.141690289343046,
+      "learning_rate": 9.604811490163443e-05,
+      "loss": 3.2607,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6442138866719872,
+      "grad_norm": 1.2120336940002405,
+      "learning_rate": 9.604328801261746e-05,
+      "loss": 3.2254,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6444267092311785,
+      "grad_norm": 1.274184719976231,
+      "learning_rate": 9.603845829902224e-05,
+      "loss": 3.2669,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6446395317903698,
+      "grad_norm": 1.166676062188135,
+      "learning_rate": 9.603362576114512e-05,
+      "loss": 3.2337,
+      "step": 3029
+    },
+    {
+      "epoch": 0.644852354349561,
+      "grad_norm": 1.253896518090154,
+      "learning_rate": 9.602879039928249e-05,
+      "loss": 3.2158,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6450651769087523,
+      "grad_norm": 1.1192036123299982,
+      "learning_rate": 9.602395221373104e-05,
+      "loss": 3.2152,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6452779994679436,
+      "grad_norm": 1.3998637847375999,
+      "learning_rate": 9.601911120478754e-05,
+      "loss": 3.2149,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6454908220271349,
+      "grad_norm": 1.2915571093720377,
+      "learning_rate": 9.601426737274896e-05,
+      "loss": 3.2147,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6457036445863261,
+      "grad_norm": 1.3272808889999912,
+      "learning_rate": 9.600942071791248e-05,
+      "loss": 3.2389,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6459164671455174,
+      "grad_norm": 1.079875243349078,
+      "learning_rate": 9.600457124057539e-05,
+      "loss": 3.1864,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6461292897047087,
+      "grad_norm": 1.4035392226916443,
+      "learning_rate": 9.59997189410352e-05,
+      "loss": 3.2287,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6463421122638999,
+      "grad_norm": 1.0302591609415688,
+      "learning_rate": 9.59948638195896e-05,
+      "loss": 3.3056,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6465549348230912,
+      "grad_norm": 1.3219524441887356,
+      "learning_rate": 9.599000587653642e-05,
+      "loss": 3.2193,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6467677573822825,
+      "grad_norm": 1.0169140986539482,
+      "learning_rate": 9.598514511217367e-05,
+      "loss": 3.2122,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6469805799414738,
+      "grad_norm": 1.425044656644629,
+      "learning_rate": 9.598028152679954e-05,
+      "loss": 3.2496,
+      "step": 3040
+    },
+    {
+      "epoch": 0.647193402500665,
+      "grad_norm": 1.0959036541587148,
+      "learning_rate": 9.59754151207124e-05,
+      "loss": 3.2197,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6474062250598563,
+      "grad_norm": 1.2602140689984929,
+      "learning_rate": 9.597054589421077e-05,
+      "loss": 3.2161,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 1.1663323087985027,
+      "learning_rate": 9.596567384759336e-05,
+      "loss": 3.2584,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6478318701782388,
+      "grad_norm": 1.131051342216679,
+      "learning_rate": 9.596079898115909e-05,
+      "loss": 3.232,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6480446927374302,
+      "grad_norm": 1.1782945707070092,
+      "learning_rate": 9.595592129520695e-05,
+      "loss": 3.2314,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6482575152966215,
+      "grad_norm": 1.1820721081840229,
+      "learning_rate": 9.595104079003621e-05,
+      "loss": 3.2256,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6484703378558128,
+      "grad_norm": 1.348616684478313,
+      "learning_rate": 9.594615746594627e-05,
+      "loss": 3.267,
+      "step": 3047
+    },
+    {
+      "epoch": 0.648683160415004,
+      "grad_norm": 1.1421213459632853,
+      "learning_rate": 9.594127132323669e-05,
+      "loss": 3.2317,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6488959829741953,
+      "grad_norm": 1.1530466821273337,
+      "learning_rate": 9.593638236220723e-05,
+      "loss": 3.2377,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6491088055333866,
+      "grad_norm": 1.3147357368457604,
+      "learning_rate": 9.593149058315778e-05,
+      "loss": 3.2065,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6493216280925778,
+      "grad_norm": 0.9763078515534374,
+      "learning_rate": 9.592659598638847e-05,
+      "loss": 3.2021,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6495344506517691,
+      "grad_norm": 1.3976451907884222,
+      "learning_rate": 9.592169857219952e-05,
+      "loss": 3.2158,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6497472732109604,
+      "grad_norm": 1.0940846022571489,
+      "learning_rate": 9.591679834089142e-05,
+      "loss": 3.2399,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6499600957701517,
+      "grad_norm": 1.574638038052169,
+      "learning_rate": 9.591189529276474e-05,
+      "loss": 3.2498,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6501729183293429,
+      "grad_norm": 1.003870191813818,
+      "learning_rate": 9.590698942812029e-05,
+      "loss": 3.2036,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6503857408885342,
+      "grad_norm": 1.569202345430979,
+      "learning_rate": 9.5902080747259e-05,
+      "loss": 3.2468,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6505985634477255,
+      "grad_norm": 1.067282371736521,
+      "learning_rate": 9.5897169250482e-05,
+      "loss": 3.2664,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6508113860069167,
+      "grad_norm": 1.5500440775633246,
+      "learning_rate": 9.589225493809063e-05,
+      "loss": 3.246,
+      "step": 3058
+    },
+    {
+      "epoch": 0.651024208566108,
+      "grad_norm": 1.1367104290009225,
+      "learning_rate": 9.58873378103863e-05,
+      "loss": 3.2113,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6512370311252993,
+      "grad_norm": 1.328665211520704,
+      "learning_rate": 9.588241786767071e-05,
+      "loss": 3.2192,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6514498536844906,
+      "grad_norm": 1.244315013495178,
+      "learning_rate": 9.587749511024568e-05,
+      "loss": 3.1999,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6516626762436818,
+      "grad_norm": 1.0560659894862583,
+      "learning_rate": 9.587256953841317e-05,
+      "loss": 3.2642,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6518754988028731,
+      "grad_norm": 1.1661318289346958,
+      "learning_rate": 9.586764115247535e-05,
+      "loss": 3.1991,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6520883213620644,
+      "grad_norm": 1.1567669345404892,
+      "learning_rate": 9.586270995273459e-05,
+      "loss": 3.2127,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6523011439212556,
+      "grad_norm": 1.1617012229265413,
+      "learning_rate": 9.585777593949336e-05,
+      "loss": 3.2252,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6525139664804469,
+      "grad_norm": 1.2607318033325539,
+      "learning_rate": 9.585283911305437e-05,
+      "loss": 3.2424,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6527267890396382,
+      "grad_norm": 1.2088592659875748,
+      "learning_rate": 9.584789947372046e-05,
+      "loss": 3.2674,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6529396115988295,
+      "grad_norm": 1.2803887437415324,
+      "learning_rate": 9.584295702179465e-05,
+      "loss": 3.1931,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6531524341580207,
+      "grad_norm": 1.145453961618909,
+      "learning_rate": 9.583801175758018e-05,
+      "loss": 3.2035,
+      "step": 3069
+    },
+    {
+      "epoch": 0.653365256717212,
+      "grad_norm": 1.2872543958527245,
+      "learning_rate": 9.583306368138039e-05,
+      "loss": 3.2403,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6535780792764033,
+      "grad_norm": 1.2774041168709445,
+      "learning_rate": 9.582811279349882e-05,
+      "loss": 3.1948,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6537909018355945,
+      "grad_norm": 1.164905419239589,
+      "learning_rate": 9.58231590942392e-05,
+      "loss": 3.2348,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6540037243947858,
+      "grad_norm": 1.2375413789117888,
+      "learning_rate": 9.581820258390543e-05,
+      "loss": 3.2104,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6542165469539771,
+      "grad_norm": 1.1898981089716563,
+      "learning_rate": 9.581324326280156e-05,
+      "loss": 3.1977,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6544293695131684,
+      "grad_norm": 1.1690879541242931,
+      "learning_rate": 9.580828113123181e-05,
+      "loss": 3.2283,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6546421920723596,
+      "grad_norm": 1.1816993459734586,
+      "learning_rate": 9.580331618950063e-05,
+      "loss": 3.2358,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6548550146315509,
+      "grad_norm": 1.1564567912701715,
+      "learning_rate": 9.579834843791256e-05,
+      "loss": 3.2041,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6550678371907422,
+      "grad_norm": 1.2331754001766797,
+      "learning_rate": 9.579337787677239e-05,
+      "loss": 3.2271,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6552806597499335,
+      "grad_norm": 1.1100661363046809,
+      "learning_rate": 9.5788404506385e-05,
+      "loss": 3.1933,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6554934823091247,
+      "grad_norm": 1.156544686537575,
+      "learning_rate": 9.578342832705552e-05,
+      "loss": 3.2105,
+      "step": 3080
+    },
+    {
+      "epoch": 0.655706304868316,
+      "grad_norm": 1.3651257299543325,
+      "learning_rate": 9.577844933908922e-05,
+      "loss": 3.2531,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6559191274275074,
+      "grad_norm": 1.1958672471565888,
+      "learning_rate": 9.57734675427915e-05,
+      "loss": 3.2025,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6561319499866985,
+      "grad_norm": 1.3461408024368926,
+      "learning_rate": 9.576848293846803e-05,
+      "loss": 3.229,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6563447725458899,
+      "grad_norm": 1.0817599627005072,
+      "learning_rate": 9.576349552642456e-05,
+      "loss": 3.2371,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6565575951050812,
+      "grad_norm": 1.229614240860312,
+      "learning_rate": 9.575850530696707e-05,
+      "loss": 3.2268,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6567704176642725,
+      "grad_norm": 1.1944864921680427,
+      "learning_rate": 9.575351228040168e-05,
+      "loss": 3.2246,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6569832402234637,
+      "grad_norm": 1.2099986386178387,
+      "learning_rate": 9.574851644703469e-05,
+      "loss": 3.1962,
+      "step": 3087
+    },
+    {
+      "epoch": 0.657196062782655,
+      "grad_norm": 1.071119539302682,
+      "learning_rate": 9.574351780717258e-05,
+      "loss": 3.1882,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6574088853418463,
+      "grad_norm": 1.2665528542773183,
+      "learning_rate": 9.573851636112199e-05,
+      "loss": 3.2572,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6576217079010375,
+      "grad_norm": 1.3802698147539505,
+      "learning_rate": 9.573351210918974e-05,
+      "loss": 3.2993,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6578345304602288,
+      "grad_norm": 1.4086163011375212,
+      "learning_rate": 9.572850505168285e-05,
+      "loss": 3.2287,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6580473530194201,
+      "grad_norm": 0.9744970248061033,
+      "learning_rate": 9.572349518890844e-05,
+      "loss": 3.2436,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6582601755786114,
+      "grad_norm": 1.3330549033948047,
+      "learning_rate": 9.571848252117388e-05,
+      "loss": 3.1975,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6584729981378026,
+      "grad_norm": 1.0543015105416822,
+      "learning_rate": 9.571346704878666e-05,
+      "loss": 3.2323,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6586858206969939,
+      "grad_norm": 1.417413431726246,
+      "learning_rate": 9.570844877205446e-05,
+      "loss": 3.2166,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6588986432561852,
+      "grad_norm": 1.0363759690018444,
+      "learning_rate": 9.570342769128514e-05,
+      "loss": 3.2293,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6591114658153764,
+      "grad_norm": 1.4421443092810773,
+      "learning_rate": 9.569840380678672e-05,
+      "loss": 3.1743,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6593242883745677,
+      "grad_norm": 1.0858943804537653,
+      "learning_rate": 9.569337711886739e-05,
+      "loss": 3.231,
+      "step": 3098
+    },
+    {
+      "epoch": 0.659537110933759,
+      "grad_norm": 1.2840482419109644,
+      "learning_rate": 9.568834762783554e-05,
+      "loss": 3.2493,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6597499334929503,
+      "grad_norm": 1.2654382764354988,
+      "learning_rate": 9.568331533399967e-05,
+      "loss": 3.2094,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6599627560521415,
+      "grad_norm": 1.0834276368207019,
+      "learning_rate": 9.567828023766853e-05,
+      "loss": 3.1815,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6601755786113328,
+      "grad_norm": 1.4281916391425278,
+      "learning_rate": 9.567324233915099e-05,
+      "loss": 3.2126,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6603884011705241,
+      "grad_norm": 1.102079213039178,
+      "learning_rate": 9.56682016387561e-05,
+      "loss": 3.2158,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6606012237297153,
+      "grad_norm": 1.205848881103214,
+      "learning_rate": 9.56631581367931e-05,
+      "loss": 3.2187,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6608140462889066,
+      "grad_norm": 1.2003172205575114,
+      "learning_rate": 9.565811183357137e-05,
+      "loss": 3.2357,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6610268688480979,
+      "grad_norm": 1.4710185032928915,
+      "learning_rate": 9.56530627294005e-05,
+      "loss": 3.2555,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6612396914072892,
+      "grad_norm": 1.089510052598004,
+      "learning_rate": 9.564801082459022e-05,
+      "loss": 3.237,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6614525139664804,
+      "grad_norm": 1.3477221136901225,
+      "learning_rate": 9.564295611945046e-05,
+      "loss": 3.2486,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6616653365256717,
+      "grad_norm": 1.2765206470315353,
+      "learning_rate": 9.563789861429131e-05,
+      "loss": 3.2681,
+      "step": 3109
+    },
+    {
+      "epoch": 0.661878159084863,
+      "grad_norm": 1.0650588295653325,
+      "learning_rate": 9.5632838309423e-05,
+      "loss": 3.1999,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6620909816440542,
+      "grad_norm": 1.2228041005485466,
+      "learning_rate": 9.562777520515598e-05,
+      "loss": 3.1988,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6623038042032455,
+      "grad_norm": 1.014638355422869,
+      "learning_rate": 9.562270930180083e-05,
+      "loss": 3.197,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6625166267624368,
+      "grad_norm": 1.3225964591508852,
+      "learning_rate": 9.561764059966837e-05,
+      "loss": 3.2409,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6627294493216281,
+      "grad_norm": 1.0198084975547057,
+      "learning_rate": 9.561256909906951e-05,
+      "loss": 3.2208,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6629422718808193,
+      "grad_norm": 1.431834095804018,
+      "learning_rate": 9.560749480031537e-05,
+      "loss": 3.2223,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6631550944400106,
+      "grad_norm": 1.055967974816523,
+      "learning_rate": 9.560241770371723e-05,
+      "loss": 3.1817,
+      "step": 3116
+    },
+    {
+      "epoch": 0.663367916999202,
+      "grad_norm": 1.3213860393778425,
+      "learning_rate": 9.559733780958659e-05,
+      "loss": 3.1782,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6635807395583931,
+      "grad_norm": 1.1142737737387582,
+      "learning_rate": 9.559225511823504e-05,
+      "loss": 3.1945,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6637935621175844,
+      "grad_norm": 1.153151151127635,
+      "learning_rate": 9.55871696299744e-05,
+      "loss": 3.2251,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6640063846767758,
+      "grad_norm": 1.2775934785339162,
+      "learning_rate": 9.558208134511664e-05,
+      "loss": 3.2398,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6642192072359671,
+      "grad_norm": 1.3663678202459582,
+      "learning_rate": 9.557699026397392e-05,
+      "loss": 3.2133,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6644320297951583,
+      "grad_norm": 1.119082115777685,
+      "learning_rate": 9.557189638685855e-05,
+      "loss": 3.1808,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6646448523543496,
+      "grad_norm": 1.2533769827924641,
+      "learning_rate": 9.556679971408301e-05,
+      "loss": 3.239,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6648576749135409,
+      "grad_norm": 1.097554092873656,
+      "learning_rate": 9.556170024595997e-05,
+      "loss": 3.2347,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6650704974727321,
+      "grad_norm": 1.367314506992631,
+      "learning_rate": 9.555659798280226e-05,
+      "loss": 3.2164,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6652833200319234,
+      "grad_norm": 1.0776196666711382,
+      "learning_rate": 9.555149292492289e-05,
+      "loss": 3.1795,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6654961425911147,
+      "grad_norm": 1.314687228674194,
+      "learning_rate": 9.554638507263504e-05,
+      "loss": 3.1875,
+      "step": 3127
+    },
+    {
+      "epoch": 0.665708965150306,
+      "grad_norm": 1.1169664333014813,
+      "learning_rate": 9.554127442625205e-05,
+      "loss": 3.2694,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6659217877094972,
+      "grad_norm": 1.2831832161499297,
+      "learning_rate": 9.553616098608742e-05,
+      "loss": 3.2381,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6661346102686885,
+      "grad_norm": 1.1143164584084466,
+      "learning_rate": 9.553104475245487e-05,
+      "loss": 3.1917,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6663474328278798,
+      "grad_norm": 1.3043455064614027,
+      "learning_rate": 9.552592572566826e-05,
+      "loss": 3.2149,
+      "step": 3131
+    },
+    {
+      "epoch": 0.666560255387071,
+      "grad_norm": 1.0445284829461383,
+      "learning_rate": 9.55208039060416e-05,
+      "loss": 3.2069,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6667730779462623,
+      "grad_norm": 1.1756253484729842,
+      "learning_rate": 9.551567929388909e-05,
+      "loss": 3.2383,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6669859005054536,
+      "grad_norm": 1.3005885987356995,
+      "learning_rate": 9.551055188952514e-05,
+      "loss": 3.1933,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6671987230646449,
+      "grad_norm": 1.0988916882379691,
+      "learning_rate": 9.550542169326427e-05,
+      "loss": 3.1952,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6674115456238361,
+      "grad_norm": 1.2412403477314191,
+      "learning_rate": 9.55002887054212e-05,
+      "loss": 3.2326,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6676243681830274,
+      "grad_norm": 1.1594599526427183,
+      "learning_rate": 9.549515292631084e-05,
+      "loss": 3.2221,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6678371907422187,
+      "grad_norm": 1.1462068374907024,
+      "learning_rate": 9.549001435624823e-05,
+      "loss": 3.197,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6680500133014099,
+      "grad_norm": 1.3428022250066622,
+      "learning_rate": 9.54848729955486e-05,
+      "loss": 3.2125,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6682628358606012,
+      "grad_norm": 1.1537109798741136,
+      "learning_rate": 9.547972884452735e-05,
+      "loss": 3.2082,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6684756584197925,
+      "grad_norm": 1.121697940295118,
+      "learning_rate": 9.547458190350008e-05,
+      "loss": 3.2207,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6686884809789838,
+      "grad_norm": 1.0578857251104945,
+      "learning_rate": 9.54694321727825e-05,
+      "loss": 3.2576,
+      "step": 3142
+    },
+    {
+      "epoch": 0.668901303538175,
+      "grad_norm": 1.35400853592918,
+      "learning_rate": 9.546427965269054e-05,
+      "loss": 3.2381,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6691141260973663,
+      "grad_norm": 1.0163800665828484,
+      "learning_rate": 9.54591243435403e-05,
+      "loss": 3.2021,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6693269486565576,
+      "grad_norm": 1.3815477964764706,
+      "learning_rate": 9.545396624564802e-05,
+      "loss": 3.23,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6695397712157489,
+      "grad_norm": 1.1003786363379666,
+      "learning_rate": 9.544880535933015e-05,
+      "loss": 3.2348,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6697525937749401,
+      "grad_norm": 1.4127941883906592,
+      "learning_rate": 9.544364168490328e-05,
+      "loss": 3.1985,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6699654163341314,
+      "grad_norm": 1.0129455993358196,
+      "learning_rate": 9.543847522268416e-05,
+      "loss": 3.2215,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6701782388933227,
+      "grad_norm": 1.260070886539418,
+      "learning_rate": 9.543330597298979e-05,
+      "loss": 3.2021,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6703910614525139,
+      "grad_norm": 1.1881901669637875,
+      "learning_rate": 9.542813393613721e-05,
+      "loss": 3.2019,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6706038840117052,
+      "grad_norm": 1.2401269075830947,
+      "learning_rate": 9.542295911244376e-05,
+      "loss": 3.2177,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6708167065708965,
+      "grad_norm": 1.0948786053720032,
+      "learning_rate": 9.541778150222686e-05,
+      "loss": 3.1879,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6710295291300878,
+      "grad_norm": 1.4202641987336537,
+      "learning_rate": 9.541260110580417e-05,
+      "loss": 3.2102,
+      "step": 3153
+    },
+    {
+      "epoch": 0.671242351689279,
+      "grad_norm": 1.0842341815245333,
+      "learning_rate": 9.540741792349346e-05,
+      "loss": 3.1957,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6714551742484703,
+      "grad_norm": 1.320635182808839,
+      "learning_rate": 9.540223195561271e-05,
+      "loss": 3.2172,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6716679968076616,
+      "grad_norm": 1.0630543841195637,
+      "learning_rate": 9.539704320248007e-05,
+      "loss": 3.1968,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6718808193668528,
+      "grad_norm": 1.3914336956877507,
+      "learning_rate": 9.539185166441382e-05,
+      "loss": 3.2803,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6720936419260441,
+      "grad_norm": 1.137557456631497,
+      "learning_rate": 9.538665734173246e-05,
+      "loss": 3.1677,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6723064644852355,
+      "grad_norm": 1.205321137839855,
+      "learning_rate": 9.538146023475464e-05,
+      "loss": 3.2226,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6725192870444268,
+      "grad_norm": 1.1773604379027753,
+      "learning_rate": 9.537626034379919e-05,
+      "loss": 3.2309,
+      "step": 3160
+    },
+    {
+      "epoch": 0.672732109603618,
+      "grad_norm": 1.139267359403809,
+      "learning_rate": 9.537105766918509e-05,
+      "loss": 3.1948,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6729449321628093,
+      "grad_norm": 1.1347250624328296,
+      "learning_rate": 9.536585221123151e-05,
+      "loss": 3.2526,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6731577547220006,
+      "grad_norm": 1.2528835034512396,
+      "learning_rate": 9.53606439702578e-05,
+      "loss": 3.2584,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6733705772811918,
+      "grad_norm": 1.1893383930308148,
+      "learning_rate": 9.535543294658342e-05,
+      "loss": 3.2059,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6735833998403831,
+      "grad_norm": 1.3467702728821185,
+      "learning_rate": 9.535021914052808e-05,
+      "loss": 3.2513,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6737962223995744,
+      "grad_norm": 1.0620855434222967,
+      "learning_rate": 9.534500255241164e-05,
+      "loss": 3.1689,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6740090449587657,
+      "grad_norm": 1.28248364636991,
+      "learning_rate": 9.53397831825541e-05,
+      "loss": 3.2322,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6742218675179569,
+      "grad_norm": 1.225555170495291,
+      "learning_rate": 9.533456103127566e-05,
+      "loss": 3.2522,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6744346900771482,
+      "grad_norm": 0.96656066953877,
+      "learning_rate": 9.532933609889663e-05,
+      "loss": 3.1865,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6746475126363395,
+      "grad_norm": 1.3706927939794835,
+      "learning_rate": 9.53241083857376e-05,
+      "loss": 3.2245,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6748603351955307,
+      "grad_norm": 1.040343804346695,
+      "learning_rate": 9.531887789211925e-05,
+      "loss": 3.1938,
+      "step": 3171
+    },
+    {
+      "epoch": 0.675073157754722,
+      "grad_norm": 1.435513699162156,
+      "learning_rate": 9.531364461836245e-05,
+      "loss": 3.2176,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6752859803139133,
+      "grad_norm": 0.9979781857800953,
+      "learning_rate": 9.530840856478822e-05,
+      "loss": 3.1865,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6754988028731046,
+      "grad_norm": 1.6205545960553245,
+      "learning_rate": 9.53031697317178e-05,
+      "loss": 3.1815,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6757116254322958,
+      "grad_norm": 0.8871150122014366,
+      "learning_rate": 9.529792811947259e-05,
+      "loss": 3.2401,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6759244479914871,
+      "grad_norm": 1.4153951039147596,
+      "learning_rate": 9.529268372837407e-05,
+      "loss": 3.225,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6761372705506784,
+      "grad_norm": 1.0588833655869088,
+      "learning_rate": 9.528743655874404e-05,
+      "loss": 3.2462,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6763500931098696,
+      "grad_norm": 1.4204639605292053,
+      "learning_rate": 9.528218661090434e-05,
+      "loss": 3.2246,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6765629156690609,
+      "grad_norm": 1.1377710577928677,
+      "learning_rate": 9.527693388517706e-05,
+      "loss": 3.1968,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6767757382282522,
+      "grad_norm": 1.233242092956062,
+      "learning_rate": 9.527167838188446e-05,
+      "loss": 3.2053,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6769885607874435,
+      "grad_norm": 1.2098242988566288,
+      "learning_rate": 9.52664201013489e-05,
+      "loss": 3.1962,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6772013833466347,
+      "grad_norm": 1.0076923169932666,
+      "learning_rate": 9.526115904389296e-05,
+      "loss": 3.2308,
+      "step": 3182
+    },
+    {
+      "epoch": 0.677414205905826,
+      "grad_norm": 1.371073501886988,
+      "learning_rate": 9.52558952098394e-05,
+      "loss": 3.2092,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6776270284650173,
+      "grad_norm": 1.1501377404919069,
+      "learning_rate": 9.525062859951113e-05,
+      "loss": 3.1627,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6778398510242085,
+      "grad_norm": 1.1444866451727955,
+      "learning_rate": 9.524535921323125e-05,
+      "loss": 3.1905,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6780526735833998,
+      "grad_norm": 1.2261971698348735,
+      "learning_rate": 9.5240087051323e-05,
+      "loss": 3.1971,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6782654961425911,
+      "grad_norm": 1.1010598653634631,
+      "learning_rate": 9.523481211410981e-05,
+      "loss": 3.2306,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6784783187017824,
+      "grad_norm": 1.0982873736195642,
+      "learning_rate": 9.522953440191528e-05,
+      "loss": 3.1744,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6786911412609736,
+      "grad_norm": 1.3149629563224983,
+      "learning_rate": 9.522425391506318e-05,
+      "loss": 3.2149,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6789039638201649,
+      "grad_norm": 0.9971945556961135,
+      "learning_rate": 9.521897065387743e-05,
+      "loss": 3.169,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6791167863793562,
+      "grad_norm": 1.2098779877241992,
+      "learning_rate": 9.521368461868216e-05,
+      "loss": 3.1679,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6793296089385474,
+      "grad_norm": 1.2420378016942184,
+      "learning_rate": 9.520839580980166e-05,
+      "loss": 3.2105,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6795424314977387,
+      "grad_norm": 1.2304913630329866,
+      "learning_rate": 9.520310422756034e-05,
+      "loss": 3.2219,
+      "step": 3193
+    },
+    {
+      "epoch": 0.67975525405693,
+      "grad_norm": 1.13496004845846,
+      "learning_rate": 9.519780987228283e-05,
+      "loss": 3.245,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6799680766161214,
+      "grad_norm": 1.2904773656942692,
+      "learning_rate": 9.519251274429393e-05,
+      "loss": 3.214,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6801808991753125,
+      "grad_norm": 1.0772884233618907,
+      "learning_rate": 9.518721284391858e-05,
+      "loss": 3.1914,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6803937217345039,
+      "grad_norm": 1.3230338864142,
+      "learning_rate": 9.518191017148195e-05,
+      "loss": 3.2332,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6806065442936952,
+      "grad_norm": 1.1108680962449424,
+      "learning_rate": 9.51766047273093e-05,
+      "loss": 3.2109,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6808193668528864,
+      "grad_norm": 1.421689449065264,
+      "learning_rate": 9.517129651172609e-05,
+      "loss": 3.2533,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6810321894120777,
+      "grad_norm": 1.0109146866460819,
+      "learning_rate": 9.516598552505798e-05,
+      "loss": 3.201,
+      "step": 3200
+    },
+    {
+      "epoch": 0.681245011971269,
+      "grad_norm": 1.4782299448302143,
+      "learning_rate": 9.516067176763078e-05,
+      "loss": 3.1963,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6814578345304603,
+      "grad_norm": 1.0160819590009873,
+      "learning_rate": 9.515535523977047e-05,
+      "loss": 3.2206,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6816706570896515,
+      "grad_norm": 1.1718757776548692,
+      "learning_rate": 9.515003594180319e-05,
+      "loss": 3.189,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6818834796488428,
+      "grad_norm": 1.3613667350370378,
+      "learning_rate": 9.514471387405526e-05,
+      "loss": 3.2138,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6820963022080341,
+      "grad_norm": 1.040926424148393,
+      "learning_rate": 9.513938903685316e-05,
+      "loss": 3.1607,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6823091247672253,
+      "grad_norm": 1.246652933296475,
+      "learning_rate": 9.513406143052356e-05,
+      "loss": 3.1809,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6825219473264166,
+      "grad_norm": 1.212560230692388,
+      "learning_rate": 9.512873105539329e-05,
+      "loss": 3.1694,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6827347698856079,
+      "grad_norm": 1.2200911615329268,
+      "learning_rate": 9.512339791178935e-05,
+      "loss": 3.1658,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6829475924447992,
+      "grad_norm": 1.0447195027423561,
+      "learning_rate": 9.51180620000389e-05,
+      "loss": 3.1962,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6831604150039904,
+      "grad_norm": 1.2463887805848386,
+      "learning_rate": 9.511272332046927e-05,
+      "loss": 3.2003,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6833732375631817,
+      "grad_norm": 1.0192383147921902,
+      "learning_rate": 9.510738187340797e-05,
+      "loss": 3.2078,
+      "step": 3211
+    },
+    {
+      "epoch": 0.683586060122373,
+      "grad_norm": 1.3612994749229173,
+      "learning_rate": 9.51020376591827e-05,
+      "loss": 3.2006,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6837988826815642,
+      "grad_norm": 1.0594966146631484,
+      "learning_rate": 9.509669067812128e-05,
+      "loss": 3.2234,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6840117052407555,
+      "grad_norm": 1.3258283111645224,
+      "learning_rate": 9.509134093055174e-05,
+      "loss": 3.18,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6842245277999468,
+      "grad_norm": 1.2475206675832693,
+      "learning_rate": 9.508598841680225e-05,
+      "loss": 3.1729,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6844373503591381,
+      "grad_norm": 1.1484061216514851,
+      "learning_rate": 9.508063313720119e-05,
+      "loss": 3.1634,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6846501729183293,
+      "grad_norm": 1.4626475978627178,
+      "learning_rate": 9.507527509207708e-05,
+      "loss": 3.2461,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6848629954775206,
+      "grad_norm": 1.0185639417885692,
+      "learning_rate": 9.50699142817586e-05,
+      "loss": 3.1474,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6850758180367119,
+      "grad_norm": 1.5895397261358217,
+      "learning_rate": 9.506455070657464e-05,
+      "loss": 3.2184,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6852886405959032,
+      "grad_norm": 0.9263199806875713,
+      "learning_rate": 9.505918436685422e-05,
+      "loss": 3.2129,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6855014631550944,
+      "grad_norm": 1.3885432137548122,
+      "learning_rate": 9.505381526292653e-05,
+      "loss": 3.1852,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 1.1943080730052198,
+      "learning_rate": 9.504844339512095e-05,
+      "loss": 3.2018,
+      "step": 3222
+    },
+    {
+      "epoch": 0.685927108273477,
+      "grad_norm": 1.1074924318014605,
+      "learning_rate": 9.504306876376705e-05,
+      "loss": 3.2272,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6861399308326682,
+      "grad_norm": 1.1688781088876554,
+      "learning_rate": 9.503769136919452e-05,
+      "loss": 3.1508,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6863527533918595,
+      "grad_norm": 1.092786755154763,
+      "learning_rate": 9.503231121173326e-05,
+      "loss": 3.2315,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6865655759510508,
+      "grad_norm": 1.1792557907615848,
+      "learning_rate": 9.50269282917133e-05,
+      "loss": 3.2137,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6867783985102421,
+      "grad_norm": 1.2228703219577077,
+      "learning_rate": 9.502154260946486e-05,
+      "loss": 3.1662,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6869912210694333,
+      "grad_norm": 1.1355771020699619,
+      "learning_rate": 9.501615416531836e-05,
+      "loss": 3.1451,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6872040436286246,
+      "grad_norm": 1.1423201880387934,
+      "learning_rate": 9.501076295960432e-05,
+      "loss": 3.195,
+      "step": 3229
+    },
+    {
+      "epoch": 0.687416866187816,
+      "grad_norm": 1.1495948440118173,
+      "learning_rate": 9.500536899265349e-05,
+      "loss": 3.1951,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6876296887470071,
+      "grad_norm": 1.2702095288859208,
+      "learning_rate": 9.499997226479679e-05,
+      "loss": 3.2087,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6878425113061984,
+      "grad_norm": 1.2979990305490094,
+      "learning_rate": 9.499457277636526e-05,
+      "loss": 3.2397,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6880553338653898,
+      "grad_norm": 1.112209587298845,
+      "learning_rate": 9.498917052769014e-05,
+      "loss": 3.1991,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6882681564245811,
+      "grad_norm": 1.1685579291600676,
+      "learning_rate": 9.498376551910286e-05,
+      "loss": 3.1795,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6884809789837723,
+      "grad_norm": 1.1204138324359174,
+      "learning_rate": 9.497835775093495e-05,
+      "loss": 3.1299,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6886938015429636,
+      "grad_norm": 1.2879690260810663,
+      "learning_rate": 9.497294722351822e-05,
+      "loss": 3.2144,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6889066241021549,
+      "grad_norm": 0.9814150279127819,
+      "learning_rate": 9.496753393718453e-05,
+      "loss": 3.1723,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6891194466613461,
+      "grad_norm": 1.2753423852736492,
+      "learning_rate": 9.4962117892266e-05,
+      "loss": 3.1834,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6893322692205374,
+      "grad_norm": 1.180825962451507,
+      "learning_rate": 9.495669908909484e-05,
+      "loss": 3.1732,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6895450917797287,
+      "grad_norm": 1.2996568427126702,
+      "learning_rate": 9.495127752800353e-05,
+      "loss": 3.1801,
+      "step": 3240
+    },
+    {
+      "epoch": 0.68975791433892,
+      "grad_norm": 1.0089581240391914,
+      "learning_rate": 9.494585320932462e-05,
+      "loss": 3.1863,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6899707368981112,
+      "grad_norm": 1.3978025738414868,
+      "learning_rate": 9.494042613339089e-05,
+      "loss": 3.187,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6901835594573025,
+      "grad_norm": 1.033195217045015,
+      "learning_rate": 9.493499630053526e-05,
+      "loss": 3.1613,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6903963820164938,
+      "grad_norm": 1.3539843830138372,
+      "learning_rate": 9.492956371109083e-05,
+      "loss": 3.2189,
+      "step": 3244
+    },
+    {
+      "epoch": 0.690609204575685,
+      "grad_norm": 1.0702750876358804,
+      "learning_rate": 9.492412836539088e-05,
+      "loss": 3.1926,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6908220271348763,
+      "grad_norm": 1.4743915200180657,
+      "learning_rate": 9.491869026376883e-05,
+      "loss": 3.1932,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6910348496940676,
+      "grad_norm": 0.9972769200674895,
+      "learning_rate": 9.49132494065583e-05,
+      "loss": 3.1643,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6912476722532589,
+      "grad_norm": 1.2520625288614298,
+      "learning_rate": 9.490780579409306e-05,
+      "loss": 3.202,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6914604948124501,
+      "grad_norm": 1.2575072341122824,
+      "learning_rate": 9.490235942670706e-05,
+      "loss": 3.1938,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6916733173716414,
+      "grad_norm": 1.1104090598234377,
+      "learning_rate": 9.489691030473442e-05,
+      "loss": 3.1778,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6918861399308327,
+      "grad_norm": 1.2035428678350772,
+      "learning_rate": 9.48914584285094e-05,
+      "loss": 3.1968,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6920989624900239,
+      "grad_norm": 1.2248181675957714,
+      "learning_rate": 9.488600379836647e-05,
+      "loss": 3.1886,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6923117850492152,
+      "grad_norm": 1.1540268031565606,
+      "learning_rate": 9.488054641464023e-05,
+      "loss": 3.1655,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6925246076084065,
+      "grad_norm": 1.1105164716949938,
+      "learning_rate": 9.48750862776655e-05,
+      "loss": 3.2026,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6927374301675978,
+      "grad_norm": 1.1957180760651118,
+      "learning_rate": 9.486962338777723e-05,
+      "loss": 3.2109,
+      "step": 3255
+    },
+    {
+      "epoch": 0.692950252726789,
+      "grad_norm": 1.070132056787562,
+      "learning_rate": 9.486415774531053e-05,
+      "loss": 3.1839,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6931630752859803,
+      "grad_norm": 1.2133373930995814,
+      "learning_rate": 9.48586893506007e-05,
+      "loss": 3.1748,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6933758978451716,
+      "grad_norm": 1.2338646867884526,
+      "learning_rate": 9.485321820398321e-05,
+      "loss": 3.1797,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6935887204043628,
+      "grad_norm": 1.1425499504126828,
+      "learning_rate": 9.48477443057937e-05,
+      "loss": 3.1857,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6938015429635541,
+      "grad_norm": 1.2774919651598649,
+      "learning_rate": 9.484226765636797e-05,
+      "loss": 3.2254,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6940143655227454,
+      "grad_norm": 1.192071372902788,
+      "learning_rate": 9.483678825604198e-05,
+      "loss": 3.2139,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6942271880819367,
+      "grad_norm": 1.2453368048026208,
+      "learning_rate": 9.483130610515187e-05,
+      "loss": 3.1765,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6944400106411279,
+      "grad_norm": 1.2304978146021388,
+      "learning_rate": 9.482582120403399e-05,
+      "loss": 3.1584,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6946528332003192,
+      "grad_norm": 1.1024498774182683,
+      "learning_rate": 9.482033355302475e-05,
+      "loss": 3.1656,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6948656557595105,
+      "grad_norm": 1.4408834494682923,
+      "learning_rate": 9.481484315246086e-05,
+      "loss": 3.1725,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6950784783187017,
+      "grad_norm": 1.013079103599839,
+      "learning_rate": 9.480935000267908e-05,
+      "loss": 3.212,
+      "step": 3266
+    },
+    {
+      "epoch": 0.695291300877893,
+      "grad_norm": 1.253929072726404,
+      "learning_rate": 9.480385410401641e-05,
+      "loss": 3.1676,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6955041234370843,
+      "grad_norm": 1.078488436030153,
+      "learning_rate": 9.479835545681005e-05,
+      "loss": 3.1662,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6957169459962756,
+      "grad_norm": 1.4213376124607808,
+      "learning_rate": 9.479285406139726e-05,
+      "loss": 3.199,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6959297685554668,
+      "grad_norm": 0.8988460651202428,
+      "learning_rate": 9.478734991811556e-05,
+      "loss": 3.1862,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6961425911146581,
+      "grad_norm": 1.4456253036548177,
+      "learning_rate": 9.47818430273026e-05,
+      "loss": 3.2016,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6963554136738495,
+      "grad_norm": 0.9800891357856799,
+      "learning_rate": 9.47763333892962e-05,
+      "loss": 3.211,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6965682362330406,
+      "grad_norm": 1.310258446511787,
+      "learning_rate": 9.477082100443437e-05,
+      "loss": 3.2273,
+      "step": 3273
+    },
+    {
+      "epoch": 0.696781058792232,
+      "grad_norm": 1.184526791904899,
+      "learning_rate": 9.476530587305526e-05,
+      "loss": 3.1531,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6969938813514233,
+      "grad_norm": 1.2688169366820636,
+      "learning_rate": 9.47597879954972e-05,
+      "loss": 3.1731,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6972067039106146,
+      "grad_norm": 1.404995585887335,
+      "learning_rate": 9.475426737209871e-05,
+      "loss": 3.1894,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6974195264698058,
+      "grad_norm": 0.8760880962069845,
+      "learning_rate": 9.474874400319844e-05,
+      "loss": 3.1795,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6976323490289971,
+      "grad_norm": 1.1224334755122598,
+      "learning_rate": 9.474321788913524e-05,
+      "loss": 3.1824,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6978451715881884,
+      "grad_norm": 1.3305972310475485,
+      "learning_rate": 9.47376890302481e-05,
+      "loss": 3.1983,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6980579941473796,
+      "grad_norm": 1.1617070338486586,
+      "learning_rate": 9.473215742687621e-05,
+      "loss": 3.1825,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6982708167065709,
+      "grad_norm": 1.1391713324646762,
+      "learning_rate": 9.472662307935891e-05,
+      "loss": 3.1682,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6984836392657622,
+      "grad_norm": 1.1452714441484146,
+      "learning_rate": 9.47210859880357e-05,
+      "loss": 3.1827,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6986964618249535,
+      "grad_norm": 1.3260293432713277,
+      "learning_rate": 9.471554615324628e-05,
+      "loss": 3.1767,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6989092843841447,
+      "grad_norm": 0.9988812888399677,
+      "learning_rate": 9.471000357533047e-05,
+      "loss": 3.1413,
+      "step": 3284
+    },
+    {
+      "epoch": 0.699122106943336,
+      "grad_norm": 1.3757809036714526,
+      "learning_rate": 9.47044582546283e-05,
+      "loss": 3.2282,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6993349295025273,
+      "grad_norm": 1.1912679234183197,
+      "learning_rate": 9.469891019147995e-05,
+      "loss": 3.1909,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6995477520617186,
+      "grad_norm": 1.2148031708550138,
+      "learning_rate": 9.469335938622579e-05,
+      "loss": 3.1732,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6997605746209098,
+      "grad_norm": 1.0204591008500152,
+      "learning_rate": 9.468780583920631e-05,
+      "loss": 3.2166,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6999733971801011,
+      "grad_norm": 1.4168955749904164,
+      "learning_rate": 9.468224955076224e-05,
+      "loss": 3.1928,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7001862197392924,
+      "grad_norm": 1.1052458336385595,
+      "learning_rate": 9.467669052123439e-05,
+      "loss": 3.1633,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7003990422984836,
+      "grad_norm": 1.125391014379309,
+      "learning_rate": 9.467112875096382e-05,
+      "loss": 3.1757,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7006118648576749,
+      "grad_norm": 1.117219856719665,
+      "learning_rate": 9.46655642402917e-05,
+      "loss": 3.1683,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7008246874168662,
+      "grad_norm": 1.1553969873122991,
+      "learning_rate": 9.465999698955941e-05,
+      "loss": 3.1634,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7010375099760575,
+      "grad_norm": 1.150973092695771,
+      "learning_rate": 9.465442699910847e-05,
+      "loss": 3.1953,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7012503325352487,
+      "grad_norm": 1.2743669963328021,
+      "learning_rate": 9.464885426928056e-05,
+      "loss": 3.1848,
+      "step": 3295
+    },
+    {
+      "epoch": 0.70146315509444,
+      "grad_norm": 1.0076624800691243,
+      "learning_rate": 9.464327880041759e-05,
+      "loss": 3.1669,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7016759776536313,
+      "grad_norm": 1.309628814487651,
+      "learning_rate": 9.463770059286157e-05,
+      "loss": 3.1825,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7018888002128225,
+      "grad_norm": 1.0161606676255412,
+      "learning_rate": 9.463211964695468e-05,
+      "loss": 3.1602,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7021016227720138,
+      "grad_norm": 1.2631828291632383,
+      "learning_rate": 9.462653596303932e-05,
+      "loss": 3.223,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7023144453312051,
+      "grad_norm": 1.249812206714401,
+      "learning_rate": 9.4620949541458e-05,
+      "loss": 3.1746,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7025272678903964,
+      "grad_norm": 1.0571694176751305,
+      "learning_rate": 9.461536038255346e-05,
+      "loss": 3.1744,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7027400904495876,
+      "grad_norm": 1.2690781841855192,
+      "learning_rate": 9.460976848666855e-05,
+      "loss": 3.1933,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7029529130087789,
+      "grad_norm": 0.9919944739152158,
+      "learning_rate": 9.460417385414631e-05,
+      "loss": 3.2072,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7031657355679702,
+      "grad_norm": 1.274720911796629,
+      "learning_rate": 9.459857648532998e-05,
+      "loss": 3.167,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7033785581271614,
+      "grad_norm": 1.1626889883836717,
+      "learning_rate": 9.459297638056289e-05,
+      "loss": 3.1917,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7035913806863527,
+      "grad_norm": 1.3474354561739799,
+      "learning_rate": 9.458737354018859e-05,
+      "loss": 3.1865,
+      "step": 3306
+    },
+    {
+      "epoch": 0.703804203245544,
+      "grad_norm": 1.0577211235169341,
+      "learning_rate": 9.458176796455083e-05,
+      "loss": 3.1993,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7040170258047354,
+      "grad_norm": 1.1891394136794438,
+      "learning_rate": 9.457615965399347e-05,
+      "loss": 3.1683,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7042298483639265,
+      "grad_norm": 1.1738212933390928,
+      "learning_rate": 9.457054860886057e-05,
+      "loss": 3.1843,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7044426709231179,
+      "grad_norm": 1.196813236722334,
+      "learning_rate": 9.456493482949631e-05,
+      "loss": 3.1712,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7046554934823092,
+      "grad_norm": 1.2102881809925605,
+      "learning_rate": 9.455931831624512e-05,
+      "loss": 3.1871,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7048683160415004,
+      "grad_norm": 1.1219562322238126,
+      "learning_rate": 9.455369906945151e-05,
+      "loss": 3.1919,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7050811386006917,
+      "grad_norm": 0.9629298642578851,
+      "learning_rate": 9.454807708946023e-05,
+      "loss": 3.1995,
+      "step": 3313
+    },
+    {
+      "epoch": 0.705293961159883,
+      "grad_norm": 1.3344323739140278,
+      "learning_rate": 9.454245237661616e-05,
+      "loss": 3.1731,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7055067837190743,
+      "grad_norm": 1.1014441745192967,
+      "learning_rate": 9.453682493126435e-05,
+      "loss": 3.1502,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7057196062782655,
+      "grad_norm": 1.1793834631762652,
+      "learning_rate": 9.453119475375001e-05,
+      "loss": 3.1706,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7059324288374568,
+      "grad_norm": 1.0674540680884397,
+      "learning_rate": 9.452556184441854e-05,
+      "loss": 3.1474,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7061452513966481,
+      "grad_norm": 1.3978326881125085,
+      "learning_rate": 9.451992620361552e-05,
+      "loss": 3.187,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7063580739558393,
+      "grad_norm": 1.2303418970783506,
+      "learning_rate": 9.451428783168664e-05,
+      "loss": 3.1679,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7065708965150306,
+      "grad_norm": 1.2798576575791354,
+      "learning_rate": 9.45086467289778e-05,
+      "loss": 3.2217,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7067837190742219,
+      "grad_norm": 1.044547190914594,
+      "learning_rate": 9.450300289583507e-05,
+      "loss": 3.2247,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7069965416334132,
+      "grad_norm": 1.4632980492750578,
+      "learning_rate": 9.449735633260469e-05,
+      "loss": 3.1817,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7072093641926044,
+      "grad_norm": 1.0737982463024023,
+      "learning_rate": 9.449170703963302e-05,
+      "loss": 3.1308,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7074221867517957,
+      "grad_norm": 1.371245211553948,
+      "learning_rate": 9.448605501726664e-05,
+      "loss": 3.1629,
+      "step": 3324
+    },
+    {
+      "epoch": 0.707635009310987,
+      "grad_norm": 1.2786605694755475,
+      "learning_rate": 9.44804002658523e-05,
+      "loss": 3.1553,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7078478318701782,
+      "grad_norm": 1.0819304087126167,
+      "learning_rate": 9.447474278573685e-05,
+      "loss": 3.1897,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7080606544293695,
+      "grad_norm": 1.2842264659624651,
+      "learning_rate": 9.44690825772674e-05,
+      "loss": 3.1822,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7082734769885608,
+      "grad_norm": 1.1070222360922013,
+      "learning_rate": 9.446341964079116e-05,
+      "loss": 3.1818,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7084862995477521,
+      "grad_norm": 1.3910937324612742,
+      "learning_rate": 9.445775397665554e-05,
+      "loss": 3.1983,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7086991221069433,
+      "grad_norm": 0.902712917399487,
+      "learning_rate": 9.445208558520809e-05,
+      "loss": 3.1212,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7089119446661346,
+      "grad_norm": 1.1531526551565772,
+      "learning_rate": 9.444641446679656e-05,
+      "loss": 3.2096,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7091247672253259,
+      "grad_norm": 1.299039607740832,
+      "learning_rate": 9.444074062176884e-05,
+      "loss": 3.1568,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7093375897845171,
+      "grad_norm": 0.9092521058107336,
+      "learning_rate": 9.443506405047303e-05,
+      "loss": 3.1596,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7095504123437084,
+      "grad_norm": 1.1406152522943485,
+      "learning_rate": 9.442938475325732e-05,
+      "loss": 3.1918,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7097632349028997,
+      "grad_norm": 1.2880335887977197,
+      "learning_rate": 9.442370273047011e-05,
+      "loss": 3.1612,
+      "step": 3335
+    },
+    {
+      "epoch": 0.709976057462091,
+      "grad_norm": 0.9402036335426285,
+      "learning_rate": 9.441801798246002e-05,
+      "loss": 3.1766,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7101888800212822,
+      "grad_norm": 1.261514631008428,
+      "learning_rate": 9.441233050957575e-05,
+      "loss": 3.1422,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7104017025804735,
+      "grad_norm": 1.1171794716269126,
+      "learning_rate": 9.440664031216622e-05,
+      "loss": 3.1911,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7106145251396648,
+      "grad_norm": 1.3255567760198226,
+      "learning_rate": 9.44009473905805e-05,
+      "loss": 3.2306,
+      "step": 3339
+    },
+    {
+      "epoch": 0.710827347698856,
+      "grad_norm": 1.0239770685402323,
+      "learning_rate": 9.439525174516782e-05,
+      "loss": 3.1638,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7110401702580473,
+      "grad_norm": 1.398737935338933,
+      "learning_rate": 9.43895533762776e-05,
+      "loss": 3.1596,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7112529928172386,
+      "grad_norm": 0.9589544784260385,
+      "learning_rate": 9.438385228425938e-05,
+      "loss": 3.1348,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7114658153764299,
+      "grad_norm": 1.3197791361602824,
+      "learning_rate": 9.437814846946296e-05,
+      "loss": 3.1309,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7116786379356211,
+      "grad_norm": 1.063686839583903,
+      "learning_rate": 9.437244193223818e-05,
+      "loss": 3.1614,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7118914604948124,
+      "grad_norm": 1.347614085745624,
+      "learning_rate": 9.436673267293516e-05,
+      "loss": 3.174,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7121042830540038,
+      "grad_norm": 1.0193371818095744,
+      "learning_rate": 9.436102069190414e-05,
+      "loss": 3.1718,
+      "step": 3346
+    },
+    {
+      "epoch": 0.712317105613195,
+      "grad_norm": 1.4287027068845206,
+      "learning_rate": 9.435530598949549e-05,
+      "loss": 3.1747,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7125299281723863,
+      "grad_norm": 1.1291356597663504,
+      "learning_rate": 9.434958856605982e-05,
+      "loss": 3.1832,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7127427507315776,
+      "grad_norm": 1.121291449780691,
+      "learning_rate": 9.434386842194785e-05,
+      "loss": 3.1368,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7129555732907689,
+      "grad_norm": 1.1995270933946027,
+      "learning_rate": 9.433814555751052e-05,
+      "loss": 3.1758,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7131683958499601,
+      "grad_norm": 1.0334614318521067,
+      "learning_rate": 9.433241997309888e-05,
+      "loss": 3.2148,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7133812184091514,
+      "grad_norm": 1.2935319599091752,
+      "learning_rate": 9.432669166906416e-05,
+      "loss": 3.1772,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7135940409683427,
+      "grad_norm": 1.1058106052592593,
+      "learning_rate": 9.432096064575781e-05,
+      "loss": 3.1656,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7138068635275339,
+      "grad_norm": 1.1245692165514507,
+      "learning_rate": 9.431522690353137e-05,
+      "loss": 3.1575,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7140196860867252,
+      "grad_norm": 1.1218048016990252,
+      "learning_rate": 9.43094904427366e-05,
+      "loss": 3.133,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7142325086459165,
+      "grad_norm": 1.2461423425998328,
+      "learning_rate": 9.430375126372542e-05,
+      "loss": 3.1745,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7144453312051078,
+      "grad_norm": 1.1364824343956292,
+      "learning_rate": 9.429800936684988e-05,
+      "loss": 3.1744,
+      "step": 3357
+    },
+    {
+      "epoch": 0.714658153764299,
+      "grad_norm": 1.1760037318377516,
+      "learning_rate": 9.429226475246222e-05,
+      "loss": 3.1404,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7148709763234903,
+      "grad_norm": 1.211524396643962,
+      "learning_rate": 9.42865174209149e-05,
+      "loss": 3.1989,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7150837988826816,
+      "grad_norm": 1.230550927142487,
+      "learning_rate": 9.428076737256045e-05,
+      "loss": 3.1787,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7152966214418729,
+      "grad_norm": 1.1209536553739987,
+      "learning_rate": 9.427501460775161e-05,
+      "loss": 3.1654,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7155094440010641,
+      "grad_norm": 1.1502596394825548,
+      "learning_rate": 9.426925912684132e-05,
+      "loss": 3.1722,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7157222665602554,
+      "grad_norm": 1.2181136324808801,
+      "learning_rate": 9.426350093018263e-05,
+      "loss": 3.1985,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7159350891194467,
+      "grad_norm": 1.0093585275181332,
+      "learning_rate": 9.425774001812881e-05,
+      "loss": 3.1604,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7161479116786379,
+      "grad_norm": 1.3487231064332021,
+      "learning_rate": 9.425197639103327e-05,
+      "loss": 3.1444,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7163607342378292,
+      "grad_norm": 1.046167400079857,
+      "learning_rate": 9.424621004924955e-05,
+      "loss": 3.1388,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7165735567970205,
+      "grad_norm": 1.2594252771776608,
+      "learning_rate": 9.424044099313142e-05,
+      "loss": 3.2086,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7167863793562118,
+      "grad_norm": 1.0589208650984636,
+      "learning_rate": 9.423466922303279e-05,
+      "loss": 3.1893,
+      "step": 3368
+    },
+    {
+      "epoch": 0.716999201915403,
+      "grad_norm": 1.2904020622392136,
+      "learning_rate": 9.422889473930772e-05,
+      "loss": 3.1842,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7172120244745943,
+      "grad_norm": 1.1125240309699787,
+      "learning_rate": 9.422311754231046e-05,
+      "loss": 3.1499,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7174248470337856,
+      "grad_norm": 1.1864262375832795,
+      "learning_rate": 9.421733763239543e-05,
+      "loss": 3.1504,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7176376695929768,
+      "grad_norm": 1.119446990781076,
+      "learning_rate": 9.42115550099172e-05,
+      "loss": 3.2346,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7178504921521681,
+      "grad_norm": 1.3573516835320132,
+      "learning_rate": 9.420576967523049e-05,
+      "loss": 3.1501,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7180633147113594,
+      "grad_norm": 1.2146820472447126,
+      "learning_rate": 9.419998162869024e-05,
+      "loss": 3.1672,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7182761372705507,
+      "grad_norm": 0.9036012711791509,
+      "learning_rate": 9.41941908706515e-05,
+      "loss": 3.158,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7184889598297419,
+      "grad_norm": 1.1008377777357128,
+      "learning_rate": 9.418839740146955e-05,
+      "loss": 3.1377,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7187017823889332,
+      "grad_norm": 1.3634983013170316,
+      "learning_rate": 9.418260122149974e-05,
+      "loss": 3.1833,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7189146049481245,
+      "grad_norm": 0.9713846472349524,
+      "learning_rate": 9.417680233109767e-05,
+      "loss": 3.1176,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7191274275073157,
+      "grad_norm": 1.3271434849890682,
+      "learning_rate": 9.417100073061909e-05,
+      "loss": 3.17,
+      "step": 3379
+    },
+    {
+      "epoch": 0.719340250066507,
+      "grad_norm": 1.0966037465194138,
+      "learning_rate": 9.416519642041989e-05,
+      "loss": 3.1179,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7195530726256983,
+      "grad_norm": 1.2447271727514313,
+      "learning_rate": 9.415938940085616e-05,
+      "loss": 3.1561,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7197658951848896,
+      "grad_norm": 1.0011744629660866,
+      "learning_rate": 9.415357967228411e-05,
+      "loss": 3.1689,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7199787177440808,
+      "grad_norm": 1.139010573021442,
+      "learning_rate": 9.414776723506017e-05,
+      "loss": 3.1333,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7201915403032721,
+      "grad_norm": 1.1667506543638726,
+      "learning_rate": 9.41419520895409e-05,
+      "loss": 3.1329,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7204043628624635,
+      "grad_norm": 1.1341877504299782,
+      "learning_rate": 9.413613423608305e-05,
+      "loss": 3.1535,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7206171854216546,
+      "grad_norm": 1.387604180345176,
+      "learning_rate": 9.413031367504351e-05,
+      "loss": 3.1738,
+      "step": 3386
+    },
+    {
+      "epoch": 0.720830007980846,
+      "grad_norm": 0.8789713786926409,
+      "learning_rate": 9.412449040677934e-05,
+      "loss": 3.2108,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7210428305400373,
+      "grad_norm": 1.2603931916119047,
+      "learning_rate": 9.411866443164779e-05,
+      "loss": 3.19,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7212556530992286,
+      "grad_norm": 1.1009092099787372,
+      "learning_rate": 9.411283575000625e-05,
+      "loss": 3.1738,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7214684756584198,
+      "grad_norm": 1.1614584986236143,
+      "learning_rate": 9.410700436221229e-05,
+      "loss": 3.1501,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7216812982176111,
+      "grad_norm": 1.1899094962238317,
+      "learning_rate": 9.410117026862366e-05,
+      "loss": 3.1698,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7218941207768024,
+      "grad_norm": 1.1737182633379564,
+      "learning_rate": 9.409533346959825e-05,
+      "loss": 3.1872,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7221069433359936,
+      "grad_norm": 1.0397721594084093,
+      "learning_rate": 9.408949396549412e-05,
+      "loss": 3.1486,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7223197658951849,
+      "grad_norm": 1.1250450390433928,
+      "learning_rate": 9.40836517566695e-05,
+      "loss": 3.1662,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7225325884543762,
+      "grad_norm": 1.0847595476931449,
+      "learning_rate": 9.407780684348277e-05,
+      "loss": 3.1228,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7227454110135675,
+      "grad_norm": 1.3687370471771374,
+      "learning_rate": 9.407195922629253e-05,
+      "loss": 3.1757,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7229582335727587,
+      "grad_norm": 1.0995267066254168,
+      "learning_rate": 9.406610890545748e-05,
+      "loss": 3.1351,
+      "step": 3397
+    },
+    {
+      "epoch": 0.72317105613195,
+      "grad_norm": 1.4404205156852594,
+      "learning_rate": 9.406025588133654e-05,
+      "loss": 3.1216,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7233838786911413,
+      "grad_norm": 0.9963385357121622,
+      "learning_rate": 9.405440015428875e-05,
+      "loss": 3.1732,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7235967012503325,
+      "grad_norm": 1.409416179672355,
+      "learning_rate": 9.404854172467334e-05,
+      "loss": 3.1683,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 0.9683843287264052,
+      "learning_rate": 9.40426805928497e-05,
+      "loss": 3.1772,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7240223463687151,
+      "grad_norm": 1.3643625561896107,
+      "learning_rate": 9.40368167591774e-05,
+      "loss": 3.1735,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7242351689279064,
+      "grad_norm": 0.9380725970266202,
+      "learning_rate": 9.403095022401615e-05,
+      "loss": 3.1465,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7244479914870976,
+      "grad_norm": 1.3601270809767814,
+      "learning_rate": 9.402508098772587e-05,
+      "loss": 3.2195,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7246608140462889,
+      "grad_norm": 1.0987333659661054,
+      "learning_rate": 9.401920905066657e-05,
+      "loss": 3.1645,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7248736366054802,
+      "grad_norm": 1.1014432181125553,
+      "learning_rate": 9.401333441319849e-05,
+      "loss": 3.1295,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7250864591646714,
+      "grad_norm": 1.4694345467175873,
+      "learning_rate": 9.400745707568201e-05,
+      "loss": 3.1597,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7252992817238627,
+      "grad_norm": 0.9063436387430895,
+      "learning_rate": 9.400157703847769e-05,
+      "loss": 3.1252,
+      "step": 3408
+    },
+    {
+      "epoch": 0.725512104283054,
+      "grad_norm": 1.1646995972645102,
+      "learning_rate": 9.399569430194626e-05,
+      "loss": 3.181,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7257249268422453,
+      "grad_norm": 1.1954932544346695,
+      "learning_rate": 9.398980886644859e-05,
+      "loss": 3.1669,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7259377494014365,
+      "grad_norm": 1.197742104985511,
+      "learning_rate": 9.398392073234572e-05,
+      "loss": 3.1897,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7261505719606278,
+      "grad_norm": 1.1286257789559022,
+      "learning_rate": 9.397802989999887e-05,
+      "loss": 3.152,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7263633945198191,
+      "grad_norm": 1.3759539713019007,
+      "learning_rate": 9.397213636976944e-05,
+      "loss": 3.1728,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7265762170790103,
+      "grad_norm": 0.9143843827420811,
+      "learning_rate": 9.396624014201894e-05,
+      "loss": 3.1419,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7267890396382016,
+      "grad_norm": 1.1466432493725514,
+      "learning_rate": 9.396034121710911e-05,
+      "loss": 3.1808,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7270018621973929,
+      "grad_norm": 1.1490664867683222,
+      "learning_rate": 9.395443959540182e-05,
+      "loss": 3.1451,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7272146847565842,
+      "grad_norm": 1.2159220317521577,
+      "learning_rate": 9.39485352772591e-05,
+      "loss": 3.1493,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7274275073157754,
+      "grad_norm": 1.0641231290107396,
+      "learning_rate": 9.394262826304317e-05,
+      "loss": 3.113,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7276403298749667,
+      "grad_norm": 1.1701259385068656,
+      "learning_rate": 9.393671855311638e-05,
+      "loss": 3.1395,
+      "step": 3419
+    },
+    {
+      "epoch": 0.727853152434158,
+      "grad_norm": 1.1306964493119118,
+      "learning_rate": 9.39308061478413e-05,
+      "loss": 3.17,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7280659749933492,
+      "grad_norm": 1.2895147598176546,
+      "learning_rate": 9.392489104758063e-05,
+      "loss": 3.1489,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7282787975525405,
+      "grad_norm": 1.0392126327045967,
+      "learning_rate": 9.39189732526972e-05,
+      "loss": 3.1751,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7284916201117319,
+      "grad_norm": 1.211585817414838,
+      "learning_rate": 9.391305276355407e-05,
+      "loss": 3.1438,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7287044426709232,
+      "grad_norm": 1.121522706283879,
+      "learning_rate": 9.390712958051446e-05,
+      "loss": 3.1422,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7289172652301144,
+      "grad_norm": 1.1626261442499757,
+      "learning_rate": 9.39012037039417e-05,
+      "loss": 3.1734,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7291300877893057,
+      "grad_norm": 1.1190966485103058,
+      "learning_rate": 9.389527513419934e-05,
+      "loss": 3.1467,
+      "step": 3426
+    },
+    {
+      "epoch": 0.729342910348497,
+      "grad_norm": 1.2795271781120852,
+      "learning_rate": 9.388934387165107e-05,
+      "loss": 3.1314,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7295557329076883,
+      "grad_norm": 1.0743467183249473,
+      "learning_rate": 9.388340991666076e-05,
+      "loss": 3.1453,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7297685554668795,
+      "grad_norm": 1.0931175802453548,
+      "learning_rate": 9.38774732695924e-05,
+      "loss": 3.1835,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7299813780260708,
+      "grad_norm": 1.2960724877846679,
+      "learning_rate": 9.387153393081023e-05,
+      "loss": 3.1509,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7301942005852621,
+      "grad_norm": 0.9910288380734519,
+      "learning_rate": 9.386559190067856e-05,
+      "loss": 3.1886,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7304070231444533,
+      "grad_norm": 1.377015905667843,
+      "learning_rate": 9.385964717956195e-05,
+      "loss": 3.1743,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7306198457036446,
+      "grad_norm": 1.0278162011866,
+      "learning_rate": 9.385369976782505e-05,
+      "loss": 3.173,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7308326682628359,
+      "grad_norm": 1.364694070086457,
+      "learning_rate": 9.384774966583273e-05,
+      "loss": 3.1168,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7310454908220272,
+      "grad_norm": 0.9974553657309053,
+      "learning_rate": 9.384179687395e-05,
+      "loss": 3.1246,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7312583133812184,
+      "grad_norm": 1.2460312422586417,
+      "learning_rate": 9.383584139254205e-05,
+      "loss": 3.1918,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7314711359404097,
+      "grad_norm": 1.1341086636049504,
+      "learning_rate": 9.382988322197422e-05,
+      "loss": 3.1391,
+      "step": 3437
+    },
+    {
+      "epoch": 0.731683958499601,
+      "grad_norm": 1.260418013718309,
+      "learning_rate": 9.382392236261201e-05,
+      "loss": 3.1642,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7318967810587922,
+      "grad_norm": 0.9793678523394512,
+      "learning_rate": 9.381795881482115e-05,
+      "loss": 3.16,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7321096036179835,
+      "grad_norm": 1.2073664748050548,
+      "learning_rate": 9.381199257896738e-05,
+      "loss": 3.1305,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7323224261771748,
+      "grad_norm": 1.245119123671497,
+      "learning_rate": 9.38060236554168e-05,
+      "loss": 3.1809,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7325352487363661,
+      "grad_norm": 1.1538435968793677,
+      "learning_rate": 9.380005204453553e-05,
+      "loss": 3.1553,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7327480712955573,
+      "grad_norm": 1.0376270541264592,
+      "learning_rate": 9.379407774668991e-05,
+      "loss": 3.1381,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7329608938547486,
+      "grad_norm": 1.3208528787146434,
+      "learning_rate": 9.378810076224645e-05,
+      "loss": 3.1812,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7331737164139399,
+      "grad_norm": 1.036251873539309,
+      "learning_rate": 9.378212109157183e-05,
+      "loss": 3.1583,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7333865389731311,
+      "grad_norm": 1.2037530862905665,
+      "learning_rate": 9.377613873503284e-05,
+      "loss": 3.1872,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7335993615323224,
+      "grad_norm": 1.1813875927417203,
+      "learning_rate": 9.377015369299651e-05,
+      "loss": 3.1684,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7338121840915137,
+      "grad_norm": 1.1140325473647459,
+      "learning_rate": 9.376416596582999e-05,
+      "loss": 3.2026,
+      "step": 3448
+    },
+    {
+      "epoch": 0.734025006650705,
+      "grad_norm": 1.3127734574955057,
+      "learning_rate": 9.37581755539006e-05,
+      "loss": 3.1719,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7342378292098962,
+      "grad_norm": 0.9965909273986195,
+      "learning_rate": 9.375218245757581e-05,
+      "loss": 3.1781,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7344506517690875,
+      "grad_norm": 1.2582185396645293,
+      "learning_rate": 9.374618667722333e-05,
+      "loss": 3.1779,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7346634743282788,
+      "grad_norm": 1.1340999750896683,
+      "learning_rate": 9.374018821321092e-05,
+      "loss": 3.1231,
+      "step": 3452
+    },
+    {
+      "epoch": 0.73487629688747,
+      "grad_norm": 1.2563759651580337,
+      "learning_rate": 9.373418706590658e-05,
+      "loss": 3.1368,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7350891194466613,
+      "grad_norm": 1.1090656992145842,
+      "learning_rate": 9.372818323567847e-05,
+      "loss": 3.1717,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7353019420058526,
+      "grad_norm": 1.2219771038586624,
+      "learning_rate": 9.37221767228949e-05,
+      "loss": 3.17,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7355147645650439,
+      "grad_norm": 1.3263346670092986,
+      "learning_rate": 9.371616752792431e-05,
+      "loss": 3.1526,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7357275871242351,
+      "grad_norm": 1.0982106957609548,
+      "learning_rate": 9.37101556511354e-05,
+      "loss": 3.1687,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7359404096834264,
+      "grad_norm": 1.1101099385950917,
+      "learning_rate": 9.370414109289694e-05,
+      "loss": 3.1584,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7361532322426177,
+      "grad_norm": 1.3433197832082504,
+      "learning_rate": 9.369812385357789e-05,
+      "loss": 3.123,
+      "step": 3459
+    },
+    {
+      "epoch": 0.736366054801809,
+      "grad_norm": 1.0777546058828258,
+      "learning_rate": 9.369210393354742e-05,
+      "loss": 3.1976,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7365788773610003,
+      "grad_norm": 1.2144882490104123,
+      "learning_rate": 9.36860813331748e-05,
+      "loss": 3.143,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7367916999201916,
+      "grad_norm": 1.0011397154588204,
+      "learning_rate": 9.368005605282949e-05,
+      "loss": 3.1543,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7370045224793829,
+      "grad_norm": 1.3361972404099272,
+      "learning_rate": 9.367402809288114e-05,
+      "loss": 3.1247,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7372173450385741,
+      "grad_norm": 1.1599734910415451,
+      "learning_rate": 9.366799745369953e-05,
+      "loss": 3.1446,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7374301675977654,
+      "grad_norm": 1.2468163207955743,
+      "learning_rate": 9.366196413565462e-05,
+      "loss": 3.1771,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7376429901569567,
+      "grad_norm": 1.032835656932011,
+      "learning_rate": 9.365592813911653e-05,
+      "loss": 3.1745,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7378558127161479,
+      "grad_norm": 1.4708225820658674,
+      "learning_rate": 9.364988946445554e-05,
+      "loss": 3.1239,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7380686352753392,
+      "grad_norm": 0.9145116392145062,
+      "learning_rate": 9.364384811204211e-05,
+      "loss": 3.1776,
+      "step": 3468
+    },
+    {
+      "epoch": 0.7382814578345305,
+      "grad_norm": 1.321702534637332,
+      "learning_rate": 9.363780408224686e-05,
+      "loss": 3.1396,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7384942803937218,
+      "grad_norm": 0.9721856388437469,
+      "learning_rate": 9.363175737544053e-05,
+      "loss": 3.1265,
+      "step": 3470
+    },
+    {
+      "epoch": 0.738707102952913,
+      "grad_norm": 1.2724783423057668,
+      "learning_rate": 9.36257079919941e-05,
+      "loss": 3.1325,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7389199255121043,
+      "grad_norm": 1.1166477951244302,
+      "learning_rate": 9.361965593227867e-05,
+      "loss": 3.133,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7391327480712956,
+      "grad_norm": 1.181002474781288,
+      "learning_rate": 9.361360119666549e-05,
+      "loss": 3.174,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7393455706304868,
+      "grad_norm": 1.3513528008096263,
+      "learning_rate": 9.360754378552601e-05,
+      "loss": 3.1608,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7395583931896781,
+      "grad_norm": 1.0100772298767737,
+      "learning_rate": 9.360148369923185e-05,
+      "loss": 3.1536,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7397712157488694,
+      "grad_norm": 1.2847296789612188,
+      "learning_rate": 9.359542093815472e-05,
+      "loss": 3.1688,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7399840383080607,
+      "grad_norm": 1.0220320858868537,
+      "learning_rate": 9.358935550266659e-05,
+      "loss": 3.1444,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7401968608672519,
+      "grad_norm": 1.3486599135075332,
+      "learning_rate": 9.358328739313955e-05,
+      "loss": 3.1485,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7404096834264432,
+      "grad_norm": 0.9457128002698075,
+      "learning_rate": 9.357721660994583e-05,
+      "loss": 3.1855,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7406225059856345,
+      "grad_norm": 1.3056200465888554,
+      "learning_rate": 9.357114315345788e-05,
+      "loss": 3.1419,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7408353285448257,
+      "grad_norm": 1.0843617712816198,
+      "learning_rate": 9.356506702404823e-05,
+      "loss": 3.1302,
+      "step": 3481
+    },
+    {
+      "epoch": 0.741048151104017,
+      "grad_norm": 1.1773483222653374,
+      "learning_rate": 9.35589882220897e-05,
+      "loss": 3.1471,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7412609736632083,
+      "grad_norm": 1.1395849417405683,
+      "learning_rate": 9.355290674795516e-05,
+      "loss": 3.1295,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7414737962223996,
+      "grad_norm": 1.3183350024845937,
+      "learning_rate": 9.354682260201769e-05,
+      "loss": 3.1004,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7416866187815908,
+      "grad_norm": 1.0850325830400398,
+      "learning_rate": 9.354073578465051e-05,
+      "loss": 3.1894,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7418994413407821,
+      "grad_norm": 1.3130705954874315,
+      "learning_rate": 9.353464629622706e-05,
+      "loss": 3.1108,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7421122638999734,
+      "grad_norm": 1.041008788555277,
+      "learning_rate": 9.352855413712088e-05,
+      "loss": 3.124,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7423250864591646,
+      "grad_norm": 1.3640340183709714,
+      "learning_rate": 9.35224593077057e-05,
+      "loss": 3.1552,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7425379090183559,
+      "grad_norm": 1.1209842230142875,
+      "learning_rate": 9.351636180835543e-05,
+      "loss": 3.1538,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7427507315775472,
+      "grad_norm": 1.2039332428832525,
+      "learning_rate": 9.351026163944412e-05,
+      "loss": 3.1416,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7429635541367385,
+      "grad_norm": 1.2041839852632927,
+      "learning_rate": 9.350415880134599e-05,
+      "loss": 3.137,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7431763766959297,
+      "grad_norm": 0.9541074859675166,
+      "learning_rate": 9.349805329443544e-05,
+      "loss": 3.1477,
+      "step": 3492
+    },
+    {
+      "epoch": 0.743389199255121,
+      "grad_norm": 1.1691848096792086,
+      "learning_rate": 9.349194511908701e-05,
+      "loss": 3.1966,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7436020218143123,
+      "grad_norm": 1.0929180778654888,
+      "learning_rate": 9.348583427567539e-05,
+      "loss": 3.1382,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7438148443735035,
+      "grad_norm": 1.2439293432684677,
+      "learning_rate": 9.34797207645755e-05,
+      "loss": 3.1252,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7440276669326948,
+      "grad_norm": 1.0235618728361726,
+      "learning_rate": 9.347360458616233e-05,
+      "loss": 3.1747,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7442404894918861,
+      "grad_norm": 1.2721660468048754,
+      "learning_rate": 9.346748574081112e-05,
+      "loss": 3.1611,
+      "step": 3497
+    },
+    {
+      "epoch": 0.7444533120510775,
+      "grad_norm": 1.036162839735352,
+      "learning_rate": 9.346136422889724e-05,
+      "loss": 3.1307,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7446661346102686,
+      "grad_norm": 1.216818121569334,
+      "learning_rate": 9.345524005079621e-05,
+      "loss": 3.1445,
+      "step": 3499
+    },
+    {
+      "epoch": 0.74487895716946,
+      "grad_norm": 1.2259195749966325,
+      "learning_rate": 9.344911320688372e-05,
+      "loss": 3.1374,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7450917797286513,
+      "grad_norm": 1.0688302633726394,
+      "learning_rate": 9.344298369753564e-05,
+      "loss": 3.1175,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7453046022878426,
+      "grad_norm": 1.148970381284232,
+      "learning_rate": 9.343685152312797e-05,
+      "loss": 3.1386,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7455174248470338,
+      "grad_norm": 1.3189273958565262,
+      "learning_rate": 9.343071668403692e-05,
+      "loss": 3.1253,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7457302474062251,
+      "grad_norm": 0.9093263218001914,
+      "learning_rate": 9.342457918063882e-05,
+      "loss": 3.1204,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7459430699654164,
+      "grad_norm": 1.0858081252699898,
+      "learning_rate": 9.341843901331021e-05,
+      "loss": 3.1294,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7461558925246076,
+      "grad_norm": 1.2125701749566333,
+      "learning_rate": 9.341229618242774e-05,
+      "loss": 3.1287,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7463687150837989,
+      "grad_norm": 1.0431564466562873,
+      "learning_rate": 9.340615068836824e-05,
+      "loss": 3.1711,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7465815376429902,
+      "grad_norm": 1.308010314629898,
+      "learning_rate": 9.340000253150876e-05,
+      "loss": 3.1567,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7467943602021815,
+      "grad_norm": 1.086200378157022,
+      "learning_rate": 9.339385171222641e-05,
+      "loss": 3.2036,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7470071827613727,
+      "grad_norm": 1.1964148117226523,
+      "learning_rate": 9.338769823089853e-05,
+      "loss": 3.1416,
+      "step": 3510
+    },
+    {
+      "epoch": 0.747220005320564,
+      "grad_norm": 1.1363609819063591,
+      "learning_rate": 9.338154208790265e-05,
+      "loss": 3.1962,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7474328278797553,
+      "grad_norm": 1.25374893056187,
+      "learning_rate": 9.337538328361639e-05,
+      "loss": 3.1252,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7476456504389465,
+      "grad_norm": 1.1465215142838496,
+      "learning_rate": 9.336922181841759e-05,
+      "loss": 3.1253,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7478584729981378,
+      "grad_norm": 1.0651054636892843,
+      "learning_rate": 9.336305769268421e-05,
+      "loss": 3.1576,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7480712955573291,
+      "grad_norm": 1.0261040528034955,
+      "learning_rate": 9.33568909067944e-05,
+      "loss": 3.1126,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7482841181165204,
+      "grad_norm": 1.1576474087477944,
+      "learning_rate": 9.335072146112648e-05,
+      "loss": 3.141,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7484969406757116,
+      "grad_norm": 1.187035675578085,
+      "learning_rate": 9.334454935605894e-05,
+      "loss": 3.1264,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7487097632349029,
+      "grad_norm": 1.2238844873881403,
+      "learning_rate": 9.333837459197038e-05,
+      "loss": 3.1382,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7489225857940942,
+      "grad_norm": 1.0112207252789043,
+      "learning_rate": 9.33321971692396e-05,
+      "loss": 3.1378,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7491354083532854,
+      "grad_norm": 1.3079295177031458,
+      "learning_rate": 9.332601708824558e-05,
+      "loss": 3.1767,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7493482309124767,
+      "grad_norm": 1.1296200758531452,
+      "learning_rate": 9.331983434936745e-05,
+      "loss": 3.1294,
+      "step": 3521
+    },
+    {
+      "epoch": 0.749561053471668,
+      "grad_norm": 1.2337040255236398,
+      "learning_rate": 9.331364895298444e-05,
+      "loss": 3.1403,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7497738760308593,
+      "grad_norm": 0.9581597349067922,
+      "learning_rate": 9.330746089947609e-05,
+      "loss": 3.1432,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7499866985900505,
+      "grad_norm": 1.3030511664796733,
+      "learning_rate": 9.330127018922194e-05,
+      "loss": 3.1346,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7501995211492418,
+      "grad_norm": 1.0986407162774599,
+      "learning_rate": 9.32950768226018e-05,
+      "loss": 3.167,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7504123437084331,
+      "grad_norm": 1.2204062716957693,
+      "learning_rate": 9.32888807999956e-05,
+      "loss": 3.1139,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7506251662676243,
+      "grad_norm": 1.1759072959268975,
+      "learning_rate": 9.328268212178343e-05,
+      "loss": 3.1391,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7508379888268156,
+      "grad_norm": 1.1342903340510995,
+      "learning_rate": 9.327648078834558e-05,
+      "loss": 3.1621,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7510508113860069,
+      "grad_norm": 1.3791050934323457,
+      "learning_rate": 9.327027680006248e-05,
+      "loss": 3.117,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7512636339451982,
+      "grad_norm": 0.9370023919670392,
+      "learning_rate": 9.326407015731468e-05,
+      "loss": 3.1176,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7514764565043894,
+      "grad_norm": 1.105960982874337,
+      "learning_rate": 9.325786086048299e-05,
+      "loss": 3.1705,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7516892790635807,
+      "grad_norm": 1.2726613402277744,
+      "learning_rate": 9.325164890994828e-05,
+      "loss": 3.1633,
+      "step": 3532
+    },
+    {
+      "epoch": 0.751902101622772,
+      "grad_norm": 0.9810479282103401,
+      "learning_rate": 9.324543430609165e-05,
+      "loss": 3.0952,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7521149241819632,
+      "grad_norm": 1.3359771903878144,
+      "learning_rate": 9.323921704929434e-05,
+      "loss": 3.1436,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7523277467411545,
+      "grad_norm": 0.968562758169872,
+      "learning_rate": 9.323299713993776e-05,
+      "loss": 3.128,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7525405693003459,
+      "grad_norm": 1.3628393405484305,
+      "learning_rate": 9.322677457840346e-05,
+      "loss": 3.1157,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7527533918595372,
+      "grad_norm": 0.996440305085063,
+      "learning_rate": 9.322054936507319e-05,
+      "loss": 3.1335,
+      "step": 3537
+    },
+    {
+      "epoch": 0.7529662144187284,
+      "grad_norm": 1.28067369891306,
+      "learning_rate": 9.321432150032883e-05,
+      "loss": 3.1642,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7531790369779197,
+      "grad_norm": 1.0786411024551286,
+      "learning_rate": 9.320809098455243e-05,
+      "loss": 3.1263,
+      "step": 3539
+    },
+    {
+      "epoch": 0.753391859537111,
+      "grad_norm": 1.3973086820521587,
+      "learning_rate": 9.320185781812623e-05,
+      "loss": 3.1701,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7536046820963022,
+      "grad_norm": 1.0390855453261523,
+      "learning_rate": 9.319562200143258e-05,
+      "loss": 3.1882,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7538175046554935,
+      "grad_norm": 1.1745173514517284,
+      "learning_rate": 9.318938353485404e-05,
+      "loss": 3.1399,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7540303272146848,
+      "grad_norm": 1.177596908870592,
+      "learning_rate": 9.318314241877335e-05,
+      "loss": 3.121,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7542431497738761,
+      "grad_norm": 1.0304364558634935,
+      "learning_rate": 9.317689865357333e-05,
+      "loss": 3.1416,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7544559723330673,
+      "grad_norm": 1.217238437094838,
+      "learning_rate": 9.317065223963702e-05,
+      "loss": 3.1345,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7546687948922586,
+      "grad_norm": 0.9945142134543115,
+      "learning_rate": 9.316440317734763e-05,
+      "loss": 3.1608,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7548816174514499,
+      "grad_norm": 1.182183576184337,
+      "learning_rate": 9.315815146708848e-05,
+      "loss": 3.09,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7550944400106411,
+      "grad_norm": 1.2348882292364203,
+      "learning_rate": 9.315189710924315e-05,
+      "loss": 3.1056,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7553072625698324,
+      "grad_norm": 1.1225217711252384,
+      "learning_rate": 9.314564010419527e-05,
+      "loss": 3.1735,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7555200851290237,
+      "grad_norm": 1.2813877575110153,
+      "learning_rate": 9.313938045232869e-05,
+      "loss": 3.0751,
+      "step": 3550
+    },
+    {
+      "epoch": 0.755732907688215,
+      "grad_norm": 0.8936253772042932,
+      "learning_rate": 9.313311815402743e-05,
+      "loss": 3.1246,
+      "step": 3551
+    },
+    {
+      "epoch": 0.7559457302474062,
+      "grad_norm": 1.2441762808784065,
+      "learning_rate": 9.312685320967564e-05,
+      "loss": 3.1852,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7561585528065975,
+      "grad_norm": 1.1441826050504837,
+      "learning_rate": 9.312058561965768e-05,
+      "loss": 3.1331,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7563713753657888,
+      "grad_norm": 1.1325542879591015,
+      "learning_rate": 9.311431538435802e-05,
+      "loss": 3.1426,
+      "step": 3554
+    },
+    {
+      "epoch": 0.75658419792498,
+      "grad_norm": 1.147765036351735,
+      "learning_rate": 9.310804250416131e-05,
+      "loss": 3.1302,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7567970204841713,
+      "grad_norm": 1.1976216696043869,
+      "learning_rate": 9.310176697945239e-05,
+      "loss": 3.1467,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7570098430433626,
+      "grad_norm": 1.1537925257392843,
+      "learning_rate": 9.309548881061621e-05,
+      "loss": 3.0936,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7572226656025539,
+      "grad_norm": 1.0943922458238802,
+      "learning_rate": 9.308920799803793e-05,
+      "loss": 3.1665,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7574354881617451,
+      "grad_norm": 1.2348890344414296,
+      "learning_rate": 9.308292454210286e-05,
+      "loss": 3.1639,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7576483107209364,
+      "grad_norm": 1.0135547051619664,
+      "learning_rate": 9.307663844319645e-05,
+      "loss": 3.1569,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7578611332801277,
+      "grad_norm": 1.2403430950645875,
+      "learning_rate": 9.307034970170434e-05,
+      "loss": 3.1095,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7580739558393189,
+      "grad_norm": 1.0820643815764373,
+      "learning_rate": 9.30640583180123e-05,
+      "loss": 3.1651,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7582867783985102,
+      "grad_norm": 1.2240620670373403,
+      "learning_rate": 9.30577642925063e-05,
+      "loss": 3.1492,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7584996009577015,
+      "grad_norm": 1.0277035219357151,
+      "learning_rate": 9.305146762557246e-05,
+      "loss": 3.1398,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7587124235168928,
+      "grad_norm": 1.0324398125435343,
+      "learning_rate": 9.304516831759704e-05,
+      "loss": 3.1603,
+      "step": 3565
+    },
+    {
+      "epoch": 0.758925246076084,
+      "grad_norm": 1.28648250357555,
+      "learning_rate": 9.303886636896649e-05,
+      "loss": 3.148,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7591380686352753,
+      "grad_norm": 1.0671676820408436,
+      "learning_rate": 9.303256178006741e-05,
+      "loss": 3.0633,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7593508911944666,
+      "grad_norm": 1.1897404745142115,
+      "learning_rate": 9.302625455128656e-05,
+      "loss": 3.0952,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7595637137536579,
+      "grad_norm": 0.9851987065271356,
+      "learning_rate": 9.301994468301082e-05,
+      "loss": 3.1755,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7597765363128491,
+      "grad_norm": 1.4117664158215486,
+      "learning_rate": 9.301363217562737e-05,
+      "loss": 3.1947,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7599893588720404,
+      "grad_norm": 1.0786132521833582,
+      "learning_rate": 9.300731702952337e-05,
+      "loss": 3.0961,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7602021814312317,
+      "grad_norm": 1.274407360911647,
+      "learning_rate": 9.300099924508629e-05,
+      "loss": 3.1191,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7604150039904229,
+      "grad_norm": 1.0150351092316607,
+      "learning_rate": 9.299467882270367e-05,
+      "loss": 3.0924,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7606278265496142,
+      "grad_norm": 1.2166643989262171,
+      "learning_rate": 9.298835576276324e-05,
+      "loss": 3.142,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7608406491088056,
+      "grad_norm": 1.084268241293663,
+      "learning_rate": 9.298203006565291e-05,
+      "loss": 3.1372,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7610534716679969,
+      "grad_norm": 1.249891851441728,
+      "learning_rate": 9.297570173176075e-05,
+      "loss": 3.1586,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7612662942271881,
+      "grad_norm": 1.1481547228923452,
+      "learning_rate": 9.296937076147494e-05,
+      "loss": 3.1397,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7614791167863794,
+      "grad_norm": 1.1234682174207684,
+      "learning_rate": 9.296303715518389e-05,
+      "loss": 3.1213,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7616919393455707,
+      "grad_norm": 1.173488422308225,
+      "learning_rate": 9.295670091327615e-05,
+      "loss": 3.1312,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 1.0856685220930926,
+      "learning_rate": 9.295036203614037e-05,
+      "loss": 3.1358,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7621175844639532,
+      "grad_norm": 1.220535382519675,
+      "learning_rate": 9.29440205241655e-05,
+      "loss": 3.1366,
+      "step": 3581
+    },
+    {
+      "epoch": 0.7623304070231445,
+      "grad_norm": 1.016238181058433,
+      "learning_rate": 9.29376763777405e-05,
+      "loss": 3.1258,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7625432295823358,
+      "grad_norm": 1.1933463414516474,
+      "learning_rate": 9.29313295972546e-05,
+      "loss": 3.1538,
+      "step": 3583
+    },
+    {
+      "epoch": 0.762756052141527,
+      "grad_norm": 1.2358841182261937,
+      "learning_rate": 9.29249801830971e-05,
+      "loss": 3.1191,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7629688747007183,
+      "grad_norm": 1.168450248978824,
+      "learning_rate": 9.291862813565755e-05,
+      "loss": 3.1567,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7631816972599096,
+      "grad_norm": 1.2852771202214393,
+      "learning_rate": 9.291227345532562e-05,
+      "loss": 3.1322,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7633945198191008,
+      "grad_norm": 1.062383222288822,
+      "learning_rate": 9.290591614249115e-05,
+      "loss": 3.0998,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7636073423782921,
+      "grad_norm": 1.2526040941546015,
+      "learning_rate": 9.289955619754414e-05,
+      "loss": 3.0864,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7638201649374834,
+      "grad_norm": 1.0393240359362248,
+      "learning_rate": 9.289319362087472e-05,
+      "loss": 3.1207,
+      "step": 3589
+    },
+    {
+      "epoch": 0.7640329874966747,
+      "grad_norm": 1.1687792059718427,
+      "learning_rate": 9.288682841287324e-05,
+      "loss": 3.1047,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7642458100558659,
+      "grad_norm": 1.1014546258237994,
+      "learning_rate": 9.288046057393016e-05,
+      "loss": 3.1253,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7644586326150572,
+      "grad_norm": 1.225305391626283,
+      "learning_rate": 9.287409010443612e-05,
+      "loss": 3.1335,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7646714551742485,
+      "grad_norm": 1.150371220878044,
+      "learning_rate": 9.286771700478194e-05,
+      "loss": 3.1189,
+      "step": 3593
+    },
+    {
+      "epoch": 0.7648842777334397,
+      "grad_norm": 1.1230078064344886,
+      "learning_rate": 9.286134127535859e-05,
+      "loss": 3.1004,
+      "step": 3594
+    },
+    {
+      "epoch": 0.765097100292631,
+      "grad_norm": 1.1201673076972631,
+      "learning_rate": 9.285496291655717e-05,
+      "loss": 3.0521,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7653099228518223,
+      "grad_norm": 1.0284679559689627,
+      "learning_rate": 9.2848581928769e-05,
+      "loss": 3.1182,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7655227454110136,
+      "grad_norm": 1.279315633481805,
+      "learning_rate": 9.284219831238549e-05,
+      "loss": 3.1732,
+      "step": 3597
+    },
+    {
+      "epoch": 0.7657355679702048,
+      "grad_norm": 1.039997567713725,
+      "learning_rate": 9.28358120677983e-05,
+      "loss": 3.1295,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7659483905293961,
+      "grad_norm": 1.1318560271758955,
+      "learning_rate": 9.282942319539916e-05,
+      "loss": 3.1255,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7661612130885874,
+      "grad_norm": 1.012681224566348,
+      "learning_rate": 9.282303169558e-05,
+      "loss": 3.1103,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7663740356477786,
+      "grad_norm": 1.219461131520515,
+      "learning_rate": 9.281663756873296e-05,
+      "loss": 3.1351,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7665868582069699,
+      "grad_norm": 1.0415217426086525,
+      "learning_rate": 9.281024081525024e-05,
+      "loss": 3.1406,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7667996807661612,
+      "grad_norm": 1.1498914852068456,
+      "learning_rate": 9.28038414355243e-05,
+      "loss": 3.1402,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7670125033253525,
+      "grad_norm": 1.1221360509392295,
+      "learning_rate": 9.279743942994767e-05,
+      "loss": 3.0987,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7672253258845437,
+      "grad_norm": 1.1303591729863431,
+      "learning_rate": 9.279103479891315e-05,
+      "loss": 3.0735,
+      "step": 3605
+    },
+    {
+      "epoch": 0.767438148443735,
+      "grad_norm": 1.20216725069169,
+      "learning_rate": 9.27846275428136e-05,
+      "loss": 3.1476,
+      "step": 3606
+    },
+    {
+      "epoch": 0.7676509710029263,
+      "grad_norm": 1.3198885574649573,
+      "learning_rate": 9.277821766204208e-05,
+      "loss": 3.119,
+      "step": 3607
+    },
+    {
+      "epoch": 0.7678637935621175,
+      "grad_norm": 0.9771776579906697,
+      "learning_rate": 9.277180515699182e-05,
+      "loss": 3.1102,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7680766161213088,
+      "grad_norm": 1.5311710041614413,
+      "learning_rate": 9.27653900280562e-05,
+      "loss": 3.1248,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7682894386805001,
+      "grad_norm": 0.8820098772833975,
+      "learning_rate": 9.275897227562875e-05,
+      "loss": 3.1382,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7685022612396915,
+      "grad_norm": 1.3119612117829715,
+      "learning_rate": 9.275255190010321e-05,
+      "loss": 3.103,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7687150837988826,
+      "grad_norm": 1.004553141332029,
+      "learning_rate": 9.274612890187342e-05,
+      "loss": 3.1041,
+      "step": 3612
+    },
+    {
+      "epoch": 0.768927906358074,
+      "grad_norm": 1.2117072723481528,
+      "learning_rate": 9.273970328133342e-05,
+      "loss": 3.1089,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7691407289172653,
+      "grad_norm": 0.9737423229819716,
+      "learning_rate": 9.273327503887737e-05,
+      "loss": 3.1351,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7693535514764565,
+      "grad_norm": 1.1169692214959053,
+      "learning_rate": 9.272684417489963e-05,
+      "loss": 3.0929,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7695663740356478,
+      "grad_norm": 1.0786086974162759,
+      "learning_rate": 9.272041068979473e-05,
+      "loss": 3.119,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7697791965948391,
+      "grad_norm": 1.1198024233748198,
+      "learning_rate": 9.271397458395732e-05,
+      "loss": 3.1034,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7699920191540304,
+      "grad_norm": 1.2105742763894443,
+      "learning_rate": 9.270753585778222e-05,
+      "loss": 3.1249,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7702048417132216,
+      "grad_norm": 1.129361068509693,
+      "learning_rate": 9.270109451166446e-05,
+      "loss": 3.1356,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7704176642724129,
+      "grad_norm": 1.1756485779321322,
+      "learning_rate": 9.269465054599915e-05,
+      "loss": 3.1586,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7706304868316042,
+      "grad_norm": 1.144023572486594,
+      "learning_rate": 9.268820396118163e-05,
+      "loss": 3.1145,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7708433093907954,
+      "grad_norm": 1.0826394835690223,
+      "learning_rate": 9.268175475760734e-05,
+      "loss": 3.0974,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7710561319499867,
+      "grad_norm": 1.2970582179427876,
+      "learning_rate": 9.267530293567197e-05,
+      "loss": 3.1198,
+      "step": 3623
+    },
+    {
+      "epoch": 0.771268954509178,
+      "grad_norm": 1.0091142902247376,
+      "learning_rate": 9.266884849577124e-05,
+      "loss": 3.1371,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7714817770683693,
+      "grad_norm": 1.2401560790358817,
+      "learning_rate": 9.266239143830118e-05,
+      "loss": 3.0979,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7716945996275605,
+      "grad_norm": 1.042317671540701,
+      "learning_rate": 9.265593176365786e-05,
+      "loss": 3.1098,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7719074221867518,
+      "grad_norm": 1.3098243417741158,
+      "learning_rate": 9.264946947223755e-05,
+      "loss": 3.12,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7721202447459431,
+      "grad_norm": 1.0325675141051684,
+      "learning_rate": 9.264300456443672e-05,
+      "loss": 3.0781,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7723330673051343,
+      "grad_norm": 1.1693842844077862,
+      "learning_rate": 9.263653704065194e-05,
+      "loss": 3.132,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7725458898643256,
+      "grad_norm": 1.17538661814043,
+      "learning_rate": 9.263006690127999e-05,
+      "loss": 3.1233,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7727587124235169,
+      "grad_norm": 1.3153481027371414,
+      "learning_rate": 9.262359414671777e-05,
+      "loss": 3.1455,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7729715349827082,
+      "grad_norm": 1.087574810507851,
+      "learning_rate": 9.261711877736236e-05,
+      "loss": 3.0955,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7731843575418994,
+      "grad_norm": 1.245868614636346,
+      "learning_rate": 9.261064079361101e-05,
+      "loss": 3.1233,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7733971801010907,
+      "grad_norm": 1.0008236296215005,
+      "learning_rate": 9.260416019586112e-05,
+      "loss": 3.0719,
+      "step": 3634
+    },
+    {
+      "epoch": 0.773610002660282,
+      "grad_norm": 1.3064135273808801,
+      "learning_rate": 9.259767698451025e-05,
+      "loss": 3.1824,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7738228252194732,
+      "grad_norm": 1.1965973707505135,
+      "learning_rate": 9.259119115995609e-05,
+      "loss": 3.0988,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7740356477786645,
+      "grad_norm": 1.196200977260708,
+      "learning_rate": 9.258470272259656e-05,
+      "loss": 3.143,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7742484703378558,
+      "grad_norm": 1.0516532689869886,
+      "learning_rate": 9.257821167282968e-05,
+      "loss": 3.1272,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7744612928970471,
+      "grad_norm": 1.0405712818397432,
+      "learning_rate": 9.257171801105364e-05,
+      "loss": 3.0928,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7746741154562383,
+      "grad_norm": 1.2020869375879597,
+      "learning_rate": 9.256522173766683e-05,
+      "loss": 3.1546,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7748869380154296,
+      "grad_norm": 1.1306482811037148,
+      "learning_rate": 9.255872285306777e-05,
+      "loss": 3.1293,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7750997605746209,
+      "grad_norm": 1.1481215551742006,
+      "learning_rate": 9.255222135765511e-05,
+      "loss": 3.0587,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7753125831338122,
+      "grad_norm": 0.9484333193072864,
+      "learning_rate": 9.254571725182773e-05,
+      "loss": 3.0919,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7755254056930034,
+      "grad_norm": 1.1960769735193562,
+      "learning_rate": 9.253921053598459e-05,
+      "loss": 3.1009,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7757382282521947,
+      "grad_norm": 1.0969672875187617,
+      "learning_rate": 9.253270121052488e-05,
+      "loss": 3.1257,
+      "step": 3645
+    },
+    {
+      "epoch": 0.775951050811386,
+      "grad_norm": 1.081321884478779,
+      "learning_rate": 9.252618927584791e-05,
+      "loss": 3.1382,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7761638733705772,
+      "grad_norm": 1.273718477581445,
+      "learning_rate": 9.251967473235319e-05,
+      "loss": 3.1078,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7763766959297685,
+      "grad_norm": 1.1175105720483298,
+      "learning_rate": 9.251315758044033e-05,
+      "loss": 3.1427,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7765895184889599,
+      "grad_norm": 1.053995019713114,
+      "learning_rate": 9.250663782050913e-05,
+      "loss": 3.1213,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7768023410481512,
+      "grad_norm": 1.2638972512312796,
+      "learning_rate": 9.250011545295958e-05,
+      "loss": 3.1098,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7770151636073424,
+      "grad_norm": 1.285879849298085,
+      "learning_rate": 9.249359047819178e-05,
+      "loss": 3.0963,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7772279861665337,
+      "grad_norm": 1.0905131531380845,
+      "learning_rate": 9.248706289660603e-05,
+      "loss": 3.1558,
+      "step": 3652
+    },
+    {
+      "epoch": 0.777440808725725,
+      "grad_norm": 1.1192126972162533,
+      "learning_rate": 9.248053270860276e-05,
+      "loss": 3.1068,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7776536312849162,
+      "grad_norm": 1.053427156898824,
+      "learning_rate": 9.247399991458255e-05,
+      "loss": 3.1079,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7778664538441075,
+      "grad_norm": 1.2800221621028154,
+      "learning_rate": 9.24674645149462e-05,
+      "loss": 3.118,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7780792764032988,
+      "grad_norm": 0.9060430234685722,
+      "learning_rate": 9.246092651009462e-05,
+      "loss": 3.105,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7782920989624901,
+      "grad_norm": 1.2876566266297294,
+      "learning_rate": 9.245438590042887e-05,
+      "loss": 3.1121,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7785049215216813,
+      "grad_norm": 0.9652206279143325,
+      "learning_rate": 9.244784268635023e-05,
+      "loss": 3.1107,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7787177440808726,
+      "grad_norm": 1.1603940630753893,
+      "learning_rate": 9.244129686826008e-05,
+      "loss": 3.1072,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7789305666400639,
+      "grad_norm": 1.0796791563125783,
+      "learning_rate": 9.243474844655997e-05,
+      "loss": 3.1084,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7791433891992551,
+      "grad_norm": 1.2479974877086046,
+      "learning_rate": 9.242819742165163e-05,
+      "loss": 3.0761,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7793562117584464,
+      "grad_norm": 0.9244356763249432,
+      "learning_rate": 9.242164379393694e-05,
+      "loss": 3.1368,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7795690343176377,
+      "grad_norm": 1.111843202804573,
+      "learning_rate": 9.241508756381795e-05,
+      "loss": 3.1175,
+      "step": 3663
+    },
+    {
+      "epoch": 0.779781856876829,
+      "grad_norm": 1.1974986007923003,
+      "learning_rate": 9.240852873169685e-05,
+      "loss": 3.1323,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7799946794360202,
+      "grad_norm": 0.9284773161899261,
+      "learning_rate": 9.240196729797598e-05,
+      "loss": 3.113,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7802075019952115,
+      "grad_norm": 1.2048203180628818,
+      "learning_rate": 9.239540326305791e-05,
+      "loss": 3.1135,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7804203245544028,
+      "grad_norm": 1.0702873048937098,
+      "learning_rate": 9.238883662734525e-05,
+      "loss": 3.1224,
+      "step": 3667
+    },
+    {
+      "epoch": 0.780633147113594,
+      "grad_norm": 1.1021087628618484,
+      "learning_rate": 9.23822673912409e-05,
+      "loss": 3.0877,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7808459696727853,
+      "grad_norm": 1.1309746649853474,
+      "learning_rate": 9.237569555514783e-05,
+      "loss": 3.1311,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7810587922319766,
+      "grad_norm": 1.0502826585729503,
+      "learning_rate": 9.236912111946919e-05,
+      "loss": 3.1417,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7812716147911679,
+      "grad_norm": 1.1541130956425194,
+      "learning_rate": 9.236254408460831e-05,
+      "loss": 3.1102,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7814844373503591,
+      "grad_norm": 1.1282167069450248,
+      "learning_rate": 9.235596445096865e-05,
+      "loss": 3.1266,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7816972599095504,
+      "grad_norm": 1.2148524375360925,
+      "learning_rate": 9.234938221895387e-05,
+      "loss": 3.105,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7819100824687417,
+      "grad_norm": 1.1003873541165863,
+      "learning_rate": 9.234279738896774e-05,
+      "loss": 3.1108,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7821229050279329,
+      "grad_norm": 1.290341099558844,
+      "learning_rate": 9.233620996141421e-05,
+      "loss": 3.0487,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7823357275871242,
+      "grad_norm": 0.9917133601457929,
+      "learning_rate": 9.232961993669743e-05,
+      "loss": 3.123,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7825485501463155,
+      "grad_norm": 1.2215767845213419,
+      "learning_rate": 9.232302731522163e-05,
+      "loss": 3.151,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7827613727055068,
+      "grad_norm": 1.0191758309054626,
+      "learning_rate": 9.231643209739128e-05,
+      "loss": 3.0994,
+      "step": 3678
+    },
+    {
+      "epoch": 0.782974195264698,
+      "grad_norm": 1.1419344843521078,
+      "learning_rate": 9.230983428361095e-05,
+      "loss": 3.116,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7831870178238893,
+      "grad_norm": 1.09351597903091,
+      "learning_rate": 9.230323387428539e-05,
+      "loss": 3.1022,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7833998403830806,
+      "grad_norm": 1.4226603921956795,
+      "learning_rate": 9.22966308698195e-05,
+      "loss": 3.1127,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7836126629422718,
+      "grad_norm": 0.9889650010708033,
+      "learning_rate": 9.229002527061837e-05,
+      "loss": 3.1245,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7838254855014631,
+      "grad_norm": 1.458101080464433,
+      "learning_rate": 9.228341707708723e-05,
+      "loss": 3.1242,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7840383080606544,
+      "grad_norm": 0.9069292279038167,
+      "learning_rate": 9.227680628963145e-05,
+      "loss": 3.0824,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7842511306198457,
+      "grad_norm": 1.284316412571442,
+      "learning_rate": 9.227019290865657e-05,
+      "loss": 3.0981,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7844639531790369,
+      "grad_norm": 1.0747036026554235,
+      "learning_rate": 9.226357693456832e-05,
+      "loss": 3.1289,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7846767757382282,
+      "grad_norm": 1.1086096432120103,
+      "learning_rate": 9.225695836777256e-05,
+      "loss": 3.1125,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7848895982974196,
+      "grad_norm": 1.2711613604835166,
+      "learning_rate": 9.225033720867531e-05,
+      "loss": 3.101,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7851024208566107,
+      "grad_norm": 1.0864638153968282,
+      "learning_rate": 9.224371345768273e-05,
+      "loss": 3.089,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7853152434158021,
+      "grad_norm": 1.0518073221618587,
+      "learning_rate": 9.22370871152012e-05,
+      "loss": 3.1017,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7855280659749934,
+      "grad_norm": 1.0186802216828987,
+      "learning_rate": 9.223045818163719e-05,
+      "loss": 3.0896,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7857408885341847,
+      "grad_norm": 1.1802340512513017,
+      "learning_rate": 9.222382665739738e-05,
+      "loss": 3.0454,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7859537110933759,
+      "grad_norm": 1.1099462261177344,
+      "learning_rate": 9.221719254288856e-05,
+      "loss": 3.0415,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7861665336525672,
+      "grad_norm": 1.172652472376083,
+      "learning_rate": 9.221055583851774e-05,
+      "loss": 3.0998,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7863793562117585,
+      "grad_norm": 1.0393413866130101,
+      "learning_rate": 9.220391654469204e-05,
+      "loss": 3.118,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7865921787709497,
+      "grad_norm": 1.3928515010933091,
+      "learning_rate": 9.219727466181877e-05,
+      "loss": 3.1195,
+      "step": 3696
+    },
+    {
+      "epoch": 0.786805001330141,
+      "grad_norm": 0.9701160697003134,
+      "learning_rate": 9.219063019030536e-05,
+      "loss": 3.0859,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7870178238893323,
+      "grad_norm": 1.209129408303041,
+      "learning_rate": 9.218398313055945e-05,
+      "loss": 3.0461,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7872306464485236,
+      "grad_norm": 1.0249950440186932,
+      "learning_rate": 9.217733348298878e-05,
+      "loss": 3.1189,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7874434690077148,
+      "grad_norm": 1.2267884756910912,
+      "learning_rate": 9.21706812480013e-05,
+      "loss": 3.1051,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7876562915669061,
+      "grad_norm": 1.0835329806567777,
+      "learning_rate": 9.21640264260051e-05,
+      "loss": 3.0787,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7878691141260974,
+      "grad_norm": 1.0043816896607345,
+      "learning_rate": 9.215736901740842e-05,
+      "loss": 3.1249,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7880819366852886,
+      "grad_norm": 1.2159145478943147,
+      "learning_rate": 9.215070902261966e-05,
+      "loss": 3.1446,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7882947592444799,
+      "grad_norm": 1.1043379091788754,
+      "learning_rate": 9.214404644204741e-05,
+      "loss": 3.1354,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7885075818036712,
+      "grad_norm": 1.252918661906897,
+      "learning_rate": 9.213738127610036e-05,
+      "loss": 3.0843,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7887204043628625,
+      "grad_norm": 0.9798403646149875,
+      "learning_rate": 9.213071352518744e-05,
+      "loss": 3.1461,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7889332269220537,
+      "grad_norm": 1.3091612201540619,
+      "learning_rate": 9.212404318971763e-05,
+      "loss": 3.0819,
+      "step": 3707
+    },
+    {
+      "epoch": 0.789146049481245,
+      "grad_norm": 0.9787767633371811,
+      "learning_rate": 9.211737027010017e-05,
+      "loss": 3.1072,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7893588720404363,
+      "grad_norm": 1.2655675431917852,
+      "learning_rate": 9.21106947667444e-05,
+      "loss": 3.12,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7895716945996276,
+      "grad_norm": 1.0569953658072124,
+      "learning_rate": 9.210401668005986e-05,
+      "loss": 3.0991,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7897845171588188,
+      "grad_norm": 1.2008951459163768,
+      "learning_rate": 9.209733601045622e-05,
+      "loss": 3.0843,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7899973397180101,
+      "grad_norm": 1.0372617766833538,
+      "learning_rate": 9.209065275834329e-05,
+      "loss": 3.0895,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7902101622772014,
+      "grad_norm": 1.2628517421998113,
+      "learning_rate": 9.208396692413106e-05,
+      "loss": 3.0788,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7904229848363926,
+      "grad_norm": 0.9582621400594628,
+      "learning_rate": 9.207727850822971e-05,
+      "loss": 3.1113,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7906358073955839,
+      "grad_norm": 1.362781269074649,
+      "learning_rate": 9.207058751104954e-05,
+      "loss": 3.1348,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7908486299547752,
+      "grad_norm": 1.0256080115446204,
+      "learning_rate": 9.2063893933001e-05,
+      "loss": 3.0982,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7910614525139665,
+      "grad_norm": 0.9958419784125835,
+      "learning_rate": 9.205719777449473e-05,
+      "loss": 3.1118,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7912742750731577,
+      "grad_norm": 1.0438683031066116,
+      "learning_rate": 9.20504990359415e-05,
+      "loss": 3.1175,
+      "step": 3718
+    },
+    {
+      "epoch": 0.791487097632349,
+      "grad_norm": 1.1766833051302454,
+      "learning_rate": 9.204379771775227e-05,
+      "loss": 3.1235,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7916999201915403,
+      "grad_norm": 1.139563382881418,
+      "learning_rate": 9.203709382033814e-05,
+      "loss": 3.0896,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7919127427507315,
+      "grad_norm": 1.0683088130594132,
+      "learning_rate": 9.203038734411033e-05,
+      "loss": 3.125,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7921255653099228,
+      "grad_norm": 1.2830366439707572,
+      "learning_rate": 9.202367828948031e-05,
+      "loss": 3.0955,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7923383878691141,
+      "grad_norm": 1.120558644743241,
+      "learning_rate": 9.201696665685963e-05,
+      "loss": 3.1239,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7925512104283055,
+      "grad_norm": 1.0716370628659884,
+      "learning_rate": 9.201025244666002e-05,
+      "loss": 3.1284,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7927640329874966,
+      "grad_norm": 1.1386238233811892,
+      "learning_rate": 9.200353565929336e-05,
+      "loss": 3.131,
+      "step": 3725
+    },
+    {
+      "epoch": 0.792976855546688,
+      "grad_norm": 1.1182930459948017,
+      "learning_rate": 9.199681629517173e-05,
+      "loss": 3.1487,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7931896781058793,
+      "grad_norm": 1.048959862859761,
+      "learning_rate": 9.199009435470731e-05,
+      "loss": 3.0708,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7934025006650705,
+      "grad_norm": 1.0827734872836265,
+      "learning_rate": 9.198336983831249e-05,
+      "loss": 3.0926,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7936153232242618,
+      "grad_norm": 1.0690440600499518,
+      "learning_rate": 9.197664274639977e-05,
+      "loss": 3.1249,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7938281457834531,
+      "grad_norm": 1.167305887704574,
+      "learning_rate": 9.196991307938182e-05,
+      "loss": 3.0761,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7940409683426444,
+      "grad_norm": 1.091738239814202,
+      "learning_rate": 9.196318083767152e-05,
+      "loss": 3.0808,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7942537909018356,
+      "grad_norm": 1.1763173539565295,
+      "learning_rate": 9.195644602168184e-05,
+      "loss": 3.1239,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7944666134610269,
+      "grad_norm": 1.0225261529711291,
+      "learning_rate": 9.194970863182594e-05,
+      "loss": 3.1457,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7946794360202182,
+      "grad_norm": 1.2230221958378524,
+      "learning_rate": 9.194296866851714e-05,
+      "loss": 3.1381,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7948922585794094,
+      "grad_norm": 1.0379404075358991,
+      "learning_rate": 9.193622613216891e-05,
+      "loss": 3.1547,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7951050811386007,
+      "grad_norm": 1.1977439786743171,
+      "learning_rate": 9.192948102319485e-05,
+      "loss": 3.1168,
+      "step": 3736
+    },
+    {
+      "epoch": 0.795317903697792,
+      "grad_norm": 0.9349059903369048,
+      "learning_rate": 9.192273334200879e-05,
+      "loss": 3.1055,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7955307262569833,
+      "grad_norm": 1.0982796440180247,
+      "learning_rate": 9.191598308902463e-05,
+      "loss": 3.1155,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7957435488161745,
+      "grad_norm": 1.162998240520515,
+      "learning_rate": 9.190923026465652e-05,
+      "loss": 3.1251,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7959563713753658,
+      "grad_norm": 1.0995033245494885,
+      "learning_rate": 9.190247486931869e-05,
+      "loss": 3.0902,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7961691939345571,
+      "grad_norm": 1.1398240048623385,
+      "learning_rate": 9.189571690342555e-05,
+      "loss": 3.0805,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7963820164937483,
+      "grad_norm": 1.0666566020633652,
+      "learning_rate": 9.18889563673917e-05,
+      "loss": 3.0391,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7965948390529396,
+      "grad_norm": 1.1457822027053322,
+      "learning_rate": 9.188219326163185e-05,
+      "loss": 3.1116,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7968076616121309,
+      "grad_norm": 0.9764426700412062,
+      "learning_rate": 9.187542758656091e-05,
+      "loss": 3.1184,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7970204841713222,
+      "grad_norm": 1.2136895653890727,
+      "learning_rate": 9.186865934259392e-05,
+      "loss": 3.1136,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7972333067305134,
+      "grad_norm": 0.9956458816188714,
+      "learning_rate": 9.186188853014609e-05,
+      "loss": 3.1177,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7974461292897047,
+      "grad_norm": 1.2094134165750712,
+      "learning_rate": 9.185511514963277e-05,
+      "loss": 3.1208,
+      "step": 3747
+    },
+    {
+      "epoch": 0.797658951848896,
+      "grad_norm": 0.9271837025539532,
+      "learning_rate": 9.184833920146949e-05,
+      "loss": 3.0919,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7978717744080872,
+      "grad_norm": 1.0318347540489794,
+      "learning_rate": 9.184156068607193e-05,
+      "loss": 3.0844,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7980845969672785,
+      "grad_norm": 1.326846256307054,
+      "learning_rate": 9.183477960385592e-05,
+      "loss": 3.1288,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7982974195264698,
+      "grad_norm": 0.9692502639552834,
+      "learning_rate": 9.182799595523746e-05,
+      "loss": 3.0944,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7985102420856611,
+      "grad_norm": 1.4368304335083155,
+      "learning_rate": 9.182120974063269e-05,
+      "loss": 3.1279,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7987230646448523,
+      "grad_norm": 0.8761427292996312,
+      "learning_rate": 9.181442096045794e-05,
+      "loss": 3.0846,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7989358872040436,
+      "grad_norm": 1.1115416908006799,
+      "learning_rate": 9.180762961512964e-05,
+      "loss": 3.127,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7991487097632349,
+      "grad_norm": 1.128720982519039,
+      "learning_rate": 9.180083570506444e-05,
+      "loss": 3.0776,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7993615323224261,
+      "grad_norm": 1.2147927589825784,
+      "learning_rate": 9.179403923067913e-05,
+      "loss": 3.0977,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7995743548816174,
+      "grad_norm": 1.106603678487679,
+      "learning_rate": 9.178724019239062e-05,
+      "loss": 3.0939,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7997871774408087,
+      "grad_norm": 1.1257444528863223,
+      "learning_rate": 9.178043859061602e-05,
+      "loss": 3.0622,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0402794837995513,
+      "learning_rate": 9.177363442577255e-05,
+      "loss": 3.1179,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8002128225591912,
+      "grad_norm": 1.262785992195679,
+      "learning_rate": 9.176682769827769e-05,
+      "loss": 3.1232,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8004256451183825,
+      "grad_norm": 1.1931269362449974,
+      "learning_rate": 9.176001840854893e-05,
+      "loss": 3.0563,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8006384676775739,
+      "grad_norm": 1.0385577080649913,
+      "learning_rate": 9.175320655700406e-05,
+      "loss": 3.1141,
+      "step": 3762
+    },
+    {
+      "epoch": 0.800851290236765,
+      "grad_norm": 1.2518400824340772,
+      "learning_rate": 9.17463921440609e-05,
+      "loss": 3.069,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8010641127959564,
+      "grad_norm": 0.9817110340378123,
+      "learning_rate": 9.173957517013752e-05,
+      "loss": 3.0772,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8012769353551477,
+      "grad_norm": 1.2416828116921148,
+      "learning_rate": 9.173275563565212e-05,
+      "loss": 3.0948,
+      "step": 3765
+    },
+    {
+      "epoch": 0.801489757914339,
+      "grad_norm": 1.0367612669483297,
+      "learning_rate": 9.172593354102304e-05,
+      "loss": 3.1549,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8017025804735302,
+      "grad_norm": 1.2634140927439692,
+      "learning_rate": 9.171910888666879e-05,
+      "loss": 3.1034,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8019154030327215,
+      "grad_norm": 0.9411706401412081,
+      "learning_rate": 9.171228167300804e-05,
+      "loss": 3.082,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8021282255919128,
+      "grad_norm": 1.1869793262944648,
+      "learning_rate": 9.170545190045963e-05,
+      "loss": 3.1353,
+      "step": 3769
+    },
+    {
+      "epoch": 0.802341048151104,
+      "grad_norm": 1.1584720043877907,
+      "learning_rate": 9.16986195694425e-05,
+      "loss": 3.1398,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8025538707102953,
+      "grad_norm": 1.1127883999978174,
+      "learning_rate": 9.169178468037582e-05,
+      "loss": 3.1004,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8027666932694866,
+      "grad_norm": 1.2180056448516559,
+      "learning_rate": 9.168494723367888e-05,
+      "loss": 3.1274,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8029795158286779,
+      "grad_norm": 1.169636923943185,
+      "learning_rate": 9.167810722977111e-05,
+      "loss": 3.0476,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8031923383878691,
+      "grad_norm": 0.93068936907036,
+      "learning_rate": 9.167126466907214e-05,
+      "loss": 3.1013,
+      "step": 3774
+    },
+    {
+      "epoch": 0.8034051609470604,
+      "grad_norm": 1.1541519044052746,
+      "learning_rate": 9.166441955200172e-05,
+      "loss": 3.1317,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8036179835062517,
+      "grad_norm": 1.2177901392462285,
+      "learning_rate": 9.165757187897979e-05,
+      "loss": 3.0953,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8038308060654429,
+      "grad_norm": 1.0937071024095115,
+      "learning_rate": 9.16507216504264e-05,
+      "loss": 3.1157,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8040436286246342,
+      "grad_norm": 1.0504592189898272,
+      "learning_rate": 9.164386886676181e-05,
+      "loss": 3.1291,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8042564511838255,
+      "grad_norm": 1.1174983624199364,
+      "learning_rate": 9.163701352840643e-05,
+      "loss": 3.1466,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8044692737430168,
+      "grad_norm": 1.1218872253643644,
+      "learning_rate": 9.163015563578074e-05,
+      "loss": 3.1401,
+      "step": 3780
+    },
+    {
+      "epoch": 0.804682096302208,
+      "grad_norm": 1.1394718168988918,
+      "learning_rate": 9.162329518930552e-05,
+      "loss": 3.1057,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8048949188613993,
+      "grad_norm": 1.1770856010307738,
+      "learning_rate": 9.161643218940159e-05,
+      "loss": 3.1048,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8051077414205906,
+      "grad_norm": 0.9062666425663767,
+      "learning_rate": 9.160956663648998e-05,
+      "loss": 3.0958,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8053205639797819,
+      "grad_norm": 1.1626754148185534,
+      "learning_rate": 9.160269853099186e-05,
+      "loss": 3.1326,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8055333865389731,
+      "grad_norm": 1.2344148031343263,
+      "learning_rate": 9.159582787332857e-05,
+      "loss": 3.0725,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8057462090981644,
+      "grad_norm": 1.0290964665946394,
+      "learning_rate": 9.158895466392158e-05,
+      "loss": 3.0931,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8059590316573557,
+      "grad_norm": 1.2432093242285356,
+      "learning_rate": 9.158207890319256e-05,
+      "loss": 3.1251,
+      "step": 3787
+    },
+    {
+      "epoch": 0.8061718542165469,
+      "grad_norm": 1.074063204119403,
+      "learning_rate": 9.15752005915633e-05,
+      "loss": 3.1112,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8063846767757382,
+      "grad_norm": 1.11420778549016,
+      "learning_rate": 9.156831972945576e-05,
+      "loss": 3.1097,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8065974993349295,
+      "grad_norm": 0.9907920351133488,
+      "learning_rate": 9.156143631729204e-05,
+      "loss": 3.0727,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8068103218941208,
+      "grad_norm": 1.1879273847699,
+      "learning_rate": 9.155455035549444e-05,
+      "loss": 3.1234,
+      "step": 3791
+    },
+    {
+      "epoch": 0.807023144453312,
+      "grad_norm": 1.1417844901503045,
+      "learning_rate": 9.154766184448535e-05,
+      "loss": 3.1024,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8072359670125033,
+      "grad_norm": 1.1131585344012456,
+      "learning_rate": 9.154077078468739e-05,
+      "loss": 3.0847,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8074487895716946,
+      "grad_norm": 1.2375508852696824,
+      "learning_rate": 9.153387717652329e-05,
+      "loss": 3.114,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8076616121308858,
+      "grad_norm": 0.9486440175509149,
+      "learning_rate": 9.152698102041592e-05,
+      "loss": 3.1126,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8078744346900771,
+      "grad_norm": 1.2185895449539832,
+      "learning_rate": 9.152008231678839e-05,
+      "loss": 3.0948,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8080872572492684,
+      "grad_norm": 1.0963421816922168,
+      "learning_rate": 9.151318106606384e-05,
+      "loss": 3.0947,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8083000798084597,
+      "grad_norm": 1.2113143152579118,
+      "learning_rate": 9.150627726866567e-05,
+      "loss": 3.0886,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8085129023676509,
+      "grad_norm": 1.1165939209097262,
+      "learning_rate": 9.149937092501738e-05,
+      "loss": 3.109,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8087257249268422,
+      "grad_norm": 1.1376802138906792,
+      "learning_rate": 9.14924620355427e-05,
+      "loss": 3.1109,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8089385474860336,
+      "grad_norm": 1.1133770666812417,
+      "learning_rate": 9.148555060066541e-05,
+      "loss": 3.1109,
+      "step": 3801
+    },
+    {
+      "epoch": 0.8091513700452247,
+      "grad_norm": 1.3886069002940362,
+      "learning_rate": 9.147863662080953e-05,
+      "loss": 3.1123,
+      "step": 3802
+    },
+    {
+      "epoch": 0.809364192604416,
+      "grad_norm": 0.931887510279454,
+      "learning_rate": 9.147172009639921e-05,
+      "loss": 3.1121,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8095770151636074,
+      "grad_norm": 1.4121467023747332,
+      "learning_rate": 9.14648010278587e-05,
+      "loss": 3.1247,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8097898377227987,
+      "grad_norm": 1.0330107100728638,
+      "learning_rate": 9.145787941561252e-05,
+      "loss": 3.1097,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8100026602819899,
+      "grad_norm": 1.2479208709488798,
+      "learning_rate": 9.145095526008527e-05,
+      "loss": 3.1155,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8102154828411812,
+      "grad_norm": 0.8680980700903604,
+      "learning_rate": 9.144402856170168e-05,
+      "loss": 3.0789,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8104283054003725,
+      "grad_norm": 1.0569212017579699,
+      "learning_rate": 9.143709932088673e-05,
+      "loss": 3.045,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8106411279595637,
+      "grad_norm": 1.1164063225587886,
+      "learning_rate": 9.143016753806548e-05,
+      "loss": 3.0942,
+      "step": 3809
+    },
+    {
+      "epoch": 0.810853950518755,
+      "grad_norm": 1.1000749230513203,
+      "learning_rate": 9.142323321366315e-05,
+      "loss": 3.0839,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8110667730779463,
+      "grad_norm": 1.2262396258115258,
+      "learning_rate": 9.141629634810516e-05,
+      "loss": 3.0926,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8112795956371376,
+      "grad_norm": 0.9846223675409627,
+      "learning_rate": 9.140935694181705e-05,
+      "loss": 3.0915,
+      "step": 3812
+    },
+    {
+      "epoch": 0.8114924181963288,
+      "grad_norm": 1.2442298730429733,
+      "learning_rate": 9.140241499522452e-05,
+      "loss": 3.0652,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8117052407555201,
+      "grad_norm": 1.0023165338139601,
+      "learning_rate": 9.139547050875345e-05,
+      "loss": 3.0737,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8119180633147114,
+      "grad_norm": 1.130513308964487,
+      "learning_rate": 9.138852348282982e-05,
+      "loss": 3.137,
+      "step": 3815
+    },
+    {
+      "epoch": 0.8121308858739026,
+      "grad_norm": 1.122588645718107,
+      "learning_rate": 9.138157391787985e-05,
+      "loss": 3.111,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8123437084330939,
+      "grad_norm": 1.144269456498953,
+      "learning_rate": 9.137462181432985e-05,
+      "loss": 3.1081,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8125565309922852,
+      "grad_norm": 1.1322070388796206,
+      "learning_rate": 9.136766717260631e-05,
+      "loss": 3.12,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8127693535514765,
+      "grad_norm": 1.1285518551186133,
+      "learning_rate": 9.136070999313585e-05,
+      "loss": 3.0835,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8129821761106677,
+      "grad_norm": 1.074604199306683,
+      "learning_rate": 9.135375027634529e-05,
+      "loss": 3.0994,
+      "step": 3820
+    },
+    {
+      "epoch": 0.813194998669859,
+      "grad_norm": 1.0647382914428454,
+      "learning_rate": 9.134678802266156e-05,
+      "loss": 3.1563,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8134078212290503,
+      "grad_norm": 1.1151679070667873,
+      "learning_rate": 9.133982323251178e-05,
+      "loss": 3.0963,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8136206437882415,
+      "grad_norm": 1.0335728088896725,
+      "learning_rate": 9.13328559063232e-05,
+      "loss": 3.0808,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8138334663474328,
+      "grad_norm": 1.281180559061652,
+      "learning_rate": 9.132588604452328e-05,
+      "loss": 3.0876,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8140462889066241,
+      "grad_norm": 0.965719592417635,
+      "learning_rate": 9.131891364753955e-05,
+      "loss": 3.0774,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8142591114658154,
+      "grad_norm": 1.2180026849761865,
+      "learning_rate": 9.131193871579975e-05,
+      "loss": 3.1037,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8144719340250066,
+      "grad_norm": 0.9693873991548946,
+      "learning_rate": 9.130496124973177e-05,
+      "loss": 3.0858,
+      "step": 3827
+    },
+    {
+      "epoch": 0.8146847565841979,
+      "grad_norm": 1.2131613681984452,
+      "learning_rate": 9.129798124976366e-05,
+      "loss": 3.1126,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8148975791433892,
+      "grad_norm": 0.9860623464033139,
+      "learning_rate": 9.129099871632358e-05,
+      "loss": 3.0956,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8151104017025804,
+      "grad_norm": 1.0467499360183903,
+      "learning_rate": 9.128401364983994e-05,
+      "loss": 3.1152,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8153232242617717,
+      "grad_norm": 1.1687996356279204,
+      "learning_rate": 9.12770260507412e-05,
+      "loss": 3.0832,
+      "step": 3831
+    },
+    {
+      "epoch": 0.815536046820963,
+      "grad_norm": 1.0747789815201627,
+      "learning_rate": 9.127003591945605e-05,
+      "loss": 3.1186,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8157488693801543,
+      "grad_norm": 1.1456123553183117,
+      "learning_rate": 9.126304325641326e-05,
+      "loss": 3.0838,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8159616919393455,
+      "grad_norm": 1.1256931029918646,
+      "learning_rate": 9.125604806204186e-05,
+      "loss": 3.1153,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8161745144985368,
+      "grad_norm": 1.2250255270012442,
+      "learning_rate": 9.124905033677096e-05,
+      "loss": 3.1164,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8163873370577281,
+      "grad_norm": 1.0696156463589694,
+      "learning_rate": 9.124205008102982e-05,
+      "loss": 3.0972,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8166001596169193,
+      "grad_norm": 1.2345312220626563,
+      "learning_rate": 9.123504729524792e-05,
+      "loss": 3.0827,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8168129821761106,
+      "grad_norm": 0.9825703364118951,
+      "learning_rate": 9.122804197985482e-05,
+      "loss": 3.0879,
+      "step": 3838
+    },
+    {
+      "epoch": 0.817025804735302,
+      "grad_norm": 1.1690210569714885,
+      "learning_rate": 9.122103413528028e-05,
+      "loss": 3.0518,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8172386272944933,
+      "grad_norm": 1.096888125425123,
+      "learning_rate": 9.121402376195422e-05,
+      "loss": 3.0917,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8174514498536845,
+      "grad_norm": 1.1647181266459132,
+      "learning_rate": 9.120701086030667e-05,
+      "loss": 3.086,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8176642724128758,
+      "grad_norm": 1.1082637736391843,
+      "learning_rate": 9.119999543076785e-05,
+      "loss": 3.1014,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8178770949720671,
+      "grad_norm": 1.0080855941868956,
+      "learning_rate": 9.119297747376816e-05,
+      "loss": 3.0972,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8180899175312583,
+      "grad_norm": 1.2391177696557003,
+      "learning_rate": 9.118595698973808e-05,
+      "loss": 3.1109,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8183027400904496,
+      "grad_norm": 1.0588306155951568,
+      "learning_rate": 9.117893397910833e-05,
+      "loss": 3.072,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8185155626496409,
+      "grad_norm": 1.0831009811525427,
+      "learning_rate": 9.117190844230971e-05,
+      "loss": 3.1114,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8187283852088322,
+      "grad_norm": 1.0725645363536316,
+      "learning_rate": 9.116488037977324e-05,
+      "loss": 3.0688,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8189412077680234,
+      "grad_norm": 1.0727925148212973,
+      "learning_rate": 9.115784979193006e-05,
+      "loss": 3.1022,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8191540303272147,
+      "grad_norm": 1.1150887239654899,
+      "learning_rate": 9.115081667921143e-05,
+      "loss": 3.055,
+      "step": 3849
+    },
+    {
+      "epoch": 0.819366852886406,
+      "grad_norm": 1.162893953044874,
+      "learning_rate": 9.114378104204887e-05,
+      "loss": 3.0959,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8195796754455973,
+      "grad_norm": 1.0801829553399749,
+      "learning_rate": 9.113674288087394e-05,
+      "loss": 3.074,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8197924980047885,
+      "grad_norm": 1.1765000868393483,
+      "learning_rate": 9.112970219611842e-05,
+      "loss": 3.0581,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8200053205639798,
+      "grad_norm": 1.1661020801457453,
+      "learning_rate": 9.112265898821422e-05,
+      "loss": 3.1119,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8202181431231711,
+      "grad_norm": 1.0618825063884176,
+      "learning_rate": 9.111561325759343e-05,
+      "loss": 3.0609,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8204309656823623,
+      "grad_norm": 1.0708626032428579,
+      "learning_rate": 9.110856500468827e-05,
+      "loss": 3.1317,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8206437882415536,
+      "grad_norm": 1.197208371639124,
+      "learning_rate": 9.11015142299311e-05,
+      "loss": 3.098,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8208566108007449,
+      "grad_norm": 1.0101072279867698,
+      "learning_rate": 9.109446093375451e-05,
+      "loss": 3.0842,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8210694333599362,
+      "grad_norm": 1.2435270936220408,
+      "learning_rate": 9.108740511659116e-05,
+      "loss": 3.1496,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8212822559191274,
+      "grad_norm": 1.045002884813218,
+      "learning_rate": 9.108034677887387e-05,
+      "loss": 3.0404,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8214950784783187,
+      "grad_norm": 1.1592821995726676,
+      "learning_rate": 9.10732859210357e-05,
+      "loss": 3.0545,
+      "step": 3860
+    },
+    {
+      "epoch": 0.82170790103751,
+      "grad_norm": 1.01119358692847,
+      "learning_rate": 9.106622254350976e-05,
+      "loss": 3.0781,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8219207235967012,
+      "grad_norm": 1.4563948219007852,
+      "learning_rate": 9.10591566467294e-05,
+      "loss": 3.1233,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8221335461558925,
+      "grad_norm": 0.9346641472162956,
+      "learning_rate": 9.105208823112805e-05,
+      "loss": 3.1092,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8223463687150838,
+      "grad_norm": 1.2014307689601402,
+      "learning_rate": 9.104501729713935e-05,
+      "loss": 3.064,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8225591912742751,
+      "grad_norm": 1.1597137285989318,
+      "learning_rate": 9.103794384519707e-05,
+      "loss": 3.0795,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8227720138334663,
+      "grad_norm": 1.0220851316702995,
+      "learning_rate": 9.103086787573513e-05,
+      "loss": 3.0971,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8229848363926576,
+      "grad_norm": 1.1991937762082698,
+      "learning_rate": 9.102378938918764e-05,
+      "loss": 3.087,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8231976589518489,
+      "grad_norm": 1.0937353423913385,
+      "learning_rate": 9.101670838598881e-05,
+      "loss": 3.075,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8234104815110401,
+      "grad_norm": 1.4183039755053064,
+      "learning_rate": 9.100962486657303e-05,
+      "loss": 3.1535,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8236233040702314,
+      "grad_norm": 0.8183590340991536,
+      "learning_rate": 9.100253883137488e-05,
+      "loss": 3.0576,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8238361266294227,
+      "grad_norm": 1.2645065786552605,
+      "learning_rate": 9.099545028082903e-05,
+      "loss": 3.0464,
+      "step": 3871
+    },
+    {
+      "epoch": 0.824048949188614,
+      "grad_norm": 0.9958813392393912,
+      "learning_rate": 9.098835921537034e-05,
+      "loss": 3.1075,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8242617717478052,
+      "grad_norm": 1.1938799835229472,
+      "learning_rate": 9.098126563543383e-05,
+      "loss": 3.0723,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8244745943069965,
+      "grad_norm": 0.915609798267072,
+      "learning_rate": 9.097416954145466e-05,
+      "loss": 3.104,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8246874168661879,
+      "grad_norm": 0.9936735579963167,
+      "learning_rate": 9.096707093386814e-05,
+      "loss": 3.1003,
+      "step": 3875
+    },
+    {
+      "epoch": 0.824900239425379,
+      "grad_norm": 1.3232183401658824,
+      "learning_rate": 9.095996981310975e-05,
+      "loss": 3.1072,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8251130619845704,
+      "grad_norm": 0.9719375058140847,
+      "learning_rate": 9.09528661796151e-05,
+      "loss": 3.0525,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8253258845437617,
+      "grad_norm": 1.1525408690595995,
+      "learning_rate": 9.094576003382001e-05,
+      "loss": 3.1226,
+      "step": 3878
+    },
+    {
+      "epoch": 0.825538707102953,
+      "grad_norm": 1.0984713022681505,
+      "learning_rate": 9.093865137616038e-05,
+      "loss": 3.0903,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8257515296621442,
+      "grad_norm": 1.1339997127323225,
+      "learning_rate": 9.093154020707231e-05,
+      "loss": 3.0852,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8259643522213355,
+      "grad_norm": 1.0166793071122477,
+      "learning_rate": 9.092442652699203e-05,
+      "loss": 3.1035,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8261771747805268,
+      "grad_norm": 1.2182627339807537,
+      "learning_rate": 9.091731033635595e-05,
+      "loss": 3.0628,
+      "step": 3882
+    },
+    {
+      "epoch": 0.826389997339718,
+      "grad_norm": 0.9839627944767986,
+      "learning_rate": 9.091019163560064e-05,
+      "loss": 3.0607,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8266028198989093,
+      "grad_norm": 1.2636476590055872,
+      "learning_rate": 9.090307042516277e-05,
+      "loss": 3.0932,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8268156424581006,
+      "grad_norm": 1.112256187426526,
+      "learning_rate": 9.08959467054792e-05,
+      "loss": 3.1036,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8270284650172919,
+      "grad_norm": 0.9913325580315672,
+      "learning_rate": 9.088882047698697e-05,
+      "loss": 3.104,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8272412875764831,
+      "grad_norm": 1.3322231972287681,
+      "learning_rate": 9.088169174012323e-05,
+      "loss": 3.0771,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8274541101356744,
+      "grad_norm": 1.1372665197359924,
+      "learning_rate": 9.08745604953253e-05,
+      "loss": 3.0889,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8276669326948657,
+      "grad_norm": 1.3009987256710926,
+      "learning_rate": 9.086742674303067e-05,
+      "loss": 3.0968,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8278797552540569,
+      "grad_norm": 1.0346103370934758,
+      "learning_rate": 9.086029048367694e-05,
+      "loss": 3.1118,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8280925778132482,
+      "grad_norm": 1.2503917410803274,
+      "learning_rate": 9.085315171770191e-05,
+      "loss": 3.1309,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8283054003724395,
+      "grad_norm": 0.964047723451423,
+      "learning_rate": 9.084601044554352e-05,
+      "loss": 3.0594,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8285182229316308,
+      "grad_norm": 1.10975959229598,
+      "learning_rate": 9.083886666763985e-05,
+      "loss": 3.0907,
+      "step": 3893
+    },
+    {
+      "epoch": 0.828731045490822,
+      "grad_norm": 1.1812503559773897,
+      "learning_rate": 9.083172038442914e-05,
+      "loss": 3.0932,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8289438680500133,
+      "grad_norm": 0.9442724388418393,
+      "learning_rate": 9.08245715963498e-05,
+      "loss": 3.0672,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8291566906092046,
+      "grad_norm": 1.1623012299938833,
+      "learning_rate": 9.081742030384038e-05,
+      "loss": 3.0835,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8293695131683958,
+      "grad_norm": 1.012802710187365,
+      "learning_rate": 9.081026650733955e-05,
+      "loss": 3.1187,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8295823357275871,
+      "grad_norm": 1.2114187059255255,
+      "learning_rate": 9.080311020728621e-05,
+      "loss": 3.0981,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8297951582867784,
+      "grad_norm": 1.1366263134488095,
+      "learning_rate": 9.079595140411936e-05,
+      "loss": 3.0893,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8300079808459697,
+      "grad_norm": 1.0230825938355668,
+      "learning_rate": 9.078879009827817e-05,
+      "loss": 3.1339,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8302208034051609,
+      "grad_norm": 1.0560169797525296,
+      "learning_rate": 9.078162629020193e-05,
+      "loss": 3.0474,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8304336259643522,
+      "grad_norm": 1.187404786486249,
+      "learning_rate": 9.077445998033015e-05,
+      "loss": 3.0815,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8306464485235435,
+      "grad_norm": 1.1285820023316617,
+      "learning_rate": 9.076729116910243e-05,
+      "loss": 3.0761,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8308592710827347,
+      "grad_norm": 1.029350330265869,
+      "learning_rate": 9.076011985695856e-05,
+      "loss": 3.0629,
+      "step": 3904
+    },
+    {
+      "epoch": 0.831072093641926,
+      "grad_norm": 1.377409784039754,
+      "learning_rate": 9.075294604433846e-05,
+      "loss": 3.076,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8312849162011173,
+      "grad_norm": 0.7568644259367258,
+      "learning_rate": 9.074576973168223e-05,
+      "loss": 3.026,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8314977387603086,
+      "grad_norm": 1.1001035339035852,
+      "learning_rate": 9.073859091943011e-05,
+      "loss": 3.0544,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8317105613194998,
+      "grad_norm": 1.1566260718016987,
+      "learning_rate": 9.073140960802248e-05,
+      "loss": 3.0443,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8319233838786911,
+      "grad_norm": 0.9616775089112792,
+      "learning_rate": 9.07242257978999e-05,
+      "loss": 3.0998,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8321362064378824,
+      "grad_norm": 1.3646059468874383,
+      "learning_rate": 9.071703948950305e-05,
+      "loss": 3.057,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8323490289970736,
+      "grad_norm": 0.841957044362567,
+      "learning_rate": 9.07098506832728e-05,
+      "loss": 3.0484,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8325618515562649,
+      "grad_norm": 1.0481382311166758,
+      "learning_rate": 9.070265937965015e-05,
+      "loss": 3.0939,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8327746741154562,
+      "grad_norm": 1.1741413453086402,
+      "learning_rate": 9.069546557907624e-05,
+      "loss": 3.0658,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8329874966746476,
+      "grad_norm": 1.2634524983988913,
+      "learning_rate": 9.068826928199243e-05,
+      "loss": 3.0814,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8332003192338387,
+      "grad_norm": 0.9570292416973911,
+      "learning_rate": 9.068107048884014e-05,
+      "loss": 3.1063,
+      "step": 3915
+    },
+    {
+      "epoch": 0.83341314179303,
+      "grad_norm": 1.2483565801337675,
+      "learning_rate": 9.0673869200061e-05,
+      "loss": 3.0975,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8336259643522214,
+      "grad_norm": 1.1061491678724813,
+      "learning_rate": 9.066666541609679e-05,
+      "loss": 3.0606,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8338387869114126,
+      "grad_norm": 1.0923791210352887,
+      "learning_rate": 9.065945913738942e-05,
+      "loss": 3.0845,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8340516094706039,
+      "grad_norm": 1.1940288919774424,
+      "learning_rate": 9.065225036438097e-05,
+      "loss": 3.0754,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8342644320297952,
+      "grad_norm": 1.029307021230822,
+      "learning_rate": 9.064503909751369e-05,
+      "loss": 3.0792,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8344772545889865,
+      "grad_norm": 1.1858142972674925,
+      "learning_rate": 9.063782533722995e-05,
+      "loss": 3.0746,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8346900771481777,
+      "grad_norm": 0.9949245920815127,
+      "learning_rate": 9.063060908397228e-05,
+      "loss": 3.0848,
+      "step": 3922
+    },
+    {
+      "epoch": 0.834902899707369,
+      "grad_norm": 1.190156434602505,
+      "learning_rate": 9.06233903381834e-05,
+      "loss": 3.1096,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8351157222665603,
+      "grad_norm": 1.1238769381387603,
+      "learning_rate": 9.061616910030609e-05,
+      "loss": 3.0981,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8353285448257516,
+      "grad_norm": 0.9879063389171239,
+      "learning_rate": 9.06089453707834e-05,
+      "loss": 3.0787,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8355413673849428,
+      "grad_norm": 1.111182117929075,
+      "learning_rate": 9.060171915005846e-05,
+      "loss": 3.1049,
+      "step": 3926
+    },
+    {
+      "epoch": 0.8357541899441341,
+      "grad_norm": 1.1867981137595185,
+      "learning_rate": 9.059449043857458e-05,
+      "loss": 3.0514,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8359670125033254,
+      "grad_norm": 1.1100560716297962,
+      "learning_rate": 9.05872592367752e-05,
+      "loss": 3.1102,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8361798350625166,
+      "grad_norm": 1.008860717435422,
+      "learning_rate": 9.058002554510392e-05,
+      "loss": 3.0627,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8363926576217079,
+      "grad_norm": 1.083480707769841,
+      "learning_rate": 9.057278936400453e-05,
+      "loss": 3.0922,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8366054801808992,
+      "grad_norm": 1.2359158561817922,
+      "learning_rate": 9.056555069392091e-05,
+      "loss": 3.0909,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8368183027400905,
+      "grad_norm": 1.0370397942014364,
+      "learning_rate": 9.055830953529714e-05,
+      "loss": 3.0557,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8370311252992817,
+      "grad_norm": 1.0866204123580205,
+      "learning_rate": 9.055106588857744e-05,
+      "loss": 3.0788,
+      "step": 3933
+    },
+    {
+      "epoch": 0.837243947858473,
+      "grad_norm": 1.0715121846738074,
+      "learning_rate": 9.054381975420616e-05,
+      "loss": 3.0515,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8374567704176643,
+      "grad_norm": 1.0958516094014534,
+      "learning_rate": 9.053657113262785e-05,
+      "loss": 3.1112,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8376695929768555,
+      "grad_norm": 1.1931848647572898,
+      "learning_rate": 9.052932002428715e-05,
+      "loss": 3.0978,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8378824155360468,
+      "grad_norm": 0.9423639624787957,
+      "learning_rate": 9.052206642962895e-05,
+      "loss": 3.0967,
+      "step": 3937
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 1.1889843341131838,
+      "learning_rate": 9.051481034909815e-05,
+      "loss": 3.0794,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8383080606544294,
+      "grad_norm": 1.1581961266133198,
+      "learning_rate": 9.050755178313994e-05,
+      "loss": 3.0818,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8385208832136206,
+      "grad_norm": 1.0680271752729356,
+      "learning_rate": 9.050029073219959e-05,
+      "loss": 3.0694,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8387337057728119,
+      "grad_norm": 1.1758734713458978,
+      "learning_rate": 9.049302719672252e-05,
+      "loss": 3.0901,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8389465283320032,
+      "grad_norm": 1.056903468158178,
+      "learning_rate": 9.048576117715435e-05,
+      "loss": 3.114,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8391593508911944,
+      "grad_norm": 1.119680592200488,
+      "learning_rate": 9.04784926739408e-05,
+      "loss": 3.0789,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8393721734503857,
+      "grad_norm": 1.2612031699520543,
+      "learning_rate": 9.047122168752775e-05,
+      "loss": 3.0464,
+      "step": 3944
+    },
+    {
+      "epoch": 0.839584996009577,
+      "grad_norm": 0.860675522909129,
+      "learning_rate": 9.04639482183613e-05,
+      "loss": 3.059,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8397978185687683,
+      "grad_norm": 1.0591023004155966,
+      "learning_rate": 9.04566722668876e-05,
+      "loss": 3.028,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8400106411279595,
+      "grad_norm": 1.126684451684832,
+      "learning_rate": 9.044939383355302e-05,
+      "loss": 3.0288,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8402234636871508,
+      "grad_norm": 1.050161431806724,
+      "learning_rate": 9.044211291880407e-05,
+      "loss": 3.0931,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8404362862463421,
+      "grad_norm": 1.2570176340081216,
+      "learning_rate": 9.04348295230874e-05,
+      "loss": 3.0866,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8406491088055333,
+      "grad_norm": 1.0728296175308871,
+      "learning_rate": 9.04275436468498e-05,
+      "loss": 3.0436,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8408619313647246,
+      "grad_norm": 1.1114498590539412,
+      "learning_rate": 9.042025529053825e-05,
+      "loss": 3.0684,
+      "step": 3951
+    },
+    {
+      "epoch": 0.841074753923916,
+      "grad_norm": 0.9822606238322775,
+      "learning_rate": 9.041296445459988e-05,
+      "loss": 3.1051,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8412875764831073,
+      "grad_norm": 1.1598902935227717,
+      "learning_rate": 9.040567113948192e-05,
+      "loss": 3.0733,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8415003990422985,
+      "grad_norm": 1.0254495448487382,
+      "learning_rate": 9.03983753456318e-05,
+      "loss": 3.1022,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8417132216014898,
+      "grad_norm": 1.1771872898472542,
+      "learning_rate": 9.03910770734971e-05,
+      "loss": 3.0814,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8419260441606811,
+      "grad_norm": 1.090479600986494,
+      "learning_rate": 9.038377632352553e-05,
+      "loss": 3.0745,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8421388667198723,
+      "grad_norm": 1.0621735263384855,
+      "learning_rate": 9.037647309616495e-05,
+      "loss": 3.0697,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8423516892790636,
+      "grad_norm": 1.2771411934298575,
+      "learning_rate": 9.03691673918634e-05,
+      "loss": 3.0543,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8425645118382549,
+      "grad_norm": 0.9037001060362183,
+      "learning_rate": 9.036185921106907e-05,
+      "loss": 3.0524,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8427773343974462,
+      "grad_norm": 1.190003528186318,
+      "learning_rate": 9.035454855423026e-05,
+      "loss": 3.0763,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8429901569566374,
+      "grad_norm": 1.0615319170368283,
+      "learning_rate": 9.034723542179548e-05,
+      "loss": 3.0856,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8432029795158287,
+      "grad_norm": 1.1610574041032224,
+      "learning_rate": 9.033991981421335e-05,
+      "loss": 3.0937,
+      "step": 3962
+    },
+    {
+      "epoch": 0.84341580207502,
+      "grad_norm": 0.8590822735672177,
+      "learning_rate": 9.033260173193263e-05,
+      "loss": 3.0643,
+      "step": 3963
+    },
+    {
+      "epoch": 0.8436286246342112,
+      "grad_norm": 1.0148336764148318,
+      "learning_rate": 9.032528117540228e-05,
+      "loss": 3.0793,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8438414471934025,
+      "grad_norm": 1.1716097569826363,
+      "learning_rate": 9.031795814507142e-05,
+      "loss": 3.0641,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8440542697525938,
+      "grad_norm": 0.8685348285236253,
+      "learning_rate": 9.031063264138922e-05,
+      "loss": 3.0575,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8442670923117851,
+      "grad_norm": 1.0289241598023862,
+      "learning_rate": 9.030330466480512e-05,
+      "loss": 3.0897,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8444799148709763,
+      "grad_norm": 1.3407805357563731,
+      "learning_rate": 9.029597421576865e-05,
+      "loss": 3.0483,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8446927374301676,
+      "grad_norm": 0.9783515328117219,
+      "learning_rate": 9.02886412947295e-05,
+      "loss": 3.0608,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8449055599893589,
+      "grad_norm": 1.1134714352080362,
+      "learning_rate": 9.028130590213752e-05,
+      "loss": 3.0841,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8451183825485501,
+      "grad_norm": 1.1333282632155837,
+      "learning_rate": 9.027396803844272e-05,
+      "loss": 3.0507,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8453312051077414,
+      "grad_norm": 0.9776672226323506,
+      "learning_rate": 9.026662770409522e-05,
+      "loss": 3.0656,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8455440276669327,
+      "grad_norm": 1.1573454373197019,
+      "learning_rate": 9.025928489954537e-05,
+      "loss": 3.0906,
+      "step": 3973
+    },
+    {
+      "epoch": 0.845756850226124,
+      "grad_norm": 1.1039655408162707,
+      "learning_rate": 9.025193962524357e-05,
+      "loss": 3.0841,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8459696727853152,
+      "grad_norm": 1.0137335497077615,
+      "learning_rate": 9.024459188164045e-05,
+      "loss": 3.0597,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8461824953445065,
+      "grad_norm": 1.224375513737479,
+      "learning_rate": 9.023724166918676e-05,
+      "loss": 3.0868,
+      "step": 3976
+    },
+    {
+      "epoch": 0.8463953179036978,
+      "grad_norm": 0.9433528074271975,
+      "learning_rate": 9.022988898833342e-05,
+      "loss": 3.0127,
+      "step": 3977
+    },
+    {
+      "epoch": 0.846608140462889,
+      "grad_norm": 1.1492464537569727,
+      "learning_rate": 9.022253383953148e-05,
+      "loss": 3.0436,
+      "step": 3978
+    },
+    {
+      "epoch": 0.8468209630220803,
+      "grad_norm": 1.1085221692182115,
+      "learning_rate": 9.021517622323215e-05,
+      "loss": 3.0592,
+      "step": 3979
+    },
+    {
+      "epoch": 0.8470337855812716,
+      "grad_norm": 0.9751264215038096,
+      "learning_rate": 9.020781613988676e-05,
+      "loss": 3.1057,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8472466081404629,
+      "grad_norm": 1.1836720141355697,
+      "learning_rate": 9.020045358994689e-05,
+      "loss": 3.0322,
+      "step": 3981
+    },
+    {
+      "epoch": 0.8474594306996541,
+      "grad_norm": 1.0773194053956068,
+      "learning_rate": 9.019308857386415e-05,
+      "loss": 3.0591,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8476722532588454,
+      "grad_norm": 1.1058726101232688,
+      "learning_rate": 9.018572109209038e-05,
+      "loss": 3.0609,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8478850758180367,
+      "grad_norm": 1.2472282947627495,
+      "learning_rate": 9.017835114507754e-05,
+      "loss": 3.0937,
+      "step": 3984
+    },
+    {
+      "epoch": 0.8480978983772279,
+      "grad_norm": 0.8987098670254702,
+      "learning_rate": 9.017097873327773e-05,
+      "loss": 3.0764,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8483107209364192,
+      "grad_norm": 1.0819574304756303,
+      "learning_rate": 9.016360385714324e-05,
+      "loss": 3.0578,
+      "step": 3986
+    },
+    {
+      "epoch": 0.8485235434956105,
+      "grad_norm": 1.0394241357232155,
+      "learning_rate": 9.01562265171265e-05,
+      "loss": 3.0645,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8487363660548018,
+      "grad_norm": 1.0964530164629929,
+      "learning_rate": 9.014884671368006e-05,
+      "loss": 3.0471,
+      "step": 3988
+    },
+    {
+      "epoch": 0.848949188613993,
+      "grad_norm": 1.0220402450824866,
+      "learning_rate": 9.014146444725664e-05,
+      "loss": 3.0807,
+      "step": 3989
+    },
+    {
+      "epoch": 0.8491620111731844,
+      "grad_norm": 1.1483051735184848,
+      "learning_rate": 9.013407971830914e-05,
+      "loss": 3.0921,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8493748337323757,
+      "grad_norm": 1.1133741042299066,
+      "learning_rate": 9.012669252729056e-05,
+      "loss": 3.0574,
+      "step": 3991
+    },
+    {
+      "epoch": 0.8495876562915669,
+      "grad_norm": 1.0945864532572933,
+      "learning_rate": 9.011930287465409e-05,
+      "loss": 3.069,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8498004788507582,
+      "grad_norm": 1.0882867044986742,
+      "learning_rate": 9.011191076085304e-05,
+      "loss": 3.0589,
+      "step": 3993
+    },
+    {
+      "epoch": 0.8500133014099495,
+      "grad_norm": 1.035078494063724,
+      "learning_rate": 9.010451618634091e-05,
+      "loss": 3.0347,
+      "step": 3994
+    },
+    {
+      "epoch": 0.8502261239691408,
+      "grad_norm": 1.0972264613176468,
+      "learning_rate": 9.00971191515713e-05,
+      "loss": 3.0367,
+      "step": 3995
+    },
+    {
+      "epoch": 0.850438946528332,
+      "grad_norm": 1.3254763239999945,
+      "learning_rate": 9.008971965699801e-05,
+      "loss": 3.0842,
+      "step": 3996
+    },
+    {
+      "epoch": 0.8506517690875233,
+      "grad_norm": 0.9182793401598445,
+      "learning_rate": 9.008231770307498e-05,
+      "loss": 3.0597,
+      "step": 3997
+    },
+    {
+      "epoch": 0.8508645916467146,
+      "grad_norm": 1.0703490833406035,
+      "learning_rate": 9.007491329025628e-05,
+      "loss": 3.0524,
+      "step": 3998
+    },
+    {
+      "epoch": 0.8510774142059059,
+      "grad_norm": 1.265183624423311,
+      "learning_rate": 9.006750641899612e-05,
+      "loss": 3.0855,
+      "step": 3999
+    },
+    {
+      "epoch": 0.8512902367650971,
+      "grad_norm": 1.0225587356483046,
+      "learning_rate": 9.006009708974892e-05,
+      "loss": 3.0643,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8515030593242884,
+      "grad_norm": 1.2388654414875064,
+      "learning_rate": 9.00526853029692e-05,
+      "loss": 3.0574,
+      "step": 4001
+    },
+    {
+      "epoch": 0.8517158818834797,
+      "grad_norm": 1.042639997077809,
+      "learning_rate": 9.004527105911162e-05,
+      "loss": 3.1493,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8519287044426709,
+      "grad_norm": 1.2190337823684847,
+      "learning_rate": 9.003785435863106e-05,
+      "loss": 3.0504,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8521415270018622,
+      "grad_norm": 0.9686250872188834,
+      "learning_rate": 9.003043520198246e-05,
+      "loss": 3.0852,
+      "step": 4004
+    },
+    {
+      "epoch": 0.8523543495610535,
+      "grad_norm": 1.2309411295810082,
+      "learning_rate": 9.002301358962101e-05,
+      "loss": 3.1059,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8525671721202448,
+      "grad_norm": 0.9764732844536461,
+      "learning_rate": 9.001558952200196e-05,
+      "loss": 3.0714,
+      "step": 4006
+    },
+    {
+      "epoch": 0.852779994679436,
+      "grad_norm": 1.3407199250044113,
+      "learning_rate": 9.000816299958072e-05,
+      "loss": 3.084,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8529928172386273,
+      "grad_norm": 0.8731733075366057,
+      "learning_rate": 9.000073402281296e-05,
+      "loss": 3.0454,
+      "step": 4008
+    },
+    {
+      "epoch": 0.8532056397978186,
+      "grad_norm": 1.1560197552582427,
+      "learning_rate": 8.999330259215435e-05,
+      "loss": 3.0621,
+      "step": 4009
+    },
+    {
+      "epoch": 0.8534184623570098,
+      "grad_norm": 1.045403821693114,
+      "learning_rate": 8.99858687080608e-05,
+      "loss": 3.0862,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8536312849162011,
+      "grad_norm": 1.2616093307951994,
+      "learning_rate": 8.997843237098835e-05,
+      "loss": 3.0509,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8538441074753924,
+      "grad_norm": 0.9569075782862253,
+      "learning_rate": 8.997099358139321e-05,
+      "loss": 3.0291,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8540569300345837,
+      "grad_norm": 1.2064213064225549,
+      "learning_rate": 8.996355233973168e-05,
+      "loss": 3.0826,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8542697525937749,
+      "grad_norm": 1.0581224684669415,
+      "learning_rate": 8.99561086464603e-05,
+      "loss": 3.0114,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8544825751529662,
+      "grad_norm": 1.0249584498255488,
+      "learning_rate": 8.994866250203567e-05,
+      "loss": 3.029,
+      "step": 4015
+    },
+    {
+      "epoch": 0.8546953977121575,
+      "grad_norm": 1.232204169962972,
+      "learning_rate": 8.99412139069146e-05,
+      "loss": 3.0587,
+      "step": 4016
+    },
+    {
+      "epoch": 0.8549082202713487,
+      "grad_norm": 0.9533424964986927,
+      "learning_rate": 8.993376286155403e-05,
+      "loss": 3.0599,
+      "step": 4017
+    },
+    {
+      "epoch": 0.85512104283054,
+      "grad_norm": 1.0442173159663273,
+      "learning_rate": 8.992630936641106e-05,
+      "loss": 3.0944,
+      "step": 4018
+    },
+    {
+      "epoch": 0.8553338653897313,
+      "grad_norm": 1.2637803179818758,
+      "learning_rate": 8.991885342194293e-05,
+      "loss": 3.043,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8555466879489226,
+      "grad_norm": 0.9169123153290956,
+      "learning_rate": 8.991139502860703e-05,
+      "loss": 3.0772,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8557595105081138,
+      "grad_norm": 1.0361742059003372,
+      "learning_rate": 8.990393418686092e-05,
+      "loss": 3.0241,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8559723330673051,
+      "grad_norm": 1.0350579965048237,
+      "learning_rate": 8.989647089716227e-05,
+      "loss": 3.0917,
+      "step": 4022
+    },
+    {
+      "epoch": 0.8561851556264964,
+      "grad_norm": 1.2468561456519993,
+      "learning_rate": 8.988900515996892e-05,
+      "loss": 3.0501,
+      "step": 4023
+    },
+    {
+      "epoch": 0.8563979781856876,
+      "grad_norm": 0.8891702439286762,
+      "learning_rate": 8.988153697573889e-05,
+      "loss": 3.1011,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8566108007448789,
+      "grad_norm": 1.2247845824393093,
+      "learning_rate": 8.98740663449303e-05,
+      "loss": 3.078,
+      "step": 4025
+    },
+    {
+      "epoch": 0.8568236233040702,
+      "grad_norm": 1.0447499453752165,
+      "learning_rate": 8.986659326800147e-05,
+      "loss": 3.0456,
+      "step": 4026
+    },
+    {
+      "epoch": 0.8570364458632616,
+      "grad_norm": 1.1597707682637188,
+      "learning_rate": 8.985911774541083e-05,
+      "loss": 3.0587,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8572492684224527,
+      "grad_norm": 1.0381445642858689,
+      "learning_rate": 8.985163977761696e-05,
+      "loss": 3.0572,
+      "step": 4028
+    },
+    {
+      "epoch": 0.857462090981644,
+      "grad_norm": 1.2254417373657973,
+      "learning_rate": 8.984415936507863e-05,
+      "loss": 3.0657,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8576749135408354,
+      "grad_norm": 1.0154199912606234,
+      "learning_rate": 8.983667650825473e-05,
+      "loss": 3.0598,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8578877361000266,
+      "grad_norm": 1.0854662950249996,
+      "learning_rate": 8.982919120760428e-05,
+      "loss": 3.044,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8581005586592179,
+      "grad_norm": 1.039691982450658,
+      "learning_rate": 8.982170346358651e-05,
+      "loss": 3.0494,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8583133812184092,
+      "grad_norm": 1.1098370887863174,
+      "learning_rate": 8.981421327666075e-05,
+      "loss": 3.087,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8585262037776005,
+      "grad_norm": 1.154944959616273,
+      "learning_rate": 8.980672064728647e-05,
+      "loss": 3.1131,
+      "step": 4034
+    },
+    {
+      "epoch": 0.8587390263367917,
+      "grad_norm": 1.0991542883562462,
+      "learning_rate": 8.979922557592335e-05,
+      "loss": 3.0696,
+      "step": 4035
+    },
+    {
+      "epoch": 0.858951848895983,
+      "grad_norm": 1.0330892450402802,
+      "learning_rate": 8.979172806303118e-05,
+      "loss": 3.0825,
+      "step": 4036
+    },
+    {
+      "epoch": 0.8591646714551743,
+      "grad_norm": 1.355569716897149,
+      "learning_rate": 8.978422810906987e-05,
+      "loss": 3.0492,
+      "step": 4037
+    },
+    {
+      "epoch": 0.8593774940143655,
+      "grad_norm": 0.9544985391591007,
+      "learning_rate": 8.977672571449955e-05,
+      "loss": 3.0662,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8595903165735568,
+      "grad_norm": 1.0985677244417555,
+      "learning_rate": 8.976922087978044e-05,
+      "loss": 3.079,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8598031391327481,
+      "grad_norm": 1.1715655028013314,
+      "learning_rate": 8.976171360537295e-05,
+      "loss": 3.0689,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8600159616919394,
+      "grad_norm": 1.1038747935693927,
+      "learning_rate": 8.975420389173761e-05,
+      "loss": 3.0415,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8602287842511306,
+      "grad_norm": 0.9836805706329546,
+      "learning_rate": 8.974669173933511e-05,
+      "loss": 3.0844,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8604416068103219,
+      "grad_norm": 1.1802551847311495,
+      "learning_rate": 8.97391771486263e-05,
+      "loss": 3.0694,
+      "step": 4043
+    },
+    {
+      "epoch": 0.8606544293695132,
+      "grad_norm": 0.9218842060214087,
+      "learning_rate": 8.973166012007217e-05,
+      "loss": 3.0571,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8608672519287044,
+      "grad_norm": 1.2225257046596836,
+      "learning_rate": 8.972414065413386e-05,
+      "loss": 3.1201,
+      "step": 4045
+    },
+    {
+      "epoch": 0.8610800744878957,
+      "grad_norm": 1.0691038096690118,
+      "learning_rate": 8.971661875127266e-05,
+      "loss": 3.0865,
+      "step": 4046
+    },
+    {
+      "epoch": 0.861292897047087,
+      "grad_norm": 1.0540769438932498,
+      "learning_rate": 8.970909441195001e-05,
+      "loss": 3.0215,
+      "step": 4047
+    },
+    {
+      "epoch": 0.8615057196062783,
+      "grad_norm": 1.0392935996511679,
+      "learning_rate": 8.97015676366275e-05,
+      "loss": 3.0587,
+      "step": 4048
+    },
+    {
+      "epoch": 0.8617185421654695,
+      "grad_norm": 1.2006603054407154,
+      "learning_rate": 8.969403842576685e-05,
+      "loss": 3.0546,
+      "step": 4049
+    },
+    {
+      "epoch": 0.8619313647246608,
+      "grad_norm": 1.0623725708757772,
+      "learning_rate": 8.968650677982998e-05,
+      "loss": 3.082,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8621441872838521,
+      "grad_norm": 1.0858548190281756,
+      "learning_rate": 8.967897269927892e-05,
+      "loss": 3.1004,
+      "step": 4051
+    },
+    {
+      "epoch": 0.8623570098430433,
+      "grad_norm": 1.27050164956428,
+      "learning_rate": 8.967143618457583e-05,
+      "loss": 3.0307,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8625698324022346,
+      "grad_norm": 1.007616604981923,
+      "learning_rate": 8.966389723618307e-05,
+      "loss": 3.0617,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8627826549614259,
+      "grad_norm": 1.3186605649523777,
+      "learning_rate": 8.965635585456314e-05,
+      "loss": 3.082,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8629954775206172,
+      "grad_norm": 0.89096009897726,
+      "learning_rate": 8.964881204017864e-05,
+      "loss": 3.0288,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8632083000798084,
+      "grad_norm": 1.184643778183068,
+      "learning_rate": 8.964126579349236e-05,
+      "loss": 3.0616,
+      "step": 4056
+    },
+    {
+      "epoch": 0.8634211226389997,
+      "grad_norm": 0.9730846576651366,
+      "learning_rate": 8.963371711496726e-05,
+      "loss": 3.0424,
+      "step": 4057
+    },
+    {
+      "epoch": 0.863633945198191,
+      "grad_norm": 1.33628358512047,
+      "learning_rate": 8.96261660050664e-05,
+      "loss": 3.091,
+      "step": 4058
+    },
+    {
+      "epoch": 0.8638467677573822,
+      "grad_norm": 0.8872596544175007,
+      "learning_rate": 8.9618612464253e-05,
+      "loss": 3.0838,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8640595903165735,
+      "grad_norm": 1.0591866572223607,
+      "learning_rate": 8.961105649299048e-05,
+      "loss": 3.0541,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8642724128757648,
+      "grad_norm": 1.0109155931717773,
+      "learning_rate": 8.960349809174233e-05,
+      "loss": 3.0646,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8644852354349561,
+      "grad_norm": 1.1694860686570723,
+      "learning_rate": 8.959593726097227e-05,
+      "loss": 3.081,
+      "step": 4062
+    },
+    {
+      "epoch": 0.8646980579941473,
+      "grad_norm": 1.1262961170337105,
+      "learning_rate": 8.958837400114408e-05,
+      "loss": 3.0903,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8649108805533386,
+      "grad_norm": 1.1906902378268276,
+      "learning_rate": 8.958080831272176e-05,
+      "loss": 3.0288,
+      "step": 4064
+    },
+    {
+      "epoch": 0.86512370311253,
+      "grad_norm": 0.9508205955828732,
+      "learning_rate": 8.957324019616945e-05,
+      "loss": 3.0924,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8653365256717213,
+      "grad_norm": 1.0390811594361753,
+      "learning_rate": 8.95656696519514e-05,
+      "loss": 3.0519,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8655493482309125,
+      "grad_norm": 1.1762388066674356,
+      "learning_rate": 8.955809668053205e-05,
+      "loss": 3.0607,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8657621707901038,
+      "grad_norm": 0.9153620586706975,
+      "learning_rate": 8.955052128237597e-05,
+      "loss": 3.059,
+      "step": 4068
+    },
+    {
+      "epoch": 0.8659749933492951,
+      "grad_norm": 0.9466038632458048,
+      "learning_rate": 8.954294345794787e-05,
+      "loss": 3.0216,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8661878159084863,
+      "grad_norm": 1.1239393888162832,
+      "learning_rate": 8.953536320771263e-05,
+      "loss": 3.0417,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8664006384676776,
+      "grad_norm": 1.0444704746328475,
+      "learning_rate": 8.952778053213526e-05,
+      "loss": 3.0908,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8666134610268689,
+      "grad_norm": 1.0617133127832499,
+      "learning_rate": 8.952019543168095e-05,
+      "loss": 3.0593,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8668262835860602,
+      "grad_norm": 1.0918106706442838,
+      "learning_rate": 8.9512607906815e-05,
+      "loss": 3.0899,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8670391061452514,
+      "grad_norm": 1.1210667632054205,
+      "learning_rate": 8.950501795800287e-05,
+      "loss": 3.077,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8672519287044427,
+      "grad_norm": 1.2452674420751746,
+      "learning_rate": 8.949742558571019e-05,
+      "loss": 3.063,
+      "step": 4075
+    },
+    {
+      "epoch": 0.867464751263634,
+      "grad_norm": 0.9792418614999087,
+      "learning_rate": 8.948983079040272e-05,
+      "loss": 3.0313,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8676775738228252,
+      "grad_norm": 1.1366079268157132,
+      "learning_rate": 8.948223357254636e-05,
+      "loss": 3.0633,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8678903963820165,
+      "grad_norm": 1.0686088742333737,
+      "learning_rate": 8.947463393260718e-05,
+      "loss": 3.0373,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8681032189412078,
+      "grad_norm": 1.0459582187960381,
+      "learning_rate": 8.946703187105137e-05,
+      "loss": 3.0389,
+      "step": 4079
+    },
+    {
+      "epoch": 0.8683160415003991,
+      "grad_norm": 1.1175145530245878,
+      "learning_rate": 8.945942738834532e-05,
+      "loss": 3.0911,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8685288640595903,
+      "grad_norm": 1.0612765712132617,
+      "learning_rate": 8.945182048495551e-05,
+      "loss": 3.0708,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8687416866187816,
+      "grad_norm": 1.1088139386412903,
+      "learning_rate": 8.94442111613486e-05,
+      "loss": 3.0655,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8689545091779729,
+      "grad_norm": 1.2155032186164152,
+      "learning_rate": 8.94365994179914e-05,
+      "loss": 3.0552,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8691673317371641,
+      "grad_norm": 0.8900146890154164,
+      "learning_rate": 8.942898525535085e-05,
+      "loss": 3.0714,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8693801542963554,
+      "grad_norm": 0.8196570145124901,
+      "learning_rate": 8.942136867389406e-05,
+      "loss": 3.0935,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8695929768555467,
+      "grad_norm": 0.9879278633256486,
+      "learning_rate": 8.941374967408826e-05,
+      "loss": 3.0618,
+      "step": 4086
+    },
+    {
+      "epoch": 0.869805799414738,
+      "grad_norm": 1.2269302930954453,
+      "learning_rate": 8.940612825640088e-05,
+      "loss": 3.0342,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8700186219739292,
+      "grad_norm": 0.9671279742592147,
+      "learning_rate": 8.939850442129942e-05,
+      "loss": 3.1099,
+      "step": 4088
+    },
+    {
+      "epoch": 0.8702314445331205,
+      "grad_norm": 1.3495203803648148,
+      "learning_rate": 8.939087816925163e-05,
+      "loss": 3.0806,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8704442670923118,
+      "grad_norm": 0.8802985743327607,
+      "learning_rate": 8.938324950072529e-05,
+      "loss": 3.0715,
+      "step": 4090
+    },
+    {
+      "epoch": 0.870657089651503,
+      "grad_norm": 1.3293296072753915,
+      "learning_rate": 8.937561841618842e-05,
+      "loss": 3.0448,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8708699122106943,
+      "grad_norm": 0.9092500168899337,
+      "learning_rate": 8.936798491610916e-05,
+      "loss": 3.08,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8710827347698856,
+      "grad_norm": 1.2472987237886173,
+      "learning_rate": 8.93603490009558e-05,
+      "loss": 3.0519,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8712955573290769,
+      "grad_norm": 1.0901275393266505,
+      "learning_rate": 8.935271067119674e-05,
+      "loss": 3.0448,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8715083798882681,
+      "grad_norm": 1.0628600215288944,
+      "learning_rate": 8.934506992730062e-05,
+      "loss": 3.0381,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8717212024474594,
+      "grad_norm": 1.0739178263591105,
+      "learning_rate": 8.933742676973612e-05,
+      "loss": 3.0452,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8719340250066507,
+      "grad_norm": 1.1199237974962921,
+      "learning_rate": 8.932978119897213e-05,
+      "loss": 3.0681,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8721468475658419,
+      "grad_norm": 0.992345561509571,
+      "learning_rate": 8.93221332154777e-05,
+      "loss": 3.1022,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8723596701250332,
+      "grad_norm": 1.1922030437166378,
+      "learning_rate": 8.931448281972197e-05,
+      "loss": 3.0706,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8725724926842245,
+      "grad_norm": 0.9375086464645827,
+      "learning_rate": 8.930683001217428e-05,
+      "loss": 3.0725,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8727853152434158,
+      "grad_norm": 1.1901224826203332,
+      "learning_rate": 8.92991747933041e-05,
+      "loss": 3.0022,
+      "step": 4101
+    },
+    {
+      "epoch": 0.872998137802607,
+      "grad_norm": 1.1068153994102377,
+      "learning_rate": 8.929151716358104e-05,
+      "loss": 3.067,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8732109603617983,
+      "grad_norm": 1.0788762102200067,
+      "learning_rate": 8.928385712347489e-05,
+      "loss": 3.0457,
+      "step": 4103
+    },
+    {
+      "epoch": 0.8734237829209897,
+      "grad_norm": 1.0060109085125482,
+      "learning_rate": 8.927619467345554e-05,
+      "loss": 3.0745,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8736366054801809,
+      "grad_norm": 1.3093254988723226,
+      "learning_rate": 8.926852981399305e-05,
+      "loss": 3.0716,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8738494280393722,
+      "grad_norm": 0.927934331191235,
+      "learning_rate": 8.926086254555764e-05,
+      "loss": 3.0768,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8740622505985635,
+      "grad_norm": 1.052259945278218,
+      "learning_rate": 8.925319286861966e-05,
+      "loss": 3.0856,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8742750731577548,
+      "grad_norm": 1.3117812721365762,
+      "learning_rate": 8.924552078364963e-05,
+      "loss": 3.067,
+      "step": 4108
+    },
+    {
+      "epoch": 0.874487895716946,
+      "grad_norm": 0.99735490127652,
+      "learning_rate": 8.92378462911182e-05,
+      "loss": 3.0571,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8747007182761373,
+      "grad_norm": 1.2227531517377053,
+      "learning_rate": 8.923016939149615e-05,
+      "loss": 3.0276,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8749135408353286,
+      "grad_norm": 0.9427894353516559,
+      "learning_rate": 8.922249008525445e-05,
+      "loss": 3.0669,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8751263633945198,
+      "grad_norm": 1.2911839214972411,
+      "learning_rate": 8.921480837286417e-05,
+      "loss": 3.0379,
+      "step": 4112
+    },
+    {
+      "epoch": 0.8753391859537111,
+      "grad_norm": 0.8988383438593294,
+      "learning_rate": 8.92071242547966e-05,
+      "loss": 3.0425,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8755520085129024,
+      "grad_norm": 1.218881097929897,
+      "learning_rate": 8.919943773152309e-05,
+      "loss": 3.0425,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8757648310720937,
+      "grad_norm": 0.974004943619232,
+      "learning_rate": 8.91917488035152e-05,
+      "loss": 3.0633,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8759776536312849,
+      "grad_norm": 1.0836829023829093,
+      "learning_rate": 8.918405747124459e-05,
+      "loss": 3.0963,
+      "step": 4116
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 1.2777468139306583,
+      "learning_rate": 8.917636373518311e-05,
+      "loss": 3.0489,
+      "step": 4117
+    },
+    {
+      "epoch": 0.8764032987496675,
+      "grad_norm": 1.0170521774680266,
+      "learning_rate": 8.916866759580276e-05,
+      "loss": 3.0669,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8766161213088587,
+      "grad_norm": 1.10365674489313,
+      "learning_rate": 8.916096905357563e-05,
+      "loss": 3.0661,
+      "step": 4119
+    },
+    {
+      "epoch": 0.87682894386805,
+      "grad_norm": 1.0195958117931665,
+      "learning_rate": 8.915326810897403e-05,
+      "loss": 3.098,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8770417664272413,
+      "grad_norm": 1.0864459305895622,
+      "learning_rate": 8.914556476247037e-05,
+      "loss": 3.0388,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8772545889864326,
+      "grad_norm": 1.0845920965483822,
+      "learning_rate": 8.913785901453722e-05,
+      "loss": 3.0345,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8774674115456238,
+      "grad_norm": 1.093147317168206,
+      "learning_rate": 8.91301508656473e-05,
+      "loss": 3.0745,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8776802341048151,
+      "grad_norm": 1.0965470436762905,
+      "learning_rate": 8.912244031627345e-05,
+      "loss": 3.0091,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8778930566640064,
+      "grad_norm": 1.1572189579610968,
+      "learning_rate": 8.911472736688871e-05,
+      "loss": 3.0638,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8781058792231976,
+      "grad_norm": 1.1982115686494974,
+      "learning_rate": 8.910701201796625e-05,
+      "loss": 3.0932,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8783187017823889,
+      "grad_norm": 0.9908041462315424,
+      "learning_rate": 8.909929426997936e-05,
+      "loss": 3.0701,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8785315243415802,
+      "grad_norm": 1.0960438111920467,
+      "learning_rate": 8.90915741234015e-05,
+      "loss": 3.0618,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8787443469007715,
+      "grad_norm": 1.0405879925512396,
+      "learning_rate": 8.908385157870626e-05,
+      "loss": 3.0858,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8789571694599627,
+      "grad_norm": 1.043620854235227,
+      "learning_rate": 8.90761266363674e-05,
+      "loss": 3.0622,
+      "step": 4130
+    },
+    {
+      "epoch": 0.879169992019154,
+      "grad_norm": 1.2271466080984617,
+      "learning_rate": 8.90683992968588e-05,
+      "loss": 3.1094,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8793828145783453,
+      "grad_norm": 1.2096515585122076,
+      "learning_rate": 8.906066956065452e-05,
+      "loss": 3.0304,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8795956371375365,
+      "grad_norm": 0.8655814959389738,
+      "learning_rate": 8.905293742822875e-05,
+      "loss": 3.0486,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8798084596967278,
+      "grad_norm": 1.0526386887298498,
+      "learning_rate": 8.904520290005582e-05,
+      "loss": 3.0424,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8800212822559191,
+      "grad_norm": 1.1546354716557705,
+      "learning_rate": 8.90374659766102e-05,
+      "loss": 3.0573,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8802341048151104,
+      "grad_norm": 0.9180495387434062,
+      "learning_rate": 8.902972665836655e-05,
+      "loss": 3.0627,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8804469273743016,
+      "grad_norm": 1.0738019775141028,
+      "learning_rate": 8.902198494579962e-05,
+      "loss": 3.0295,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8806597499334929,
+      "grad_norm": 1.0844073457183379,
+      "learning_rate": 8.901424083938435e-05,
+      "loss": 3.0583,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8808725724926842,
+      "grad_norm": 0.8774786342302965,
+      "learning_rate": 8.900649433959581e-05,
+      "loss": 3.0375,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8810853950518756,
+      "grad_norm": 1.100725863448196,
+      "learning_rate": 8.89987454469092e-05,
+      "loss": 3.0738,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8812982176110667,
+      "grad_norm": 1.1336591398710523,
+      "learning_rate": 8.899099416179991e-05,
+      "loss": 3.0535,
+      "step": 4141
+    },
+    {
+      "epoch": 0.881511040170258,
+      "grad_norm": 1.0265846525365128,
+      "learning_rate": 8.898324048474343e-05,
+      "loss": 3.0817,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8817238627294494,
+      "grad_norm": 1.213542085792371,
+      "learning_rate": 8.897548441621544e-05,
+      "loss": 3.0289,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8819366852886406,
+      "grad_norm": 0.9143176628773765,
+      "learning_rate": 8.896772595669174e-05,
+      "loss": 3.0289,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8821495078478319,
+      "grad_norm": 1.0984226128149968,
+      "learning_rate": 8.895996510664826e-05,
+      "loss": 3.0356,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8823623304070232,
+      "grad_norm": 0.9788181821248421,
+      "learning_rate": 8.895220186656112e-05,
+      "loss": 3.0282,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8825751529662145,
+      "grad_norm": 1.0592523841076353,
+      "learning_rate": 8.894443623690655e-05,
+      "loss": 3.0679,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8827879755254057,
+      "grad_norm": 1.05782494825852,
+      "learning_rate": 8.893666821816096e-05,
+      "loss": 3.0116,
+      "step": 4148
+    },
+    {
+      "epoch": 0.883000798084597,
+      "grad_norm": 1.213019035931505,
+      "learning_rate": 8.892889781080087e-05,
+      "loss": 3.0309,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8832136206437883,
+      "grad_norm": 0.9171674993695229,
+      "learning_rate": 8.892112501530296e-05,
+      "loss": 3.0877,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8834264432029795,
+      "grad_norm": 1.0902289718734541,
+      "learning_rate": 8.89133498321441e-05,
+      "loss": 3.0461,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8836392657621708,
+      "grad_norm": 1.154824153788556,
+      "learning_rate": 8.890557226180122e-05,
+      "loss": 3.0341,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8838520883213621,
+      "grad_norm": 1.0817570577435502,
+      "learning_rate": 8.889779230475148e-05,
+      "loss": 3.0677,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8840649108805534,
+      "grad_norm": 1.0241167288551336,
+      "learning_rate": 8.889000996147213e-05,
+      "loss": 3.0826,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8842777334397446,
+      "grad_norm": 1.1544423981025216,
+      "learning_rate": 8.888222523244059e-05,
+      "loss": 3.0616,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8844905559989359,
+      "grad_norm": 0.9119693668704052,
+      "learning_rate": 8.887443811813443e-05,
+      "loss": 3.0335,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8847033785581272,
+      "grad_norm": 1.070926768419436,
+      "learning_rate": 8.886664861903134e-05,
+      "loss": 3.0564,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8849162011173184,
+      "grad_norm": 1.1650496779429642,
+      "learning_rate": 8.885885673560921e-05,
+      "loss": 3.0779,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8851290236765097,
+      "grad_norm": 0.9705944196980096,
+      "learning_rate": 8.8851062468346e-05,
+      "loss": 3.0317,
+      "step": 4159
+    },
+    {
+      "epoch": 0.885341846235701,
+      "grad_norm": 1.300138127935629,
+      "learning_rate": 8.884326581771989e-05,
+      "loss": 3.0628,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8855546687948923,
+      "grad_norm": 0.9874102759664082,
+      "learning_rate": 8.883546678420917e-05,
+      "loss": 3.0538,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8857674913540835,
+      "grad_norm": 1.2529535861437489,
+      "learning_rate": 8.882766536829228e-05,
+      "loss": 3.0599,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8859803139132748,
+      "grad_norm": 0.7813824523377261,
+      "learning_rate": 8.881986157044779e-05,
+      "loss": 3.054,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8861931364724661,
+      "grad_norm": 0.9493030127737202,
+      "learning_rate": 8.881205539115445e-05,
+      "loss": 3.0563,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8864059590316573,
+      "grad_norm": 1.2214989479157008,
+      "learning_rate": 8.880424683089112e-05,
+      "loss": 3.055,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8866187815908486,
+      "grad_norm": 1.05745936774456,
+      "learning_rate": 8.879643589013686e-05,
+      "loss": 3.0564,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8868316041500399,
+      "grad_norm": 1.190306642628231,
+      "learning_rate": 8.878862256937082e-05,
+      "loss": 3.1097,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8870444267092312,
+      "grad_norm": 1.336677929312089,
+      "learning_rate": 8.87808068690723e-05,
+      "loss": 3.0274,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8872572492684224,
+      "grad_norm": 0.7707862987560022,
+      "learning_rate": 8.87729887897208e-05,
+      "loss": 3.0455,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8874700718276137,
+      "grad_norm": 1.0456454174580876,
+      "learning_rate": 8.876516833179588e-05,
+      "loss": 3.0638,
+      "step": 4170
+    },
+    {
+      "epoch": 0.887682894386805,
+      "grad_norm": 1.132932551902584,
+      "learning_rate": 8.875734549577735e-05,
+      "loss": 3.0367,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8878957169459962,
+      "grad_norm": 1.0101160519929393,
+      "learning_rate": 8.874952028214509e-05,
+      "loss": 3.0285,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8881085395051875,
+      "grad_norm": 1.3161306814282756,
+      "learning_rate": 8.874169269137913e-05,
+      "loss": 3.0186,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8883213620643788,
+      "grad_norm": 0.8886215726259842,
+      "learning_rate": 8.873386272395967e-05,
+      "loss": 3.036,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8885341846235701,
+      "grad_norm": 1.210658739213297,
+      "learning_rate": 8.872603038036706e-05,
+      "loss": 3.076,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8887470071827613,
+      "grad_norm": 1.0342152554405852,
+      "learning_rate": 8.871819566108176e-05,
+      "loss": 3.0304,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8889598297419526,
+      "grad_norm": 1.1614531376033752,
+      "learning_rate": 8.871035856658445e-05,
+      "loss": 3.0489,
+      "step": 4177
+    },
+    {
+      "epoch": 0.889172652301144,
+      "grad_norm": 1.017144224508425,
+      "learning_rate": 8.870251909735586e-05,
+      "loss": 3.0219,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8893854748603351,
+      "grad_norm": 1.2345303913654802,
+      "learning_rate": 8.86946772538769e-05,
+      "loss": 3.0311,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8895982974195265,
+      "grad_norm": 0.9892625028183486,
+      "learning_rate": 8.868683303662868e-05,
+      "loss": 3.0446,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8898111199787178,
+      "grad_norm": 1.2545509224500713,
+      "learning_rate": 8.86789864460924e-05,
+      "loss": 3.0561,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8900239425379091,
+      "grad_norm": 1.1271875006351864,
+      "learning_rate": 8.86711374827494e-05,
+      "loss": 3.0823,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8902367650971003,
+      "grad_norm": 1.0567994228455724,
+      "learning_rate": 8.86632861470812e-05,
+      "loss": 3.07,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8904495876562916,
+      "grad_norm": 1.0315375064533683,
+      "learning_rate": 8.865543243956945e-05,
+      "loss": 3.0258,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8906624102154829,
+      "grad_norm": 0.9779837772090053,
+      "learning_rate": 8.864757636069591e-05,
+      "loss": 3.0677,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8908752327746741,
+      "grad_norm": 1.2139726081900757,
+      "learning_rate": 8.863971791094257e-05,
+      "loss": 3.0344,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8910880553338654,
+      "grad_norm": 0.9601423959072904,
+      "learning_rate": 8.863185709079149e-05,
+      "loss": 3.0467,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8913008778930567,
+      "grad_norm": 0.982812436642896,
+      "learning_rate": 8.862399390072491e-05,
+      "loss": 3.0544,
+      "step": 4188
+    },
+    {
+      "epoch": 0.891513700452248,
+      "grad_norm": 1.1193474454046741,
+      "learning_rate": 8.861612834122519e-05,
+      "loss": 3.0344,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8917265230114392,
+      "grad_norm": 0.9911248212330985,
+      "learning_rate": 8.860826041277488e-05,
+      "loss": 3.034,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8919393455706305,
+      "grad_norm": 1.018017433992764,
+      "learning_rate": 8.860039011585661e-05,
+      "loss": 2.9873,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8921521681298218,
+      "grad_norm": 1.037141681612242,
+      "learning_rate": 8.859251745095323e-05,
+      "loss": 3.0343,
+      "step": 4192
+    },
+    {
+      "epoch": 0.892364990689013,
+      "grad_norm": 0.9590596921192003,
+      "learning_rate": 8.858464241854767e-05,
+      "loss": 3.0463,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8925778132482043,
+      "grad_norm": 1.16359914944076,
+      "learning_rate": 8.857676501912306e-05,
+      "loss": 3.051,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8927906358073956,
+      "grad_norm": 1.041741356770768,
+      "learning_rate": 8.856888525316261e-05,
+      "loss": 3.0523,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8930034583665869,
+      "grad_norm": 1.1556423500200135,
+      "learning_rate": 8.856100312114975e-05,
+      "loss": 3.0699,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8932162809257781,
+      "grad_norm": 1.0659386753052433,
+      "learning_rate": 8.855311862356799e-05,
+      "loss": 3.0589,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8934291034849694,
+      "grad_norm": 1.0853488028054714,
+      "learning_rate": 8.854523176090104e-05,
+      "loss": 3.0504,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8936419260441607,
+      "grad_norm": 1.1477619826697751,
+      "learning_rate": 8.85373425336327e-05,
+      "loss": 3.0543,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8938547486033519,
+      "grad_norm": 1.0345681351759295,
+      "learning_rate": 8.852945094224697e-05,
+      "loss": 3.0577,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8940675711625432,
+      "grad_norm": 1.1484574676564143,
+      "learning_rate": 8.852155698722795e-05,
+      "loss": 3.0676,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8942803937217345,
+      "grad_norm": 1.0929048561883135,
+      "learning_rate": 8.851366066905993e-05,
+      "loss": 3.0362,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8944932162809258,
+      "grad_norm": 1.0957655993664464,
+      "learning_rate": 8.850576198822728e-05,
+      "loss": 3.0419,
+      "step": 4203
+    },
+    {
+      "epoch": 0.894706038840117,
+      "grad_norm": 1.0678378819022825,
+      "learning_rate": 8.849786094521458e-05,
+      "loss": 3.0494,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8949188613993083,
+      "grad_norm": 1.2755780924489186,
+      "learning_rate": 8.848995754050652e-05,
+      "loss": 3.0632,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8951316839584996,
+      "grad_norm": 0.9348039324810548,
+      "learning_rate": 8.848205177458795e-05,
+      "loss": 3.04,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8953445065176909,
+      "grad_norm": 1.0551911561730627,
+      "learning_rate": 8.847414364794385e-05,
+      "loss": 3.0822,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8955573290768821,
+      "grad_norm": 0.9681249263579307,
+      "learning_rate": 8.846623316105935e-05,
+      "loss": 3.0659,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8957701516360734,
+      "grad_norm": 1.156712791730634,
+      "learning_rate": 8.845832031441975e-05,
+      "loss": 3.0589,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8959829741952647,
+      "grad_norm": 1.0210312758383775,
+      "learning_rate": 8.845040510851045e-05,
+      "loss": 3.0671,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8961957967544559,
+      "grad_norm": 1.0677991292791962,
+      "learning_rate": 8.844248754381702e-05,
+      "loss": 3.0097,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8964086193136472,
+      "grad_norm": 0.9824584595681319,
+      "learning_rate": 8.843456762082518e-05,
+      "loss": 3.0256,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8966214418728385,
+      "grad_norm": 1.1777951716751613,
+      "learning_rate": 8.842664534002078e-05,
+      "loss": 3.0435,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8968342644320298,
+      "grad_norm": 0.9459709657882956,
+      "learning_rate": 8.841872070188983e-05,
+      "loss": 2.9872,
+      "step": 4214
+    },
+    {
+      "epoch": 0.897047086991221,
+      "grad_norm": 1.0669635796532166,
+      "learning_rate": 8.841079370691848e-05,
+      "loss": 3.0386,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8972599095504123,
+      "grad_norm": 1.0814329982165756,
+      "learning_rate": 8.8402864355593e-05,
+      "loss": 3.0628,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8974727321096037,
+      "grad_norm": 1.076902346363471,
+      "learning_rate": 8.839493264839984e-05,
+      "loss": 3.0567,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8976855546687948,
+      "grad_norm": 1.169831249525988,
+      "learning_rate": 8.838699858582558e-05,
+      "loss": 3.0419,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8978983772279862,
+      "grad_norm": 1.0605044421770053,
+      "learning_rate": 8.837906216835694e-05,
+      "loss": 3.0921,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8981111997871775,
+      "grad_norm": 1.103980832930935,
+      "learning_rate": 8.837112339648078e-05,
+      "loss": 2.9926,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8983240223463688,
+      "grad_norm": 1.005364095761531,
+      "learning_rate": 8.836318227068414e-05,
+      "loss": 3.0482,
+      "step": 4221
+    },
+    {
+      "epoch": 0.89853684490556,
+      "grad_norm": 1.1347381865828017,
+      "learning_rate": 8.835523879145413e-05,
+      "loss": 2.9918,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8987496674647513,
+      "grad_norm": 1.0205922734053738,
+      "learning_rate": 8.83472929592781e-05,
+      "loss": 3.0387,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8989624900239426,
+      "grad_norm": 1.0735696404984698,
+      "learning_rate": 8.833934477464347e-05,
+      "loss": 3.0401,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8991753125831338,
+      "grad_norm": 1.1340972116855865,
+      "learning_rate": 8.833139423803785e-05,
+      "loss": 3.0722,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8993881351423251,
+      "grad_norm": 1.2094696144236452,
+      "learning_rate": 8.832344134994897e-05,
+      "loss": 3.0352,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8996009577015164,
+      "grad_norm": 0.9244430102571194,
+      "learning_rate": 8.831548611086468e-05,
+      "loss": 2.9987,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8998137802607077,
+      "grad_norm": 1.0186171028737885,
+      "learning_rate": 8.830752852127303e-05,
+      "loss": 3.0364,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9000266028198989,
+      "grad_norm": 1.2104309560336464,
+      "learning_rate": 8.82995685816622e-05,
+      "loss": 3.0568,
+      "step": 4229
+    },
+    {
+      "epoch": 0.9002394253790902,
+      "grad_norm": 1.075122486444119,
+      "learning_rate": 8.829160629252046e-05,
+      "loss": 3.0804,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9004522479382815,
+      "grad_norm": 1.0659992882726157,
+      "learning_rate": 8.828364165433631e-05,
+      "loss": 3.0566,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9006650704974727,
+      "grad_norm": 1.095584642959942,
+      "learning_rate": 8.827567466759832e-05,
+      "loss": 3.0357,
+      "step": 4232
+    },
+    {
+      "epoch": 0.900877893056664,
+      "grad_norm": 1.1321312050243935,
+      "learning_rate": 8.826770533279522e-05,
+      "loss": 3.0425,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9010907156158553,
+      "grad_norm": 1.1343502660904927,
+      "learning_rate": 8.825973365041595e-05,
+      "loss": 3.0403,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9013035381750466,
+      "grad_norm": 1.0777629951965444,
+      "learning_rate": 8.825175962094951e-05,
+      "loss": 3.0185,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9015163607342378,
+      "grad_norm": 1.2102650212456283,
+      "learning_rate": 8.824378324488509e-05,
+      "loss": 3.0252,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9017291832934291,
+      "grad_norm": 0.8263356807071703,
+      "learning_rate": 8.823580452271198e-05,
+      "loss": 3.0304,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9019420058526204,
+      "grad_norm": 1.0414238534248526,
+      "learning_rate": 8.822782345491968e-05,
+      "loss": 3.0245,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9021548284118116,
+      "grad_norm": 1.25845501270413,
+      "learning_rate": 8.821984004199777e-05,
+      "loss": 3.0929,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9023676509710029,
+      "grad_norm": 0.8420338647721916,
+      "learning_rate": 8.821185428443601e-05,
+      "loss": 3.0303,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9025804735301942,
+      "grad_norm": 1.0172760683496234,
+      "learning_rate": 8.820386618272431e-05,
+      "loss": 3.033,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9027932960893855,
+      "grad_norm": 1.2146767940218093,
+      "learning_rate": 8.819587573735269e-05,
+      "loss": 3.0728,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9030061186485767,
+      "grad_norm": 0.7881011360246192,
+      "learning_rate": 8.818788294881136e-05,
+      "loss": 3.0183,
+      "step": 4243
+    },
+    {
+      "epoch": 0.903218941207768,
+      "grad_norm": 1.1006688271119538,
+      "learning_rate": 8.817988781759059e-05,
+      "loss": 3.0388,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9034317637669593,
+      "grad_norm": 1.1605620482978345,
+      "learning_rate": 8.81718903441809e-05,
+      "loss": 3.0367,
+      "step": 4245
+    },
+    {
+      "epoch": 0.9036445863261505,
+      "grad_norm": 1.0118607403652773,
+      "learning_rate": 8.81638905290729e-05,
+      "loss": 3.0711,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9038574088853418,
+      "grad_norm": 1.2061524513445643,
+      "learning_rate": 8.815588837275734e-05,
+      "loss": 3.0327,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9040702314445331,
+      "grad_norm": 0.9395881657515133,
+      "learning_rate": 8.814788387572513e-05,
+      "loss": 3.0224,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9042830540037244,
+      "grad_norm": 1.0195033408036156,
+      "learning_rate": 8.813987703846729e-05,
+      "loss": 3.0424,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9044958765629156,
+      "grad_norm": 1.195115057501346,
+      "learning_rate": 8.813186786147503e-05,
+      "loss": 3.0196,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9047086991221069,
+      "grad_norm": 1.01146423267464,
+      "learning_rate": 8.812385634523967e-05,
+      "loss": 3.0449,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9049215216812982,
+      "grad_norm": 0.9952927381285235,
+      "learning_rate": 8.811584249025271e-05,
+      "loss": 3.0478,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9051343442404894,
+      "grad_norm": 1.061011811925146,
+      "learning_rate": 8.810782629700574e-05,
+      "loss": 3.0454,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9053471667996807,
+      "grad_norm": 1.044696648524114,
+      "learning_rate": 8.809980776599054e-05,
+      "loss": 3.0084,
+      "step": 4254
+    },
+    {
+      "epoch": 0.905559989358872,
+      "grad_norm": 1.1581647197794818,
+      "learning_rate": 8.809178689769899e-05,
+      "loss": 3.0142,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9057728119180634,
+      "grad_norm": 0.9442250390050381,
+      "learning_rate": 8.80837636926232e-05,
+      "loss": 3.022,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9059856344772546,
+      "grad_norm": 1.1699937035195433,
+      "learning_rate": 8.807573815125527e-05,
+      "loss": 3.0307,
+      "step": 4257
+    },
+    {
+      "epoch": 0.9061984570364459,
+      "grad_norm": 1.0032179203073281,
+      "learning_rate": 8.806771027408763e-05,
+      "loss": 3.087,
+      "step": 4258
+    },
+    {
+      "epoch": 0.9064112795956372,
+      "grad_norm": 1.059419896568871,
+      "learning_rate": 8.805968006161269e-05,
+      "loss": 3.0252,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9066241021548284,
+      "grad_norm": 1.0958785678385523,
+      "learning_rate": 8.805164751432313e-05,
+      "loss": 3.0358,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9068369247140197,
+      "grad_norm": 1.0046661774802885,
+      "learning_rate": 8.804361263271166e-05,
+      "loss": 3.0116,
+      "step": 4261
+    },
+    {
+      "epoch": 0.907049747273211,
+      "grad_norm": 1.1865995300871985,
+      "learning_rate": 8.803557541727124e-05,
+      "loss": 3.0252,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9072625698324023,
+      "grad_norm": 1.01718250718681,
+      "learning_rate": 8.802753586849486e-05,
+      "loss": 2.9699,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9074753923915935,
+      "grad_norm": 1.0747451185987853,
+      "learning_rate": 8.801949398687578e-05,
+      "loss": 3.0509,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9076882149507848,
+      "grad_norm": 1.1078780564840365,
+      "learning_rate": 8.80114497729073e-05,
+      "loss": 3.0716,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9079010375099761,
+      "grad_norm": 1.065796083989821,
+      "learning_rate": 8.800340322708291e-05,
+      "loss": 3.0584,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9081138600691673,
+      "grad_norm": 1.0979975978630134,
+      "learning_rate": 8.799535434989625e-05,
+      "loss": 3.0227,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9083266826283586,
+      "grad_norm": 1.158453303477048,
+      "learning_rate": 8.798730314184107e-05,
+      "loss": 3.0443,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9085395051875499,
+      "grad_norm": 0.9795948192504764,
+      "learning_rate": 8.797924960341128e-05,
+      "loss": 3.0137,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9087523277467412,
+      "grad_norm": 1.1147472151605362,
+      "learning_rate": 8.797119373510095e-05,
+      "loss": 3.0156,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9089651503059324,
+      "grad_norm": 0.9959868518206259,
+      "learning_rate": 8.796313553740426e-05,
+      "loss": 3.0422,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9091779728651237,
+      "grad_norm": 1.214495140996774,
+      "learning_rate": 8.795507501081555e-05,
+      "loss": 3.0504,
+      "step": 4272
+    },
+    {
+      "epoch": 0.909390795424315,
+      "grad_norm": 0.9846028404967911,
+      "learning_rate": 8.79470121558293e-05,
+      "loss": 3.0445,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9096036179835062,
+      "grad_norm": 1.1304267334766915,
+      "learning_rate": 8.793894697294016e-05,
+      "loss": 3.0373,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9098164405426975,
+      "grad_norm": 1.089013725254442,
+      "learning_rate": 8.793087946264287e-05,
+      "loss": 3.0192,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9100292631018888,
+      "grad_norm": 1.0011454091204073,
+      "learning_rate": 8.792280962543235e-05,
+      "loss": 3.0273,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9102420856610801,
+      "grad_norm": 1.1126391338305288,
+      "learning_rate": 8.791473746180366e-05,
+      "loss": 3.0116,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9104549082202713,
+      "grad_norm": 1.1139431099406192,
+      "learning_rate": 8.790666297225196e-05,
+      "loss": 3.0333,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9106677307794626,
+      "grad_norm": 0.9723451268907818,
+      "learning_rate": 8.789858615727265e-05,
+      "loss": 3.0488,
+      "step": 4279
+    },
+    {
+      "epoch": 0.9108805533386539,
+      "grad_norm": 0.9940534041176223,
+      "learning_rate": 8.789050701736117e-05,
+      "loss": 3.0193,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9110933758978452,
+      "grad_norm": 1.2052960334081362,
+      "learning_rate": 8.788242555301315e-05,
+      "loss": 3.0484,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9113061984570364,
+      "grad_norm": 0.9373082550916166,
+      "learning_rate": 8.787434176472438e-05,
+      "loss": 3.0105,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9115190210162277,
+      "grad_norm": 0.9295378406940351,
+      "learning_rate": 8.786625565299074e-05,
+      "loss": 3.0662,
+      "step": 4283
+    },
+    {
+      "epoch": 0.911731843575419,
+      "grad_norm": 1.0767611216400201,
+      "learning_rate": 8.785816721830829e-05,
+      "loss": 3.032,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9119446661346102,
+      "grad_norm": 0.9707294550073895,
+      "learning_rate": 8.785007646117323e-05,
+      "loss": 3.0431,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9121574886938015,
+      "grad_norm": 1.1189968102603869,
+      "learning_rate": 8.784198338208189e-05,
+      "loss": 3.0517,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9123703112529928,
+      "grad_norm": 0.9627427444042251,
+      "learning_rate": 8.783388798153074e-05,
+      "loss": 3.0125,
+      "step": 4287
+    },
+    {
+      "epoch": 0.9125831338121841,
+      "grad_norm": 1.1408871016494628,
+      "learning_rate": 8.782579026001642e-05,
+      "loss": 3.0677,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9127959563713753,
+      "grad_norm": 1.0629021779751593,
+      "learning_rate": 8.78176902180357e-05,
+      "loss": 3.0428,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9130087789305666,
+      "grad_norm": 1.1475468802738038,
+      "learning_rate": 8.780958785608546e-05,
+      "loss": 3.0251,
+      "step": 4290
+    },
+    {
+      "epoch": 0.913221601489758,
+      "grad_norm": 1.117488820522359,
+      "learning_rate": 8.780148317466276e-05,
+      "loss": 3.0458,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9134344240489491,
+      "grad_norm": 1.021339670925438,
+      "learning_rate": 8.779337617426482e-05,
+      "loss": 3.0423,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9136472466081405,
+      "grad_norm": 1.2650784726995692,
+      "learning_rate": 8.778526685538893e-05,
+      "loss": 3.0031,
+      "step": 4293
+    },
+    {
+      "epoch": 0.9138600691673318,
+      "grad_norm": 0.9118423797019958,
+      "learning_rate": 8.777715521853258e-05,
+      "loss": 3.0417,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9140728917265231,
+      "grad_norm": 1.0473806276899975,
+      "learning_rate": 8.776904126419339e-05,
+      "loss": 3.0563,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 1.143025617887702,
+      "learning_rate": 8.776092499286912e-05,
+      "loss": 3.0251,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9144985368449056,
+      "grad_norm": 1.0591918307325903,
+      "learning_rate": 8.775280640505769e-05,
+      "loss": 3.0422,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9147113594040969,
+      "grad_norm": 1.1622603350829679,
+      "learning_rate": 8.77446855012571e-05,
+      "loss": 3.0537,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9149241819632881,
+      "grad_norm": 0.9578948304321335,
+      "learning_rate": 8.773656228196558e-05,
+      "loss": 3.0603,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9151370045224794,
+      "grad_norm": 1.0342588040883813,
+      "learning_rate": 8.772843674768143e-05,
+      "loss": 2.9999,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9153498270816707,
+      "grad_norm": 1.0358549110525355,
+      "learning_rate": 8.772030889890313e-05,
+      "loss": 3.0302,
+      "step": 4301
+    },
+    {
+      "epoch": 0.915562649640862,
+      "grad_norm": 1.070847198462848,
+      "learning_rate": 8.77121787361293e-05,
+      "loss": 3.0082,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9157754722000532,
+      "grad_norm": 1.0103680154978372,
+      "learning_rate": 8.770404625985867e-05,
+      "loss": 3.0503,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9159882947592445,
+      "grad_norm": 1.1871070630451537,
+      "learning_rate": 8.769591147059017e-05,
+      "loss": 3.0452,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9162011173184358,
+      "grad_norm": 0.9913237766936177,
+      "learning_rate": 8.768777436882282e-05,
+      "loss": 3.0767,
+      "step": 4305
+    },
+    {
+      "epoch": 0.916413939877627,
+      "grad_norm": 1.0256959444787936,
+      "learning_rate": 8.76796349550558e-05,
+      "loss": 3.0249,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9166267624368183,
+      "grad_norm": 1.1193511079492091,
+      "learning_rate": 8.767149322978843e-05,
+      "loss": 3.0582,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9168395849960096,
+      "grad_norm": 1.06389888171403,
+      "learning_rate": 8.766334919352017e-05,
+      "loss": 3.0164,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9170524075552009,
+      "grad_norm": 1.0992266371120993,
+      "learning_rate": 8.765520284675064e-05,
+      "loss": 3.0086,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9172652301143921,
+      "grad_norm": 1.1370092601217991,
+      "learning_rate": 8.764705418997957e-05,
+      "loss": 3.0664,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9174780526735834,
+      "grad_norm": 1.0221267566897185,
+      "learning_rate": 8.763890322370686e-05,
+      "loss": 2.998,
+      "step": 4311
+    },
+    {
+      "epoch": 0.9176908752327747,
+      "grad_norm": 1.025179931396014,
+      "learning_rate": 8.763074994843255e-05,
+      "loss": 3.0729,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9179036977919659,
+      "grad_norm": 0.9942636695253066,
+      "learning_rate": 8.762259436465677e-05,
+      "loss": 3.0353,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9181165203511572,
+      "grad_norm": 1.1867989256340914,
+      "learning_rate": 8.761443647287988e-05,
+      "loss": 3.0081,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9183293429103485,
+      "grad_norm": 0.9863635981322695,
+      "learning_rate": 8.760627627360232e-05,
+      "loss": 3.0487,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9185421654695398,
+      "grad_norm": 1.2040892922072106,
+      "learning_rate": 8.759811376732469e-05,
+      "loss": 3.0559,
+      "step": 4316
+    },
+    {
+      "epoch": 0.918754988028731,
+      "grad_norm": 0.9775832946393891,
+      "learning_rate": 8.75899489545477e-05,
+      "loss": 3.0193,
+      "step": 4317
+    },
+    {
+      "epoch": 0.9189678105879223,
+      "grad_norm": 1.2542197258856695,
+      "learning_rate": 8.758178183577228e-05,
+      "loss": 3.0548,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9191806331471136,
+      "grad_norm": 0.8597452919462456,
+      "learning_rate": 8.757361241149941e-05,
+      "loss": 3.0252,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9193934557063048,
+      "grad_norm": 1.1144918138201534,
+      "learning_rate": 8.756544068223027e-05,
+      "loss": 3.0152,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9196062782654961,
+      "grad_norm": 0.9295170358319575,
+      "learning_rate": 8.755726664846615e-05,
+      "loss": 2.993,
+      "step": 4321
+    },
+    {
+      "epoch": 0.9198191008246874,
+      "grad_norm": 0.9994937329768669,
+      "learning_rate": 8.754909031070851e-05,
+      "loss": 3.0234,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9200319233838787,
+      "grad_norm": 1.1488165254751774,
+      "learning_rate": 8.754091166945895e-05,
+      "loss": 3.0292,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9202447459430699,
+      "grad_norm": 1.183750980445788,
+      "learning_rate": 8.753273072521916e-05,
+      "loss": 3.0181,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9204575685022612,
+      "grad_norm": 0.9825806646977738,
+      "learning_rate": 8.752454747849103e-05,
+      "loss": 3.065,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9206703910614525,
+      "grad_norm": 1.113446110003373,
+      "learning_rate": 8.751636192977659e-05,
+      "loss": 3.0149,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9208832136206437,
+      "grad_norm": 0.9811937780426738,
+      "learning_rate": 8.750817407957796e-05,
+      "loss": 3.0304,
+      "step": 4327
+    },
+    {
+      "epoch": 0.921096036179835,
+      "grad_norm": 1.2213426624948502,
+      "learning_rate": 8.749998392839745e-05,
+      "loss": 3.0025,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9213088587390263,
+      "grad_norm": 0.9106265758469116,
+      "learning_rate": 8.74917914767375e-05,
+      "loss": 3.0847,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9215216812982177,
+      "grad_norm": 1.140089328789914,
+      "learning_rate": 8.748359672510066e-05,
+      "loss": 3.0708,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9217345038574088,
+      "grad_norm": 1.2408135628675774,
+      "learning_rate": 8.747539967398966e-05,
+      "loss": 3.0381,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9219473264166002,
+      "grad_norm": 0.8331393030219031,
+      "learning_rate": 8.746720032390737e-05,
+      "loss": 2.9766,
+      "step": 4332
+    },
+    {
+      "epoch": 0.9221601489757915,
+      "grad_norm": 1.0270836649185175,
+      "learning_rate": 8.745899867535678e-05,
+      "loss": 2.9969,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9223729715349827,
+      "grad_norm": 1.142209865577859,
+      "learning_rate": 8.745079472884103e-05,
+      "loss": 3.0266,
+      "step": 4334
+    },
+    {
+      "epoch": 0.922585794094174,
+      "grad_norm": 0.9725695215517953,
+      "learning_rate": 8.744258848486337e-05,
+      "loss": 3.0135,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9227986166533653,
+      "grad_norm": 1.084670665106283,
+      "learning_rate": 8.743437994392728e-05,
+      "loss": 3.0646,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9230114392125566,
+      "grad_norm": 1.0726867677033143,
+      "learning_rate": 8.742616910653628e-05,
+      "loss": 3.0346,
+      "step": 4337
+    },
+    {
+      "epoch": 0.9232242617717478,
+      "grad_norm": 1.1570460536652933,
+      "learning_rate": 8.741795597319409e-05,
+      "loss": 3.0052,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9234370843309391,
+      "grad_norm": 1.0495477021206234,
+      "learning_rate": 8.740974054440454e-05,
+      "loss": 2.9816,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9236499068901304,
+      "grad_norm": 1.0389274364071273,
+      "learning_rate": 8.740152282067163e-05,
+      "loss": 3.0204,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9238627294493216,
+      "grad_norm": 0.9938895360967724,
+      "learning_rate": 8.739330280249949e-05,
+      "loss": 2.9936,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9240755520085129,
+      "grad_norm": 1.1498661814859403,
+      "learning_rate": 8.738508049039235e-05,
+      "loss": 3.0075,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9242883745677042,
+      "grad_norm": 1.0241821223471383,
+      "learning_rate": 8.737685588485467e-05,
+      "loss": 3.0625,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9245011971268955,
+      "grad_norm": 1.1658032051614025,
+      "learning_rate": 8.736862898639096e-05,
+      "loss": 3.0281,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9247140196860867,
+      "grad_norm": 0.9464432638150649,
+      "learning_rate": 8.736039979550592e-05,
+      "loss": 3.0402,
+      "step": 4345
+    },
+    {
+      "epoch": 0.924926842245278,
+      "grad_norm": 1.178214308203852,
+      "learning_rate": 8.735216831270436e-05,
+      "loss": 2.9896,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9251396648044693,
+      "grad_norm": 0.9556698980955294,
+      "learning_rate": 8.73439345384913e-05,
+      "loss": 2.9879,
+      "step": 4347
+    },
+    {
+      "epoch": 0.9253524873636606,
+      "grad_norm": 1.0958788362834873,
+      "learning_rate": 8.73356984733718e-05,
+      "loss": 3.0125,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9255653099228518,
+      "grad_norm": 1.0759979479126227,
+      "learning_rate": 8.732746011785114e-05,
+      "loss": 3.0065,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9257781324820431,
+      "grad_norm": 1.1174948936307507,
+      "learning_rate": 8.731921947243469e-05,
+      "loss": 3.0246,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9259909550412344,
+      "grad_norm": 1.042007943946326,
+      "learning_rate": 8.7310976537628e-05,
+      "loss": 3.0573,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9262037776004256,
+      "grad_norm": 1.0685866663261534,
+      "learning_rate": 8.730273131393673e-05,
+      "loss": 3.049,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9264166001596169,
+      "grad_norm": 1.1103710775106441,
+      "learning_rate": 8.729448380186669e-05,
+      "loss": 3.0338,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9266294227188082,
+      "grad_norm": 0.8997586635446052,
+      "learning_rate": 8.728623400192387e-05,
+      "loss": 3.0015,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9268422452779995,
+      "grad_norm": 0.9960388534765393,
+      "learning_rate": 8.72779819146143e-05,
+      "loss": 2.9936,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9270550678371907,
+      "grad_norm": 1.1083012077914614,
+      "learning_rate": 8.726972754044427e-05,
+      "loss": 3.0324,
+      "step": 4356
+    },
+    {
+      "epoch": 0.927267890396382,
+      "grad_norm": 1.029995290288523,
+      "learning_rate": 8.726147087992012e-05,
+      "loss": 3.0075,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9274807129555733,
+      "grad_norm": 1.0444799796582251,
+      "learning_rate": 8.725321193354839e-05,
+      "loss": 3.0256,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9276935355147645,
+      "grad_norm": 1.1034064763645315,
+      "learning_rate": 8.724495070183572e-05,
+      "loss": 3.0606,
+      "step": 4359
+    },
+    {
+      "epoch": 0.9279063580739558,
+      "grad_norm": 1.1676416756548331,
+      "learning_rate": 8.723668718528889e-05,
+      "loss": 3.0487,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9281191806331471,
+      "grad_norm": 1.0381776789373494,
+      "learning_rate": 8.722842138441486e-05,
+      "loss": 3.0196,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9283320031923384,
+      "grad_norm": 1.1056242227869189,
+      "learning_rate": 8.722015329972069e-05,
+      "loss": 3.0221,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9285448257515296,
+      "grad_norm": 1.0527694103368865,
+      "learning_rate": 8.721188293171361e-05,
+      "loss": 3.0055,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9287576483107209,
+      "grad_norm": 1.1858245769193725,
+      "learning_rate": 8.720361028090095e-05,
+      "loss": 3.0363,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9289704708699122,
+      "grad_norm": 1.023833982140639,
+      "learning_rate": 8.719533534779023e-05,
+      "loss": 3.0088,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9291832934291034,
+      "grad_norm": 1.0715095221242952,
+      "learning_rate": 8.718705813288908e-05,
+      "loss": 3.0074,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9293961159882947,
+      "grad_norm": 1.1375774958516003,
+      "learning_rate": 8.717877863670527e-05,
+      "loss": 3.0133,
+      "step": 4367
+    },
+    {
+      "epoch": 0.929608938547486,
+      "grad_norm": 1.090802403796042,
+      "learning_rate": 8.717049685974673e-05,
+      "loss": 3.0094,
+      "step": 4368
+    },
+    {
+      "epoch": 0.9298217611066774,
+      "grad_norm": 1.0200919277531926,
+      "learning_rate": 8.716221280252147e-05,
+      "loss": 3.0411,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9300345836658686,
+      "grad_norm": 1.1041928271269432,
+      "learning_rate": 8.715392646553772e-05,
+      "loss": 3.0684,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9302474062250599,
+      "grad_norm": 0.99088325010941,
+      "learning_rate": 8.714563784930383e-05,
+      "loss": 3.0459,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9304602287842512,
+      "grad_norm": 1.190837494449514,
+      "learning_rate": 8.713734695432825e-05,
+      "loss": 2.9986,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9306730513434424,
+      "grad_norm": 0.9704912047558701,
+      "learning_rate": 8.712905378111961e-05,
+      "loss": 3.0321,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9308858739026337,
+      "grad_norm": 1.1851775778490743,
+      "learning_rate": 8.712075833018664e-05,
+      "loss": 3.0542,
+      "step": 4374
+    },
+    {
+      "epoch": 0.931098696461825,
+      "grad_norm": 1.0785732840490414,
+      "learning_rate": 8.711246060203826e-05,
+      "loss": 3.0443,
+      "step": 4375
+    },
+    {
+      "epoch": 0.9313115190210163,
+      "grad_norm": 1.092617526673495,
+      "learning_rate": 8.71041605971835e-05,
+      "loss": 2.9898,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9315243415802075,
+      "grad_norm": 1.0789931925214875,
+      "learning_rate": 8.70958583161315e-05,
+      "loss": 3.0035,
+      "step": 4377
+    },
+    {
+      "epoch": 0.9317371641393988,
+      "grad_norm": 1.0539510053735974,
+      "learning_rate": 8.708755375939161e-05,
+      "loss": 3.0003,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9319499866985901,
+      "grad_norm": 0.9808349021250093,
+      "learning_rate": 8.707924692747328e-05,
+      "loss": 3.0732,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9321628092577813,
+      "grad_norm": 1.1538807109135383,
+      "learning_rate": 8.707093782088609e-05,
+      "loss": 3.0371,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9323756318169726,
+      "grad_norm": 0.9133603980421874,
+      "learning_rate": 8.706262644013977e-05,
+      "loss": 3.0343,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9325884543761639,
+      "grad_norm": 0.9082327739016551,
+      "learning_rate": 8.70543127857442e-05,
+      "loss": 3.0029,
+      "step": 4382
+    },
+    {
+      "epoch": 0.9328012769353552,
+      "grad_norm": 0.9255471590779076,
+      "learning_rate": 8.704599685820937e-05,
+      "loss": 2.9952,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9330140994945464,
+      "grad_norm": 1.0977406284278852,
+      "learning_rate": 8.703767865804548e-05,
+      "loss": 3.0233,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9332269220537377,
+      "grad_norm": 1.0235849188399895,
+      "learning_rate": 8.702935818576276e-05,
+      "loss": 3.0178,
+      "step": 4385
+    },
+    {
+      "epoch": 0.933439744612929,
+      "grad_norm": 1.075530710103262,
+      "learning_rate": 8.702103544187167e-05,
+      "loss": 2.9749,
+      "step": 4386
+    },
+    {
+      "epoch": 0.9336525671721202,
+      "grad_norm": 1.1224193627462988,
+      "learning_rate": 8.701271042688277e-05,
+      "loss": 2.9973,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9338653897313115,
+      "grad_norm": 1.0249740090600887,
+      "learning_rate": 8.700438314130679e-05,
+      "loss": 3.0213,
+      "step": 4388
+    },
+    {
+      "epoch": 0.9340782122905028,
+      "grad_norm": 1.085897458712988,
+      "learning_rate": 8.699605358565454e-05,
+      "loss": 3.0026,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9342910348496941,
+      "grad_norm": 1.0642724604396472,
+      "learning_rate": 8.698772176043703e-05,
+      "loss": 3.0077,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9345038574088853,
+      "grad_norm": 0.9544238079230085,
+      "learning_rate": 8.697938766616539e-05,
+      "loss": 3.0438,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9347166799680766,
+      "grad_norm": 0.9291992462634308,
+      "learning_rate": 8.697105130335085e-05,
+      "loss": 3.0707,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9349295025272679,
+      "grad_norm": 1.1188189222035303,
+      "learning_rate": 8.696271267250484e-05,
+      "loss": 3.0256,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9351423250864591,
+      "grad_norm": 1.1191961660447998,
+      "learning_rate": 8.695437177413892e-05,
+      "loss": 3.0858,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9353551476456504,
+      "grad_norm": 0.9631952750222548,
+      "learning_rate": 8.694602860876475e-05,
+      "loss": 3.0157,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9355679702048417,
+      "grad_norm": 1.0994262999129714,
+      "learning_rate": 8.693768317689415e-05,
+      "loss": 3.029,
+      "step": 4396
+    },
+    {
+      "epoch": 0.935780792764033,
+      "grad_norm": 0.9922486076153388,
+      "learning_rate": 8.692933547903907e-05,
+      "loss": 3.0511,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9359936153232242,
+      "grad_norm": 1.0240077953393905,
+      "learning_rate": 8.692098551571164e-05,
+      "loss": 2.9951,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9362064378824155,
+      "grad_norm": 0.9751917062989829,
+      "learning_rate": 8.691263328742407e-05,
+      "loss": 3.0153,
+      "step": 4399
+    },
+    {
+      "epoch": 0.9364192604416068,
+      "grad_norm": 0.9747900065303902,
+      "learning_rate": 8.690427879468874e-05,
+      "loss": 3.0155,
+      "step": 4400
+    },
+    {
+      "epoch": 0.936632083000798,
+      "grad_norm": 1.1570668000180897,
+      "learning_rate": 8.68959220380182e-05,
+      "loss": 2.9879,
+      "step": 4401
+    },
+    {
+      "epoch": 0.9368449055599893,
+      "grad_norm": 0.9889256064132181,
+      "learning_rate": 8.688756301792507e-05,
+      "loss": 3.0627,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9370577281191806,
+      "grad_norm": 1.2686705865156953,
+      "learning_rate": 8.687920173492215e-05,
+      "loss": 3.1062,
+      "step": 4403
+    },
+    {
+      "epoch": 0.937270550678372,
+      "grad_norm": 0.8552076669341403,
+      "learning_rate": 8.687083818952235e-05,
+      "loss": 2.9672,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9374833732375631,
+      "grad_norm": 0.9441413446153616,
+      "learning_rate": 8.686247238223879e-05,
+      "loss": 3.0233,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9376961957967545,
+      "grad_norm": 0.9883922776824137,
+      "learning_rate": 8.685410431358464e-05,
+      "loss": 3.0337,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9379090183559458,
+      "grad_norm": 0.9957377477046081,
+      "learning_rate": 8.68457339840733e-05,
+      "loss": 2.9952,
+      "step": 4407
+    },
+    {
+      "epoch": 0.938121840915137,
+      "grad_norm": 0.9139605700805064,
+      "learning_rate": 8.68373613942182e-05,
+      "loss": 2.9954,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9383346634743283,
+      "grad_norm": 1.0534892172292698,
+      "learning_rate": 8.682898654453297e-05,
+      "loss": 3.029,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9385474860335196,
+      "grad_norm": 1.0282614279193663,
+      "learning_rate": 8.682060943553143e-05,
+      "loss": 3.0278,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9387603085927109,
+      "grad_norm": 1.1516899830669787,
+      "learning_rate": 8.681223006772742e-05,
+      "loss": 3.01,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9389731311519021,
+      "grad_norm": 1.1107176585479097,
+      "learning_rate": 8.680384844163501e-05,
+      "loss": 3.0303,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9391859537110934,
+      "grad_norm": 1.028047708721954,
+      "learning_rate": 8.679546455776839e-05,
+      "loss": 2.9965,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9393987762702847,
+      "grad_norm": 1.0626404619259124,
+      "learning_rate": 8.678707841664187e-05,
+      "loss": 3.015,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9396115988294759,
+      "grad_norm": 1.104186561253886,
+      "learning_rate": 8.67786900187699e-05,
+      "loss": 3.0298,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9398244213886672,
+      "grad_norm": 1.0419638304814258,
+      "learning_rate": 8.677029936466707e-05,
+      "loss": 3.0094,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9400372439478585,
+      "grad_norm": 1.0592126479211148,
+      "learning_rate": 8.676190645484813e-05,
+      "loss": 2.9931,
+      "step": 4417
+    },
+    {
+      "epoch": 0.9402500665070498,
+      "grad_norm": 1.1321129912127939,
+      "learning_rate": 8.675351128982795e-05,
+      "loss": 3.0247,
+      "step": 4418
+    },
+    {
+      "epoch": 0.940462889066241,
+      "grad_norm": 0.9812078356751334,
+      "learning_rate": 8.674511387012154e-05,
+      "loss": 3.0498,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9406757116254323,
+      "grad_norm": 1.2260096205786641,
+      "learning_rate": 8.673671419624405e-05,
+      "loss": 3.0303,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9408885341846236,
+      "grad_norm": 0.9870838751944324,
+      "learning_rate": 8.672831226871077e-05,
+      "loss": 3.0156,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9411013567438149,
+      "grad_norm": 1.1847913576970623,
+      "learning_rate": 8.671990808803712e-05,
+      "loss": 3.0113,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9413141793030061,
+      "grad_norm": 0.9248225209042467,
+      "learning_rate": 8.671150165473864e-05,
+      "loss": 2.97,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9415270018621974,
+      "grad_norm": 1.0679897774912572,
+      "learning_rate": 8.670309296933109e-05,
+      "loss": 2.9851,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9417398244213887,
+      "grad_norm": 1.024016626049789,
+      "learning_rate": 8.669468203233026e-05,
+      "loss": 3.0417,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9419526469805799,
+      "grad_norm": 1.0388838877119029,
+      "learning_rate": 8.668626884425217e-05,
+      "loss": 2.9835,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9421654695397712,
+      "grad_norm": 1.0922353845137411,
+      "learning_rate": 8.66778534056129e-05,
+      "loss": 3.0465,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9423782920989625,
+      "grad_norm": 0.9040896649729919,
+      "learning_rate": 8.666943571692872e-05,
+      "loss": 2.9869,
+      "step": 4428
+    },
+    {
+      "epoch": 0.9425911146581538,
+      "grad_norm": 0.9043386881155618,
+      "learning_rate": 8.666101577871601e-05,
+      "loss": 3.0,
+      "step": 4429
+    },
+    {
+      "epoch": 0.942803937217345,
+      "grad_norm": 1.099000349525922,
+      "learning_rate": 8.665259359149132e-05,
+      "loss": 3.0218,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9430167597765363,
+      "grad_norm": 0.8615882797389934,
+      "learning_rate": 8.664416915577132e-05,
+      "loss": 2.9925,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9432295823357276,
+      "grad_norm": 1.003880787749937,
+      "learning_rate": 8.66357424720728e-05,
+      "loss": 3.0529,
+      "step": 4432
+    },
+    {
+      "epoch": 0.9434424048949188,
+      "grad_norm": 1.036533338919063,
+      "learning_rate": 8.66273135409127e-05,
+      "loss": 3.0164,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9436552274541101,
+      "grad_norm": 0.9649073973422461,
+      "learning_rate": 8.661888236280813e-05,
+      "loss": 3.0295,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9438680500133014,
+      "grad_norm": 1.051351915214778,
+      "learning_rate": 8.661044893827628e-05,
+      "loss": 3.0661,
+      "step": 4435
+    },
+    {
+      "epoch": 0.9440808725724927,
+      "grad_norm": 1.184903241398577,
+      "learning_rate": 8.660201326783454e-05,
+      "loss": 3.0616,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9442936951316839,
+      "grad_norm": 1.0385222550995654,
+      "learning_rate": 8.659357535200038e-05,
+      "loss": 3.0369,
+      "step": 4437
+    },
+    {
+      "epoch": 0.9445065176908752,
+      "grad_norm": 0.9638495122631671,
+      "learning_rate": 8.658513519129143e-05,
+      "loss": 3.0325,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9447193402500665,
+      "grad_norm": 1.057179873919827,
+      "learning_rate": 8.65766927862255e-05,
+      "loss": 3.0413,
+      "step": 4439
+    },
+    {
+      "epoch": 0.9449321628092577,
+      "grad_norm": 1.0635811687545902,
+      "learning_rate": 8.656824813732045e-05,
+      "loss": 2.9875,
+      "step": 4440
+    },
+    {
+      "epoch": 0.945144985368449,
+      "grad_norm": 1.020916944897878,
+      "learning_rate": 8.655980124509436e-05,
+      "loss": 3.0186,
+      "step": 4441
+    },
+    {
+      "epoch": 0.9453578079276403,
+      "grad_norm": 1.0332978232439265,
+      "learning_rate": 8.655135211006541e-05,
+      "loss": 3.0228,
+      "step": 4442
+    },
+    {
+      "epoch": 0.9455706304868317,
+      "grad_norm": 1.054330759953503,
+      "learning_rate": 8.654290073275191e-05,
+      "loss": 3.0283,
+      "step": 4443
+    },
+    {
+      "epoch": 0.9457834530460228,
+      "grad_norm": 1.088536890331688,
+      "learning_rate": 8.653444711367231e-05,
+      "loss": 3.0518,
+      "step": 4444
+    },
+    {
+      "epoch": 0.9459962756052142,
+      "grad_norm": 1.0404473300174166,
+      "learning_rate": 8.652599125334524e-05,
+      "loss": 3.0405,
+      "step": 4445
+    },
+    {
+      "epoch": 0.9462090981644055,
+      "grad_norm": 1.2350456392308293,
+      "learning_rate": 8.65175331522894e-05,
+      "loss": 3.0023,
+      "step": 4446
+    },
+    {
+      "epoch": 0.9464219207235967,
+      "grad_norm": 0.8881636614859337,
+      "learning_rate": 8.65090728110237e-05,
+      "loss": 2.9953,
+      "step": 4447
+    },
+    {
+      "epoch": 0.946634743282788,
+      "grad_norm": 0.9905229348942948,
+      "learning_rate": 8.650061023006711e-05,
+      "loss": 3.0154,
+      "step": 4448
+    },
+    {
+      "epoch": 0.9468475658419793,
+      "grad_norm": 1.0430355859464984,
+      "learning_rate": 8.64921454099388e-05,
+      "loss": 3.0225,
+      "step": 4449
+    },
+    {
+      "epoch": 0.9470603884011706,
+      "grad_norm": 0.9918046594108855,
+      "learning_rate": 8.648367835115803e-05,
+      "loss": 3.0084,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9472732109603618,
+      "grad_norm": 1.144025593069989,
+      "learning_rate": 8.647520905424427e-05,
+      "loss": 2.996,
+      "step": 4451
+    },
+    {
+      "epoch": 0.9474860335195531,
+      "grad_norm": 0.9174682005505626,
+      "learning_rate": 8.646673751971703e-05,
+      "loss": 3.0177,
+      "step": 4452
+    },
+    {
+      "epoch": 0.9476988560787444,
+      "grad_norm": 1.1322236597328657,
+      "learning_rate": 8.645826374809603e-05,
+      "loss": 3.0409,
+      "step": 4453
+    },
+    {
+      "epoch": 0.9479116786379356,
+      "grad_norm": 0.990259230712121,
+      "learning_rate": 8.644978773990109e-05,
+      "loss": 3.0118,
+      "step": 4454
+    },
+    {
+      "epoch": 0.9481245011971269,
+      "grad_norm": 1.009915724302046,
+      "learning_rate": 8.644130949565219e-05,
+      "loss": 3.0105,
+      "step": 4455
+    },
+    {
+      "epoch": 0.9483373237563182,
+      "grad_norm": 1.0680380285909163,
+      "learning_rate": 8.643282901586945e-05,
+      "loss": 3.0064,
+      "step": 4456
+    },
+    {
+      "epoch": 0.9485501463155095,
+      "grad_norm": 1.2861349381782439,
+      "learning_rate": 8.642434630107307e-05,
+      "loss": 3.0197,
+      "step": 4457
+    },
+    {
+      "epoch": 0.9487629688747007,
+      "grad_norm": 0.8146771866216859,
+      "learning_rate": 8.64158613517835e-05,
+      "loss": 3.0187,
+      "step": 4458
+    },
+    {
+      "epoch": 0.948975791433892,
+      "grad_norm": 1.0494114131976953,
+      "learning_rate": 8.64073741685212e-05,
+      "loss": 3.0116,
+      "step": 4459
+    },
+    {
+      "epoch": 0.9491886139930833,
+      "grad_norm": 1.1042561313645785,
+      "learning_rate": 8.639888475180685e-05,
+      "loss": 2.9914,
+      "step": 4460
+    },
+    {
+      "epoch": 0.9494014365522745,
+      "grad_norm": 1.015946799425252,
+      "learning_rate": 8.639039310216123e-05,
+      "loss": 2.9942,
+      "step": 4461
+    },
+    {
+      "epoch": 0.9496142591114658,
+      "grad_norm": 1.0828815359012347,
+      "learning_rate": 8.638189922010529e-05,
+      "loss": 2.9963,
+      "step": 4462
+    },
+    {
+      "epoch": 0.9498270816706571,
+      "grad_norm": 0.9870573549893717,
+      "learning_rate": 8.637340310616009e-05,
+      "loss": 3.0223,
+      "step": 4463
+    },
+    {
+      "epoch": 0.9500399042298484,
+      "grad_norm": 1.0200479182405142,
+      "learning_rate": 8.636490476084682e-05,
+      "loss": 2.9921,
+      "step": 4464
+    },
+    {
+      "epoch": 0.9502527267890396,
+      "grad_norm": 1.1661228233873546,
+      "learning_rate": 8.635640418468684e-05,
+      "loss": 2.9864,
+      "step": 4465
+    },
+    {
+      "epoch": 0.9504655493482309,
+      "grad_norm": 1.0180200029172817,
+      "learning_rate": 8.634790137820162e-05,
+      "loss": 3.0352,
+      "step": 4466
+    },
+    {
+      "epoch": 0.9506783719074222,
+      "grad_norm": 1.170413211373194,
+      "learning_rate": 8.633939634191277e-05,
+      "loss": 2.9814,
+      "step": 4467
+    },
+    {
+      "epoch": 0.9508911944666134,
+      "grad_norm": 0.8608313773285586,
+      "learning_rate": 8.633088907634204e-05,
+      "loss": 3.0139,
+      "step": 4468
+    },
+    {
+      "epoch": 0.9511040170258047,
+      "grad_norm": 0.9251696807043075,
+      "learning_rate": 8.632237958201132e-05,
+      "loss": 3.0393,
+      "step": 4469
+    },
+    {
+      "epoch": 0.951316839584996,
+      "grad_norm": 0.9815837354264301,
+      "learning_rate": 8.631386785944265e-05,
+      "loss": 3.0352,
+      "step": 4470
+    },
+    {
+      "epoch": 0.9515296621441873,
+      "grad_norm": 1.161046158432766,
+      "learning_rate": 8.630535390915817e-05,
+      "loss": 2.9591,
+      "step": 4471
+    },
+    {
+      "epoch": 0.9517424847033785,
+      "grad_norm": 0.9890883598363434,
+      "learning_rate": 8.62968377316802e-05,
+      "loss": 2.9865,
+      "step": 4472
+    },
+    {
+      "epoch": 0.9519553072625698,
+      "grad_norm": 1.009411854615304,
+      "learning_rate": 8.628831932753114e-05,
+      "loss": 3.0163,
+      "step": 4473
+    },
+    {
+      "epoch": 0.9521681298217611,
+      "grad_norm": 1.2210251925302409,
+      "learning_rate": 8.62797986972336e-05,
+      "loss": 3.0231,
+      "step": 4474
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.8845269384722787,
+      "learning_rate": 8.627127584131026e-05,
+      "loss": 3.0038,
+      "step": 4475
+    },
+    {
+      "epoch": 0.9525937749401436,
+      "grad_norm": 1.1058268906400084,
+      "learning_rate": 8.626275076028397e-05,
+      "loss": 3.0439,
+      "step": 4476
+    },
+    {
+      "epoch": 0.9528065974993349,
+      "grad_norm": 0.9544152278858257,
+      "learning_rate": 8.625422345467773e-05,
+      "loss": 2.9791,
+      "step": 4477
+    },
+    {
+      "epoch": 0.9530194200585262,
+      "grad_norm": 1.0120538888644544,
+      "learning_rate": 8.624569392501463e-05,
+      "loss": 3.009,
+      "step": 4478
+    },
+    {
+      "epoch": 0.9532322426177174,
+      "grad_norm": 1.096759234503356,
+      "learning_rate": 8.623716217181795e-05,
+      "loss": 3.0621,
+      "step": 4479
+    },
+    {
+      "epoch": 0.9534450651769087,
+      "grad_norm": 1.0039031039019446,
+      "learning_rate": 8.622862819561106e-05,
+      "loss": 2.9931,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9536578877361,
+      "grad_norm": 1.0378648834887236,
+      "learning_rate": 8.622009199691749e-05,
+      "loss": 2.9979,
+      "step": 4481
+    },
+    {
+      "epoch": 0.9538707102952912,
+      "grad_norm": 1.1360635761932696,
+      "learning_rate": 8.62115535762609e-05,
+      "loss": 3.0327,
+      "step": 4482
+    },
+    {
+      "epoch": 0.9540835328544826,
+      "grad_norm": 1.0040176392256572,
+      "learning_rate": 8.62030129341651e-05,
+      "loss": 3.0319,
+      "step": 4483
+    },
+    {
+      "epoch": 0.9542963554136739,
+      "grad_norm": 1.0844831544162536,
+      "learning_rate": 8.619447007115403e-05,
+      "loss": 3.0007,
+      "step": 4484
+    },
+    {
+      "epoch": 0.9545091779728652,
+      "grad_norm": 1.0316135576445655,
+      "learning_rate": 8.618592498775175e-05,
+      "loss": 3.0143,
+      "step": 4485
+    },
+    {
+      "epoch": 0.9547220005320564,
+      "grad_norm": 0.9831515158502444,
+      "learning_rate": 8.617737768448247e-05,
+      "loss": 3.0541,
+      "step": 4486
+    },
+    {
+      "epoch": 0.9549348230912477,
+      "grad_norm": 1.0096134076947034,
+      "learning_rate": 8.616882816187053e-05,
+      "loss": 3.008,
+      "step": 4487
+    },
+    {
+      "epoch": 0.955147645650439,
+      "grad_norm": 1.099035195246314,
+      "learning_rate": 8.616027642044043e-05,
+      "loss": 3.0494,
+      "step": 4488
+    },
+    {
+      "epoch": 0.9553604682096303,
+      "grad_norm": 0.8828645175288171,
+      "learning_rate": 8.615172246071675e-05,
+      "loss": 3.0096,
+      "step": 4489
+    },
+    {
+      "epoch": 0.9555732907688215,
+      "grad_norm": 0.9473458597239137,
+      "learning_rate": 8.614316628322427e-05,
+      "loss": 3.023,
+      "step": 4490
+    },
+    {
+      "epoch": 0.9557861133280128,
+      "grad_norm": 1.0547936336649173,
+      "learning_rate": 8.613460788848788e-05,
+      "loss": 3.0013,
+      "step": 4491
+    },
+    {
+      "epoch": 0.9559989358872041,
+      "grad_norm": 1.0810983587261802,
+      "learning_rate": 8.612604727703258e-05,
+      "loss": 3.0304,
+      "step": 4492
+    },
+    {
+      "epoch": 0.9562117584463953,
+      "grad_norm": 1.0889735363500626,
+      "learning_rate": 8.611748444938357e-05,
+      "loss": 3.0293,
+      "step": 4493
+    },
+    {
+      "epoch": 0.9564245810055866,
+      "grad_norm": 0.8700063014881828,
+      "learning_rate": 8.61089194060661e-05,
+      "loss": 3.0139,
+      "step": 4494
+    },
+    {
+      "epoch": 0.9566374035647779,
+      "grad_norm": 0.8624065847510568,
+      "learning_rate": 8.610035214760565e-05,
+      "loss": 3.0054,
+      "step": 4495
+    },
+    {
+      "epoch": 0.9568502261239692,
+      "grad_norm": 0.9184586329830774,
+      "learning_rate": 8.609178267452774e-05,
+      "loss": 2.9905,
+      "step": 4496
+    },
+    {
+      "epoch": 0.9570630486831604,
+      "grad_norm": 0.9899597596094939,
+      "learning_rate": 8.608321098735811e-05,
+      "loss": 3.0073,
+      "step": 4497
+    },
+    {
+      "epoch": 0.9572758712423517,
+      "grad_norm": 1.200069522579625,
+      "learning_rate": 8.60746370866226e-05,
+      "loss": 3.0337,
+      "step": 4498
+    },
+    {
+      "epoch": 0.957488693801543,
+      "grad_norm": 1.0112815093982992,
+      "learning_rate": 8.606606097284715e-05,
+      "loss": 3.0296,
+      "step": 4499
+    },
+    {
+      "epoch": 0.9577015163607342,
+      "grad_norm": 0.9514991817673749,
+      "learning_rate": 8.605748264655791e-05,
+      "loss": 2.9908,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9579143389199255,
+      "grad_norm": 1.1811083901515371,
+      "learning_rate": 8.604890210828112e-05,
+      "loss": 3.0462,
+      "step": 4501
+    },
+    {
+      "epoch": 0.9581271614791168,
+      "grad_norm": 0.9327895585731456,
+      "learning_rate": 8.604031935854315e-05,
+      "loss": 3.0139,
+      "step": 4502
+    },
+    {
+      "epoch": 0.9583399840383081,
+      "grad_norm": 1.0092428607947301,
+      "learning_rate": 8.603173439787051e-05,
+      "loss": 3.0247,
+      "step": 4503
+    },
+    {
+      "epoch": 0.9585528065974993,
+      "grad_norm": 1.2026596201995625,
+      "learning_rate": 8.60231472267899e-05,
+      "loss": 3.0105,
+      "step": 4504
+    },
+    {
+      "epoch": 0.9587656291566906,
+      "grad_norm": 0.8756605846171066,
+      "learning_rate": 8.601455784582805e-05,
+      "loss": 3.0152,
+      "step": 4505
+    },
+    {
+      "epoch": 0.9589784517158819,
+      "grad_norm": 0.9051320494604502,
+      "learning_rate": 8.600596625551191e-05,
+      "loss": 3.0044,
+      "step": 4506
+    },
+    {
+      "epoch": 0.9591912742750731,
+      "grad_norm": 0.8628310220819909,
+      "learning_rate": 8.599737245636856e-05,
+      "loss": 3.0111,
+      "step": 4507
+    },
+    {
+      "epoch": 0.9594040968342644,
+      "grad_norm": 1.058060337454993,
+      "learning_rate": 8.598877644892519e-05,
+      "loss": 3.0196,
+      "step": 4508
+    },
+    {
+      "epoch": 0.9596169193934557,
+      "grad_norm": 1.0521654872772133,
+      "learning_rate": 8.598017823370912e-05,
+      "loss": 3.0092,
+      "step": 4509
+    },
+    {
+      "epoch": 0.959829741952647,
+      "grad_norm": 0.8811847820239408,
+      "learning_rate": 8.59715778112478e-05,
+      "loss": 2.9848,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9600425645118382,
+      "grad_norm": 0.9750523833894597,
+      "learning_rate": 8.596297518206889e-05,
+      "loss": 2.9712,
+      "step": 4511
+    },
+    {
+      "epoch": 0.9602553870710295,
+      "grad_norm": 1.272100510208057,
+      "learning_rate": 8.595437034670007e-05,
+      "loss": 3.021,
+      "step": 4512
+    },
+    {
+      "epoch": 0.9604682096302208,
+      "grad_norm": 0.9640132221670139,
+      "learning_rate": 8.594576330566922e-05,
+      "loss": 3.018,
+      "step": 4513
+    },
+    {
+      "epoch": 0.960681032189412,
+      "grad_norm": 1.0328082293519691,
+      "learning_rate": 8.593715405950437e-05,
+      "loss": 2.9811,
+      "step": 4514
+    },
+    {
+      "epoch": 0.9608938547486033,
+      "grad_norm": 1.0590662257874066,
+      "learning_rate": 8.592854260873367e-05,
+      "loss": 2.9991,
+      "step": 4515
+    },
+    {
+      "epoch": 0.9611066773077946,
+      "grad_norm": 0.9660152326812581,
+      "learning_rate": 8.591992895388539e-05,
+      "loss": 3.0014,
+      "step": 4516
+    },
+    {
+      "epoch": 0.961319499866986,
+      "grad_norm": 0.9560848570496494,
+      "learning_rate": 8.591131309548792e-05,
+      "loss": 3.0079,
+      "step": 4517
+    },
+    {
+      "epoch": 0.9615323224261771,
+      "grad_norm": 1.0135174376157847,
+      "learning_rate": 8.590269503406985e-05,
+      "loss": 2.9999,
+      "step": 4518
+    },
+    {
+      "epoch": 0.9617451449853684,
+      "grad_norm": 1.0508187117774856,
+      "learning_rate": 8.589407477015985e-05,
+      "loss": 3.0229,
+      "step": 4519
+    },
+    {
+      "epoch": 0.9619579675445598,
+      "grad_norm": 1.0410319888443746,
+      "learning_rate": 8.588545230428673e-05,
+      "loss": 3.004,
+      "step": 4520
+    },
+    {
+      "epoch": 0.962170790103751,
+      "grad_norm": 1.0148038599596798,
+      "learning_rate": 8.587682763697943e-05,
+      "loss": 3.0012,
+      "step": 4521
+    },
+    {
+      "epoch": 0.9623836126629423,
+      "grad_norm": 1.0369241025016438,
+      "learning_rate": 8.586820076876709e-05,
+      "loss": 2.9963,
+      "step": 4522
+    },
+    {
+      "epoch": 0.9625964352221336,
+      "grad_norm": 1.1862824095232243,
+      "learning_rate": 8.58595717001789e-05,
+      "loss": 2.9941,
+      "step": 4523
+    },
+    {
+      "epoch": 0.9628092577813249,
+      "grad_norm": 0.9246916994204114,
+      "learning_rate": 8.585094043174424e-05,
+      "loss": 3.0437,
+      "step": 4524
+    },
+    {
+      "epoch": 0.9630220803405161,
+      "grad_norm": 1.0091635941987025,
+      "learning_rate": 8.584230696399256e-05,
+      "loss": 3.0082,
+      "step": 4525
+    },
+    {
+      "epoch": 0.9632349028997074,
+      "grad_norm": 1.1017698787378007,
+      "learning_rate": 8.583367129745354e-05,
+      "loss": 3.0161,
+      "step": 4526
+    },
+    {
+      "epoch": 0.9634477254588987,
+      "grad_norm": 0.912491602771009,
+      "learning_rate": 8.582503343265694e-05,
+      "loss": 2.9691,
+      "step": 4527
+    },
+    {
+      "epoch": 0.9636605480180899,
+      "grad_norm": 1.0311525005528441,
+      "learning_rate": 8.581639337013262e-05,
+      "loss": 3.0059,
+      "step": 4528
+    },
+    {
+      "epoch": 0.9638733705772812,
+      "grad_norm": 1.0724057298518062,
+      "learning_rate": 8.580775111041064e-05,
+      "loss": 2.9761,
+      "step": 4529
+    },
+    {
+      "epoch": 0.9640861931364725,
+      "grad_norm": 0.9677893054319934,
+      "learning_rate": 8.579910665402118e-05,
+      "loss": 3.018,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9642990156956638,
+      "grad_norm": 1.1541195056581417,
+      "learning_rate": 8.579046000149454e-05,
+      "loss": 3.0261,
+      "step": 4531
+    },
+    {
+      "epoch": 0.964511838254855,
+      "grad_norm": 0.959598116517968,
+      "learning_rate": 8.578181115336115e-05,
+      "loss": 2.997,
+      "step": 4532
+    },
+    {
+      "epoch": 0.9647246608140463,
+      "grad_norm": 1.119630816828779,
+      "learning_rate": 8.577316011015158e-05,
+      "loss": 3.0149,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9649374833732376,
+      "grad_norm": 0.9192965690609086,
+      "learning_rate": 8.576450687239654e-05,
+      "loss": 3.0249,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9651503059324288,
+      "grad_norm": 1.07291755023516,
+      "learning_rate": 8.57558514406269e-05,
+      "loss": 2.991,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9653631284916201,
+      "grad_norm": 1.0568466710373177,
+      "learning_rate": 8.57471938153736e-05,
+      "loss": 3.0095,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9655759510508114,
+      "grad_norm": 1.172660444007652,
+      "learning_rate": 8.573853399716776e-05,
+      "loss": 3.0232,
+      "step": 4537
+    },
+    {
+      "epoch": 0.9657887736100027,
+      "grad_norm": 1.0916495084361049,
+      "learning_rate": 8.572987198654065e-05,
+      "loss": 3.0528,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9660015961691939,
+      "grad_norm": 1.2547720411108882,
+      "learning_rate": 8.572120778402362e-05,
+      "loss": 2.9924,
+      "step": 4539
+    },
+    {
+      "epoch": 0.9662144187283852,
+      "grad_norm": 0.8265673528056547,
+      "learning_rate": 8.57125413901482e-05,
+      "loss": 2.9962,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9664272412875765,
+      "grad_norm": 0.9130047803781013,
+      "learning_rate": 8.570387280544606e-05,
+      "loss": 2.9984,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9666400638467677,
+      "grad_norm": 1.1034986857035047,
+      "learning_rate": 8.569520203044893e-05,
+      "loss": 2.9638,
+      "step": 4542
+    },
+    {
+      "epoch": 0.966852886405959,
+      "grad_norm": 0.8149822484955327,
+      "learning_rate": 8.568652906568878e-05,
+      "loss": 3.032,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9670657089651503,
+      "grad_norm": 0.9476096166913395,
+      "learning_rate": 8.567785391169764e-05,
+      "loss": 3.041,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9672785315243416,
+      "grad_norm": 0.9794643200839463,
+      "learning_rate": 8.56691765690077e-05,
+      "loss": 3.032,
+      "step": 4545
+    },
+    {
+      "epoch": 0.9674913540835328,
+      "grad_norm": 1.031136713747681,
+      "learning_rate": 8.56604970381513e-05,
+      "loss": 3.0061,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9677041766427241,
+      "grad_norm": 1.1176184125800184,
+      "learning_rate": 8.565181531966086e-05,
+      "loss": 2.995,
+      "step": 4547
+    },
+    {
+      "epoch": 0.9679169992019154,
+      "grad_norm": 0.946366697649712,
+      "learning_rate": 8.5643131414069e-05,
+      "loss": 2.9691,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9681298217611066,
+      "grad_norm": 0.9753551379716667,
+      "learning_rate": 8.563444532190843e-05,
+      "loss": 2.9898,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9683426443202979,
+      "grad_norm": 1.125634356111292,
+      "learning_rate": 8.562575704371203e-05,
+      "loss": 3.0155,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9685554668794892,
+      "grad_norm": 0.9670334247145342,
+      "learning_rate": 8.561706658001275e-05,
+      "loss": 3.0132,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9687682894386805,
+      "grad_norm": 1.124116359530381,
+      "learning_rate": 8.560837393134376e-05,
+      "loss": 2.9836,
+      "step": 4552
+    },
+    {
+      "epoch": 0.9689811119978717,
+      "grad_norm": 0.9888904844078452,
+      "learning_rate": 8.559967909823829e-05,
+      "loss": 2.9894,
+      "step": 4553
+    },
+    {
+      "epoch": 0.969193934557063,
+      "grad_norm": 1.1607715058596595,
+      "learning_rate": 8.559098208122973e-05,
+      "loss": 3.0108,
+      "step": 4554
+    },
+    {
+      "epoch": 0.9694067571162543,
+      "grad_norm": 0.9660186287852598,
+      "learning_rate": 8.558228288085165e-05,
+      "loss": 3.0245,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9696195796754455,
+      "grad_norm": 1.1214006143885,
+      "learning_rate": 8.557358149763768e-05,
+      "loss": 3.0075,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9698324022346368,
+      "grad_norm": 0.9424715790187017,
+      "learning_rate": 8.556487793212162e-05,
+      "loss": 3.0092,
+      "step": 4557
+    },
+    {
+      "epoch": 0.9700452247938282,
+      "grad_norm": 1.2902581770928145,
+      "learning_rate": 8.55561721848374e-05,
+      "loss": 3.0279,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9702580473530195,
+      "grad_norm": 0.9322863973213088,
+      "learning_rate": 8.55474642563191e-05,
+      "loss": 2.9895,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9704708699122107,
+      "grad_norm": 0.9383828246892355,
+      "learning_rate": 8.553875414710088e-05,
+      "loss": 2.9811,
+      "step": 4560
+    },
+    {
+      "epoch": 0.970683692471402,
+      "grad_norm": 0.9503371607815138,
+      "learning_rate": 8.553004185771711e-05,
+      "loss": 2.9766,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9708965150305933,
+      "grad_norm": 1.1771184479665486,
+      "learning_rate": 8.552132738870226e-05,
+      "loss": 3.0303,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9711093375897846,
+      "grad_norm": 1.0503362382502448,
+      "learning_rate": 8.551261074059089e-05,
+      "loss": 2.9791,
+      "step": 4563
+    },
+    {
+      "epoch": 0.9713221601489758,
+      "grad_norm": 1.1096222880356188,
+      "learning_rate": 8.550389191391776e-05,
+      "loss": 2.9976,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9715349827081671,
+      "grad_norm": 0.968341651275381,
+      "learning_rate": 8.549517090921772e-05,
+      "loss": 2.9822,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9717478052673584,
+      "grad_norm": 1.011098338259892,
+      "learning_rate": 8.54864477270258e-05,
+      "loss": 3.0223,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9719606278265496,
+      "grad_norm": 1.0267125107171011,
+      "learning_rate": 8.54777223678771e-05,
+      "loss": 2.9853,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9721734503857409,
+      "grad_norm": 1.0092951231314544,
+      "learning_rate": 8.546899483230691e-05,
+      "loss": 2.974,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9723862729449322,
+      "grad_norm": 1.1164956545611677,
+      "learning_rate": 8.546026512085062e-05,
+      "loss": 3.0017,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9725990955041235,
+      "grad_norm": 0.9758260857309248,
+      "learning_rate": 8.545153323404376e-05,
+      "loss": 3.0002,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9728119180633147,
+      "grad_norm": 0.8731612943503658,
+      "learning_rate": 8.5442799172422e-05,
+      "loss": 2.9543,
+      "step": 4571
+    },
+    {
+      "epoch": 0.973024740622506,
+      "grad_norm": 0.9320268978460582,
+      "learning_rate": 8.543406293652117e-05,
+      "loss": 3.0126,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9732375631816973,
+      "grad_norm": 1.132463041846159,
+      "learning_rate": 8.542532452687715e-05,
+      "loss": 2.9968,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9734503857408885,
+      "grad_norm": 1.0275204885203044,
+      "learning_rate": 8.541658394402605e-05,
+      "loss": 2.9642,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9736632083000798,
+      "grad_norm": 1.231543175684484,
+      "learning_rate": 8.540784118850404e-05,
+      "loss": 3.051,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9738760308592711,
+      "grad_norm": 0.8530372034427182,
+      "learning_rate": 8.53990962608475e-05,
+      "loss": 2.9923,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9740888534184624,
+      "grad_norm": 0.9215143875163333,
+      "learning_rate": 8.539034916159285e-05,
+      "loss": 3.0098,
+      "step": 4577
+    },
+    {
+      "epoch": 0.9743016759776536,
+      "grad_norm": 1.0290888247874155,
+      "learning_rate": 8.538159989127672e-05,
+      "loss": 3.0179,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9745144985368449,
+      "grad_norm": 1.0052946602288888,
+      "learning_rate": 8.537284845043583e-05,
+      "loss": 3.0168,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9747273210960362,
+      "grad_norm": 0.9681009998398851,
+      "learning_rate": 8.536409483960704e-05,
+      "loss": 2.9945,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9749401436552274,
+      "grad_norm": 1.047856539148285,
+      "learning_rate": 8.535533905932738e-05,
+      "loss": 2.9711,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9751529662144187,
+      "grad_norm": 1.2324914919246763,
+      "learning_rate": 8.534658111013396e-05,
+      "loss": 3.0082,
+      "step": 4582
+    },
+    {
+      "epoch": 0.97536578877361,
+      "grad_norm": 0.9506995333339135,
+      "learning_rate": 8.533782099256405e-05,
+      "loss": 2.9653,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9755786113328013,
+      "grad_norm": 1.0931646015680647,
+      "learning_rate": 8.532905870715505e-05,
+      "loss": 3.0624,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9757914338919925,
+      "grad_norm": 1.0508084618448439,
+      "learning_rate": 8.53202942544445e-05,
+      "loss": 3.01,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9760042564511838,
+      "grad_norm": 1.1092990438301373,
+      "learning_rate": 8.531152763497005e-05,
+      "loss": 2.9991,
+      "step": 4586
+    },
+    {
+      "epoch": 0.9762170790103751,
+      "grad_norm": 0.957795363673452,
+      "learning_rate": 8.530275884926952e-05,
+      "loss": 2.9963,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9764299015695663,
+      "grad_norm": 1.0594367191513632,
+      "learning_rate": 8.529398789788081e-05,
+      "loss": 2.9845,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9766427241287576,
+      "grad_norm": 0.9949753679070187,
+      "learning_rate": 8.528521478134201e-05,
+      "loss": 3.0079,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9768555466879489,
+      "grad_norm": 1.260315773269589,
+      "learning_rate": 8.527643950019131e-05,
+      "loss": 3.0287,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9770683692471402,
+      "grad_norm": 0.9171476369796596,
+      "learning_rate": 8.526766205496705e-05,
+      "loss": 2.991,
+      "step": 4591
+    },
+    {
+      "epoch": 0.9772811918063314,
+      "grad_norm": 0.9802864194352426,
+      "learning_rate": 8.525888244620768e-05,
+      "loss": 2.9936,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9774940143655227,
+      "grad_norm": 1.1043355791638891,
+      "learning_rate": 8.525010067445178e-05,
+      "loss": 3.0053,
+      "step": 4593
+    },
+    {
+      "epoch": 0.977706836924714,
+      "grad_norm": 0.947674494834943,
+      "learning_rate": 8.52413167402381e-05,
+      "loss": 2.9832,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9779196594839052,
+      "grad_norm": 1.2213480073712977,
+      "learning_rate": 8.52325306441055e-05,
+      "loss": 3.0457,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9781324820430966,
+      "grad_norm": 1.0225751659639717,
+      "learning_rate": 8.522374238659296e-05,
+      "loss": 3.0306,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9783453046022879,
+      "grad_norm": 1.225472018036862,
+      "learning_rate": 8.521495196823963e-05,
+      "loss": 3.0444,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9785581271614792,
+      "grad_norm": 0.8129587749299382,
+      "learning_rate": 8.520615938958474e-05,
+      "loss": 2.9832,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9787709497206704,
+      "grad_norm": 0.9402390621363382,
+      "learning_rate": 8.51973646511677e-05,
+      "loss": 2.9808,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9789837722798617,
+      "grad_norm": 1.0543731200220423,
+      "learning_rate": 8.518856775352801e-05,
+      "loss": 2.9823,
+      "step": 4600
+    },
+    {
+      "epoch": 0.979196594839053,
+      "grad_norm": 1.0721426198398698,
+      "learning_rate": 8.517976869720537e-05,
+      "loss": 3.0115,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9794094173982442,
+      "grad_norm": 0.8834727248279804,
+      "learning_rate": 8.517096748273951e-05,
+      "loss": 3.0117,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9796222399574355,
+      "grad_norm": 1.0258669792694424,
+      "learning_rate": 8.516216411067039e-05,
+      "loss": 2.9955,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9798350625166268,
+      "grad_norm": 1.0833909604751668,
+      "learning_rate": 8.515335858153807e-05,
+      "loss": 2.9968,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9800478850758181,
+      "grad_norm": 0.9146447517153802,
+      "learning_rate": 8.51445508958827e-05,
+      "loss": 3.034,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9802607076350093,
+      "grad_norm": 1.0120926971977775,
+      "learning_rate": 8.513574105424462e-05,
+      "loss": 2.992,
+      "step": 4606
+    },
+    {
+      "epoch": 0.9804735301942006,
+      "grad_norm": 1.0494528311494136,
+      "learning_rate": 8.512692905716429e-05,
+      "loss": 3.0208,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9806863527533919,
+      "grad_norm": 1.1477792840945142,
+      "learning_rate": 8.511811490518227e-05,
+      "loss": 2.9483,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9808991753125831,
+      "grad_norm": 0.9240964836416098,
+      "learning_rate": 8.510929859883927e-05,
+      "loss": 2.9861,
+      "step": 4609
+    },
+    {
+      "epoch": 0.9811119978717744,
+      "grad_norm": 0.9666322508961107,
+      "learning_rate": 8.510048013867615e-05,
+      "loss": 3.006,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9813248204309657,
+      "grad_norm": 1.149245048465329,
+      "learning_rate": 8.509165952523391e-05,
+      "loss": 2.9967,
+      "step": 4611
+    },
+    {
+      "epoch": 0.981537642990157,
+      "grad_norm": 0.9230026541294014,
+      "learning_rate": 8.508283675905362e-05,
+      "loss": 3.034,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9817504655493482,
+      "grad_norm": 0.9633631162822898,
+      "learning_rate": 8.507401184067654e-05,
+      "loss": 2.986,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9819632881085395,
+      "grad_norm": 1.1590882724725693,
+      "learning_rate": 8.506518477064406e-05,
+      "loss": 2.9745,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9821761106677308,
+      "grad_norm": 1.0448898386588974,
+      "learning_rate": 8.505635554949767e-05,
+      "loss": 2.9818,
+      "step": 4615
+    },
+    {
+      "epoch": 0.982388933226922,
+      "grad_norm": 0.9968422192714034,
+      "learning_rate": 8.5047524177779e-05,
+      "loss": 3.0551,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9826017557861133,
+      "grad_norm": 0.9379490863174959,
+      "learning_rate": 8.503869065602983e-05,
+      "loss": 2.9801,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9828145783453046,
+      "grad_norm": 0.9680475216706556,
+      "learning_rate": 8.502985498479207e-05,
+      "loss": 3.0179,
+      "step": 4618
+    },
+    {
+      "epoch": 0.9830274009044959,
+      "grad_norm": 1.158369490031947,
+      "learning_rate": 8.502101716460775e-05,
+      "loss": 3.0093,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9832402234636871,
+      "grad_norm": 0.9664641035002763,
+      "learning_rate": 8.501217719601904e-05,
+      "loss": 2.9928,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9834530460228784,
+      "grad_norm": 1.0835561370126132,
+      "learning_rate": 8.500333507956822e-05,
+      "loss": 3.0066,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9836658685820697,
+      "grad_norm": 0.9980029275559179,
+      "learning_rate": 8.499449081579773e-05,
+      "loss": 3.0094,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9838786911412609,
+      "grad_norm": 1.1914491936985414,
+      "learning_rate": 8.498564440525014e-05,
+      "loss": 2.995,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9840915137004522,
+      "grad_norm": 0.9213309419289606,
+      "learning_rate": 8.497679584846815e-05,
+      "loss": 2.9959,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9843043362596435,
+      "grad_norm": 0.9840714519462322,
+      "learning_rate": 8.496794514599456e-05,
+      "loss": 2.9991,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9845171588188348,
+      "grad_norm": 1.2551349357624695,
+      "learning_rate": 8.495909229837234e-05,
+      "loss": 3.059,
+      "step": 4626
+    },
+    {
+      "epoch": 0.984729981378026,
+      "grad_norm": 0.8034908163593443,
+      "learning_rate": 8.495023730614457e-05,
+      "loss": 2.9796,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9849428039372173,
+      "grad_norm": 0.891123508001413,
+      "learning_rate": 8.494138016985449e-05,
+      "loss": 2.9996,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9851556264964086,
+      "grad_norm": 1.0563258205673813,
+      "learning_rate": 8.493252089004542e-05,
+      "loss": 2.9845,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9853684490556,
+      "grad_norm": 0.8719052448921978,
+      "learning_rate": 8.492365946726087e-05,
+      "loss": 3.0044,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9855812716147911,
+      "grad_norm": 1.0710807906987125,
+      "learning_rate": 8.491479590204446e-05,
+      "loss": 2.9771,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9857940941739824,
+      "grad_norm": 1.1380221342563803,
+      "learning_rate": 8.49059301949399e-05,
+      "loss": 3.0254,
+      "step": 4632
+    },
+    {
+      "epoch": 0.9860069167331738,
+      "grad_norm": 0.8075953195780368,
+      "learning_rate": 8.489706234649109e-05,
+      "loss": 3.0077,
+      "step": 4633
+    },
+    {
+      "epoch": 0.986219739292365,
+      "grad_norm": 1.0669868483968499,
+      "learning_rate": 8.488819235724204e-05,
+      "loss": 3.0007,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9864325618515563,
+      "grad_norm": 0.9371857834266868,
+      "learning_rate": 8.487932022773686e-05,
+      "loss": 2.9825,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9866453844107476,
+      "grad_norm": 0.9904368898289535,
+      "learning_rate": 8.487044595851987e-05,
+      "loss": 2.9833,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9868582069699389,
+      "grad_norm": 1.108732225398422,
+      "learning_rate": 8.486156955013545e-05,
+      "loss": 2.9797,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9870710295291301,
+      "grad_norm": 1.0038456261895925,
+      "learning_rate": 8.485269100312811e-05,
+      "loss": 3.0215,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9872838520883214,
+      "grad_norm": 1.3284425900425236,
+      "learning_rate": 8.484381031804257e-05,
+      "loss": 3.0047,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9874966746475127,
+      "grad_norm": 0.8299355637025292,
+      "learning_rate": 8.483492749542356e-05,
+      "loss": 2.9623,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9877094972067039,
+      "grad_norm": 1.0316467236195037,
+      "learning_rate": 8.482604253581605e-05,
+      "loss": 2.9904,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9879223197658952,
+      "grad_norm": 1.1433548688399064,
+      "learning_rate": 8.481715543976509e-05,
+      "loss": 3.0013,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9881351423250865,
+      "grad_norm": 0.9308799811097582,
+      "learning_rate": 8.480826620781586e-05,
+      "loss": 2.9816,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9883479648842778,
+      "grad_norm": 0.8790915187986785,
+      "learning_rate": 8.479937484051368e-05,
+      "loss": 2.9994,
+      "step": 4644
+    },
+    {
+      "epoch": 0.988560787443469,
+      "grad_norm": 1.090920156888323,
+      "learning_rate": 8.479048133840401e-05,
+      "loss": 3.0103,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9887736100026603,
+      "grad_norm": 1.097685991605724,
+      "learning_rate": 8.478158570203243e-05,
+      "loss": 3.0048,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9889864325618516,
+      "grad_norm": 1.0672739442055297,
+      "learning_rate": 8.477268793194467e-05,
+      "loss": 2.9966,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9891992551210428,
+      "grad_norm": 0.9177716845857326,
+      "learning_rate": 8.476378802868654e-05,
+      "loss": 3.0151,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9894120776802341,
+      "grad_norm": 1.0523620622806964,
+      "learning_rate": 8.475488599280402e-05,
+      "loss": 2.9968,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9896249002394254,
+      "grad_norm": 1.1665692203349152,
+      "learning_rate": 8.474598182484323e-05,
+      "loss": 3.0258,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9898377227986167,
+      "grad_norm": 1.0484048601183868,
+      "learning_rate": 8.473707552535042e-05,
+      "loss": 2.9744,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9900505453578079,
+      "grad_norm": 1.0298403642947098,
+      "learning_rate": 8.472816709487192e-05,
+      "loss": 3.0063,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9902633679169992,
+      "grad_norm": 1.1278405084763392,
+      "learning_rate": 8.471925653395426e-05,
+      "loss": 3.0115,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 1.1054056512903365,
+      "learning_rate": 8.471034384314404e-05,
+      "loss": 2.9975,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9906890130353817,
+      "grad_norm": 0.9467861906864441,
+      "learning_rate": 8.470142902298806e-05,
+      "loss": 2.9815,
+      "step": 4655
+    },
+    {
+      "epoch": 0.990901835594573,
+      "grad_norm": 1.040144526512872,
+      "learning_rate": 8.469251207403317e-05,
+      "loss": 3.0314,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9911146581537643,
+      "grad_norm": 1.1719783058454445,
+      "learning_rate": 8.468359299682641e-05,
+      "loss": 3.0108,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9913274807129556,
+      "grad_norm": 0.8886241711459665,
+      "learning_rate": 8.467467179191493e-05,
+      "loss": 3.0022,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9915403032721468,
+      "grad_norm": 0.9832429185630562,
+      "learning_rate": 8.466574845984601e-05,
+      "loss": 2.9824,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9917531258313381,
+      "grad_norm": 1.212781605744905,
+      "learning_rate": 8.465682300116704e-05,
+      "loss": 2.9678,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9919659483905294,
+      "grad_norm": 0.8153135393754267,
+      "learning_rate": 8.46478954164256e-05,
+      "loss": 2.9884,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9921787709497206,
+      "grad_norm": 1.061314134232022,
+      "learning_rate": 8.463896570616934e-05,
+      "loss": 3.023,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9923915935089119,
+      "grad_norm": 1.2147315039849906,
+      "learning_rate": 8.463003387094607e-05,
+      "loss": 3.0216,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9926044160681032,
+      "grad_norm": 1.0423842007314161,
+      "learning_rate": 8.462109991130372e-05,
+      "loss": 3.0053,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9928172386272945,
+      "grad_norm": 0.99167657870056,
+      "learning_rate": 8.461216382779036e-05,
+      "loss": 2.9832,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9930300611864857,
+      "grad_norm": 0.9727408346866591,
+      "learning_rate": 8.460322562095417e-05,
+      "loss": 2.95,
+      "step": 4666
+    },
+    {
+      "epoch": 0.993242883745677,
+      "grad_norm": 1.0601768444905602,
+      "learning_rate": 8.45942852913435e-05,
+      "loss": 2.9842,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9934557063048683,
+      "grad_norm": 0.9609478090419343,
+      "learning_rate": 8.458534283950679e-05,
+      "loss": 3.0151,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9936685288640595,
+      "grad_norm": 1.1537191155717992,
+      "learning_rate": 8.457639826599261e-05,
+      "loss": 3.0095,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9938813514232508,
+      "grad_norm": 0.9895888990186272,
+      "learning_rate": 8.456745157134971e-05,
+      "loss": 3.0182,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9940941739824422,
+      "grad_norm": 1.0326540252076768,
+      "learning_rate": 8.45585027561269e-05,
+      "loss": 2.9785,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9943069965416335,
+      "grad_norm": 1.041344328502243,
+      "learning_rate": 8.454955182087318e-05,
+      "loss": 2.9921,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9945198191008247,
+      "grad_norm": 1.2058229029482852,
+      "learning_rate": 8.454059876613765e-05,
+      "loss": 3.0364,
+      "step": 4673
+    },
+    {
+      "epoch": 0.994732641660016,
+      "grad_norm": 0.8910365977899318,
+      "learning_rate": 8.453164359246953e-05,
+      "loss": 2.9951,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9949454642192073,
+      "grad_norm": 1.002609160707218,
+      "learning_rate": 8.45226863004182e-05,
+      "loss": 2.9788,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9951582867783985,
+      "grad_norm": 1.1110493690388439,
+      "learning_rate": 8.451372689053314e-05,
+      "loss": 3.0134,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9953711093375898,
+      "grad_norm": 1.0353743418714363,
+      "learning_rate": 8.4504765363364e-05,
+      "loss": 2.9765,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9955839318967811,
+      "grad_norm": 1.0055251085575687,
+      "learning_rate": 8.449580171946052e-05,
+      "loss": 2.9974,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9957967544559724,
+      "grad_norm": 1.087711086892747,
+      "learning_rate": 8.448683595937258e-05,
+      "loss": 3.0185,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9960095770151636,
+      "grad_norm": 1.023332721157995,
+      "learning_rate": 8.447786808365022e-05,
+      "loss": 2.9892,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9962223995743549,
+      "grad_norm": 1.2889122736418386,
+      "learning_rate": 8.446889809284356e-05,
+      "loss": 3.0214,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9964352221335462,
+      "grad_norm": 0.8819313106465124,
+      "learning_rate": 8.445992598750289e-05,
+      "loss": 2.9899,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9966480446927374,
+      "grad_norm": 0.9614493251528208,
+      "learning_rate": 8.445095176817859e-05,
+      "loss": 2.9878,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9968608672519287,
+      "grad_norm": 1.0419824413447334,
+      "learning_rate": 8.444197543542121e-05,
+      "loss": 2.9881,
+      "step": 4684
+    },
+    {
+      "epoch": 0.99707368981112,
+      "grad_norm": 1.0540111957403413,
+      "learning_rate": 8.443299698978142e-05,
+      "loss": 3.0048,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9972865123703113,
+      "grad_norm": 0.9908167146527761,
+      "learning_rate": 8.442401643181e-05,
+      "loss": 2.9757,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9974993349295025,
+      "grad_norm": 1.0334782917086205,
+      "learning_rate": 8.441503376205788e-05,
+      "loss": 3.028,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9977121574886938,
+      "grad_norm": 1.204642577463568,
+      "learning_rate": 8.440604898107612e-05,
+      "loss": 2.9979,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9979249800478851,
+      "grad_norm": 0.8681628226602003,
+      "learning_rate": 8.439706208941587e-05,
+      "loss": 3.0176,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9981378026070763,
+      "grad_norm": 0.9639490108196416,
+      "learning_rate": 8.438807308762848e-05,
+      "loss": 2.9801,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9983506251662676,
+      "grad_norm": 1.2589956028728562,
+      "learning_rate": 8.437908197626536e-05,
+      "loss": 3.0121,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9985634477254589,
+      "grad_norm": 0.850063491550585,
+      "learning_rate": 8.43700887558781e-05,
+      "loss": 2.9971,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9987762702846502,
+      "grad_norm": 1.0553258756669182,
+      "learning_rate": 8.436109342701841e-05,
+      "loss": 2.9791,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9989890928438414,
+      "grad_norm": 1.0871128197977464,
+      "learning_rate": 8.435209599023808e-05,
+      "loss": 3.0084,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9992019154030327,
+      "grad_norm": 1.0188486055499406,
+      "learning_rate": 8.43430964460891e-05,
+      "loss": 2.9699,
+      "step": 4695
+    },
+    {
+      "epoch": 0.999414737962224,
+      "grad_norm": 1.0850205817207408,
+      "learning_rate": 8.433409479512356e-05,
+      "loss": 3.0195,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9996275605214152,
+      "grad_norm": 1.1739050753190632,
+      "learning_rate": 8.432509103789363e-05,
+      "loss": 2.9698,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9998403830806065,
+      "grad_norm": 0.889699535818436,
+      "learning_rate": 8.431608517495171e-05,
+      "loss": 2.983,
+      "step": 4698
+    },
+    {
+      "epoch": 1.0001596169193934,
+      "grad_norm": 1.7553438230249736,
+      "learning_rate": 8.430707720685027e-05,
+      "loss": 5.2931,
+      "step": 4699
+    },
+    {
+      "epoch": 1.0003724394785847,
+      "grad_norm": 1.312443089336489,
+      "learning_rate": 8.429806713414187e-05,
+      "loss": 2.996,
+      "step": 4700
+    },
+    {
+      "epoch": 1.000585262037776,
+      "grad_norm": 0.9633848921267233,
+      "learning_rate": 8.428905495737928e-05,
+      "loss": 2.9913,
+      "step": 4701
+    },
+    {
+      "epoch": 1.0007980845969673,
+      "grad_norm": 1.064905938788887,
+      "learning_rate": 8.428004067711536e-05,
+      "loss": 2.958,
+      "step": 4702
+    },
+    {
+      "epoch": 1.0010109071561586,
+      "grad_norm": 0.9208566632375287,
+      "learning_rate": 8.42710242939031e-05,
+      "loss": 2.9785,
+      "step": 4703
+    },
+    {
+      "epoch": 1.00122372971535,
+      "grad_norm": 1.02818804835895,
+      "learning_rate": 8.42620058082956e-05,
+      "loss": 2.9664,
+      "step": 4704
+    },
+    {
+      "epoch": 1.0014365522745412,
+      "grad_norm": 0.9453859528678653,
+      "learning_rate": 8.425298522084614e-05,
+      "loss": 3.0161,
+      "step": 4705
+    },
+    {
+      "epoch": 1.0016493748337323,
+      "grad_norm": 1.088456878946007,
+      "learning_rate": 8.42439625321081e-05,
+      "loss": 3.017,
+      "step": 4706
+    },
+    {
+      "epoch": 1.0018621973929236,
+      "grad_norm": 1.0311376603793934,
+      "learning_rate": 8.423493774263493e-05,
+      "loss": 3.0035,
+      "step": 4707
+    },
+    {
+      "epoch": 1.002075019952115,
+      "grad_norm": 1.0275589044400937,
+      "learning_rate": 8.422591085298032e-05,
+      "loss": 2.9674,
+      "step": 4708
+    },
+    {
+      "epoch": 1.0022878425113062,
+      "grad_norm": 1.0014371365856538,
+      "learning_rate": 8.421688186369804e-05,
+      "loss": 2.9852,
+      "step": 4709
+    },
+    {
+      "epoch": 1.0025006650704975,
+      "grad_norm": 1.1124369092492772,
+      "learning_rate": 8.420785077534196e-05,
+      "loss": 2.9705,
+      "step": 4710
+    },
+    {
+      "epoch": 1.0027134876296888,
+      "grad_norm": 1.024253315232201,
+      "learning_rate": 8.41988175884661e-05,
+      "loss": 2.9792,
+      "step": 4711
+    },
+    {
+      "epoch": 1.0029263101888801,
+      "grad_norm": 1.0092782904529458,
+      "learning_rate": 8.418978230362462e-05,
+      "loss": 2.982,
+      "step": 4712
+    },
+    {
+      "epoch": 1.0031391327480712,
+      "grad_norm": 1.1597748360214013,
+      "learning_rate": 8.41807449213718e-05,
+      "loss": 2.9935,
+      "step": 4713
+    },
+    {
+      "epoch": 1.0033519553072625,
+      "grad_norm": 1.0554119997418263,
+      "learning_rate": 8.417170544226204e-05,
+      "loss": 2.9764,
+      "step": 4714
+    },
+    {
+      "epoch": 1.0035647778664538,
+      "grad_norm": 0.9410040585138438,
+      "learning_rate": 8.41626638668499e-05,
+      "loss": 2.9938,
+      "step": 4715
+    },
+    {
+      "epoch": 1.0037776004256451,
+      "grad_norm": 1.1214336100692983,
+      "learning_rate": 8.415362019569002e-05,
+      "loss": 2.9774,
+      "step": 4716
+    },
+    {
+      "epoch": 1.0039904229848364,
+      "grad_norm": 0.9791694666197788,
+      "learning_rate": 8.414457442933721e-05,
+      "loss": 2.9746,
+      "step": 4717
+    },
+    {
+      "epoch": 1.0042032455440277,
+      "grad_norm": 1.1474922300191985,
+      "learning_rate": 8.413552656834637e-05,
+      "loss": 3.0012,
+      "step": 4718
+    },
+    {
+      "epoch": 1.004416068103219,
+      "grad_norm": 0.9854942108739603,
+      "learning_rate": 8.412647661327259e-05,
+      "loss": 2.9712,
+      "step": 4719
+    },
+    {
+      "epoch": 1.0046288906624101,
+      "grad_norm": 0.9933495497078636,
+      "learning_rate": 8.411742456467103e-05,
+      "loss": 2.9935,
+      "step": 4720
+    },
+    {
+      "epoch": 1.0048417132216014,
+      "grad_norm": 1.1054519572752735,
+      "learning_rate": 8.410837042309698e-05,
+      "loss": 3.0109,
+      "step": 4721
+    },
+    {
+      "epoch": 1.0050545357807927,
+      "grad_norm": 0.8884668602473781,
+      "learning_rate": 8.409931418910591e-05,
+      "loss": 2.9857,
+      "step": 4722
+    },
+    {
+      "epoch": 1.005267358339984,
+      "grad_norm": 1.0422634088358622,
+      "learning_rate": 8.409025586325336e-05,
+      "loss": 2.9905,
+      "step": 4723
+    },
+    {
+      "epoch": 1.0054801808991753,
+      "grad_norm": 0.9871129056898175,
+      "learning_rate": 8.408119544609504e-05,
+      "loss": 3.0035,
+      "step": 4724
+    },
+    {
+      "epoch": 1.0056930034583667,
+      "grad_norm": 1.0289709666510893,
+      "learning_rate": 8.407213293818676e-05,
+      "loss": 3.0082,
+      "step": 4725
+    },
+    {
+      "epoch": 1.005905826017558,
+      "grad_norm": 1.0563218884376508,
+      "learning_rate": 8.406306834008446e-05,
+      "loss": 3.0008,
+      "step": 4726
+    },
+    {
+      "epoch": 1.006118648576749,
+      "grad_norm": 0.8231093385876029,
+      "learning_rate": 8.405400165234426e-05,
+      "loss": 2.9431,
+      "step": 4727
+    },
+    {
+      "epoch": 1.0063314711359403,
+      "grad_norm": 0.9309987324167538,
+      "learning_rate": 8.404493287552232e-05,
+      "loss": 2.9755,
+      "step": 4728
+    },
+    {
+      "epoch": 1.0065442936951317,
+      "grad_norm": 0.9079377511861508,
+      "learning_rate": 8.4035862010175e-05,
+      "loss": 3.0027,
+      "step": 4729
+    },
+    {
+      "epoch": 1.006757116254323,
+      "grad_norm": 1.094845504893282,
+      "learning_rate": 8.402678905685875e-05,
+      "loss": 3.014,
+      "step": 4730
+    },
+    {
+      "epoch": 1.0069699388135143,
+      "grad_norm": 1.0306927722452706,
+      "learning_rate": 8.401771401613016e-05,
+      "loss": 2.9917,
+      "step": 4731
+    },
+    {
+      "epoch": 1.0071827613727056,
+      "grad_norm": 0.941195047149327,
+      "learning_rate": 8.400863688854597e-05,
+      "loss": 2.9705,
+      "step": 4732
+    },
+    {
+      "epoch": 1.0073955839318969,
+      "grad_norm": 0.9908198286844799,
+      "learning_rate": 8.3999557674663e-05,
+      "loss": 2.9951,
+      "step": 4733
+    },
+    {
+      "epoch": 1.007608406491088,
+      "grad_norm": 0.9762082658568045,
+      "learning_rate": 8.399047637503825e-05,
+      "loss": 2.9969,
+      "step": 4734
+    },
+    {
+      "epoch": 1.0078212290502793,
+      "grad_norm": 1.0515299581568351,
+      "learning_rate": 8.39813929902288e-05,
+      "loss": 2.9794,
+      "step": 4735
+    },
+    {
+      "epoch": 1.0080340516094706,
+      "grad_norm": 0.9780313281805151,
+      "learning_rate": 8.397230752079189e-05,
+      "loss": 3.0133,
+      "step": 4736
+    },
+    {
+      "epoch": 1.0082468741686619,
+      "grad_norm": 1.1162824525402208,
+      "learning_rate": 8.396321996728489e-05,
+      "loss": 2.9687,
+      "step": 4737
+    },
+    {
+      "epoch": 1.0084596967278532,
+      "grad_norm": 0.9827000643060806,
+      "learning_rate": 8.395413033026527e-05,
+      "loss": 2.9653,
+      "step": 4738
+    },
+    {
+      "epoch": 1.0086725192870445,
+      "grad_norm": 1.173790166229388,
+      "learning_rate": 8.394503861029063e-05,
+      "loss": 2.9859,
+      "step": 4739
+    },
+    {
+      "epoch": 1.0088853418462358,
+      "grad_norm": 1.0985282041070075,
+      "learning_rate": 8.393594480791875e-05,
+      "loss": 2.9793,
+      "step": 4740
+    },
+    {
+      "epoch": 1.0090981644054269,
+      "grad_norm": 0.993220251979564,
+      "learning_rate": 8.392684892370748e-05,
+      "loss": 2.9609,
+      "step": 4741
+    },
+    {
+      "epoch": 1.0093109869646182,
+      "grad_norm": 1.2069850351480396,
+      "learning_rate": 8.391775095821482e-05,
+      "loss": 2.9919,
+      "step": 4742
+    },
+    {
+      "epoch": 1.0095238095238095,
+      "grad_norm": 1.0252103212695791,
+      "learning_rate": 8.390865091199887e-05,
+      "loss": 2.9763,
+      "step": 4743
+    },
+    {
+      "epoch": 1.0097366320830008,
+      "grad_norm": 0.9588910159574637,
+      "learning_rate": 8.389954878561792e-05,
+      "loss": 2.9999,
+      "step": 4744
+    },
+    {
+      "epoch": 1.009949454642192,
+      "grad_norm": 1.0213007571366637,
+      "learning_rate": 8.389044457963035e-05,
+      "loss": 2.9892,
+      "step": 4745
+    },
+    {
+      "epoch": 1.0101622772013834,
+      "grad_norm": 1.0068069492297467,
+      "learning_rate": 8.388133829459464e-05,
+      "loss": 2.9948,
+      "step": 4746
+    },
+    {
+      "epoch": 1.0103750997605747,
+      "grad_norm": 1.1784743261328996,
+      "learning_rate": 8.387222993106945e-05,
+      "loss": 2.9923,
+      "step": 4747
+    },
+    {
+      "epoch": 1.0105879223197658,
+      "grad_norm": 0.9103276019452732,
+      "learning_rate": 8.386311948961352e-05,
+      "loss": 3.0023,
+      "step": 4748
+    },
+    {
+      "epoch": 1.010800744878957,
+      "grad_norm": 0.9924520757621965,
+      "learning_rate": 8.385400697078577e-05,
+      "loss": 2.9725,
+      "step": 4749
+    },
+    {
+      "epoch": 1.0110135674381484,
+      "grad_norm": 1.1797495135995337,
+      "learning_rate": 8.384489237514519e-05,
+      "loss": 2.9916,
+      "step": 4750
+    },
+    {
+      "epoch": 1.0112263899973397,
+      "grad_norm": 0.8618302998445827,
+      "learning_rate": 8.383577570325093e-05,
+      "loss": 2.9689,
+      "step": 4751
+    },
+    {
+      "epoch": 1.011439212556531,
+      "grad_norm": 0.9840788094323558,
+      "learning_rate": 8.382665695566228e-05,
+      "loss": 3.0029,
+      "step": 4752
+    },
+    {
+      "epoch": 1.0116520351157223,
+      "grad_norm": 1.1679519790779365,
+      "learning_rate": 8.381753613293861e-05,
+      "loss": 2.966,
+      "step": 4753
+    },
+    {
+      "epoch": 1.0118648576749136,
+      "grad_norm": 1.0717648581310708,
+      "learning_rate": 8.38084132356395e-05,
+      "loss": 3.042,
+      "step": 4754
+    },
+    {
+      "epoch": 1.0120776802341047,
+      "grad_norm": 1.0279257577911436,
+      "learning_rate": 8.379928826432455e-05,
+      "loss": 2.9451,
+      "step": 4755
+    },
+    {
+      "epoch": 1.012290502793296,
+      "grad_norm": 1.0782186936752078,
+      "learning_rate": 8.379016121955356e-05,
+      "loss": 2.9528,
+      "step": 4756
+    },
+    {
+      "epoch": 1.0125033253524873,
+      "grad_norm": 0.9671891177324807,
+      "learning_rate": 8.378103210188646e-05,
+      "loss": 2.9869,
+      "step": 4757
+    },
+    {
+      "epoch": 1.0127161479116786,
+      "grad_norm": 0.9583717937729627,
+      "learning_rate": 8.377190091188324e-05,
+      "loss": 2.9854,
+      "step": 4758
+    },
+    {
+      "epoch": 1.01292897047087,
+      "grad_norm": 1.0854824933272251,
+      "learning_rate": 8.376276765010412e-05,
+      "loss": 2.9835,
+      "step": 4759
+    },
+    {
+      "epoch": 1.0131417930300612,
+      "grad_norm": 0.9736606563977702,
+      "learning_rate": 8.375363231710933e-05,
+      "loss": 2.9389,
+      "step": 4760
+    },
+    {
+      "epoch": 1.0133546155892525,
+      "grad_norm": 1.0786255249525418,
+      "learning_rate": 8.374449491345935e-05,
+      "loss": 2.9849,
+      "step": 4761
+    },
+    {
+      "epoch": 1.0135674381484436,
+      "grad_norm": 0.953921103399243,
+      "learning_rate": 8.373535543971469e-05,
+      "loss": 2.9842,
+      "step": 4762
+    },
+    {
+      "epoch": 1.013780260707635,
+      "grad_norm": 1.031347040554551,
+      "learning_rate": 8.3726213896436e-05,
+      "loss": 2.9878,
+      "step": 4763
+    },
+    {
+      "epoch": 1.0139930832668262,
+      "grad_norm": 1.0797813968916459,
+      "learning_rate": 8.371707028418412e-05,
+      "loss": 2.9542,
+      "step": 4764
+    },
+    {
+      "epoch": 1.0142059058260176,
+      "grad_norm": 0.9452515541698566,
+      "learning_rate": 8.370792460351997e-05,
+      "loss": 2.958,
+      "step": 4765
+    },
+    {
+      "epoch": 1.0144187283852089,
+      "grad_norm": 0.9658587022694968,
+      "learning_rate": 8.369877685500458e-05,
+      "loss": 2.9685,
+      "step": 4766
+    },
+    {
+      "epoch": 1.0146315509444002,
+      "grad_norm": 1.1495503889277254,
+      "learning_rate": 8.368962703919913e-05,
+      "loss": 2.9975,
+      "step": 4767
+    },
+    {
+      "epoch": 1.0148443735035915,
+      "grad_norm": 0.978488223739071,
+      "learning_rate": 8.368047515666495e-05,
+      "loss": 2.9922,
+      "step": 4768
+    },
+    {
+      "epoch": 1.0150571960627826,
+      "grad_norm": 1.0883013018696417,
+      "learning_rate": 8.367132120796346e-05,
+      "loss": 2.9816,
+      "step": 4769
+    },
+    {
+      "epoch": 1.0152700186219739,
+      "grad_norm": 1.0377718227836699,
+      "learning_rate": 8.366216519365621e-05,
+      "loss": 3.0022,
+      "step": 4770
+    },
+    {
+      "epoch": 1.0154828411811652,
+      "grad_norm": 1.0971655388124681,
+      "learning_rate": 8.365300711430491e-05,
+      "loss": 2.9515,
+      "step": 4771
+    },
+    {
+      "epoch": 1.0156956637403565,
+      "grad_norm": 0.867433332398049,
+      "learning_rate": 8.364384697047134e-05,
+      "loss": 2.9885,
+      "step": 4772
+    },
+    {
+      "epoch": 1.0159084862995478,
+      "grad_norm": 0.9618709440579605,
+      "learning_rate": 8.363468476271747e-05,
+      "loss": 2.9643,
+      "step": 4773
+    },
+    {
+      "epoch": 1.016121308858739,
+      "grad_norm": 1.159873150707539,
+      "learning_rate": 8.362552049160534e-05,
+      "loss": 3.0076,
+      "step": 4774
+    },
+    {
+      "epoch": 1.0163341314179304,
+      "grad_norm": 0.9231883432590192,
+      "learning_rate": 8.361635415769718e-05,
+      "loss": 2.964,
+      "step": 4775
+    },
+    {
+      "epoch": 1.0165469539771215,
+      "grad_norm": 0.9940516185034293,
+      "learning_rate": 8.360718576155527e-05,
+      "loss": 2.971,
+      "step": 4776
+    },
+    {
+      "epoch": 1.0167597765363128,
+      "grad_norm": 0.9077342030895741,
+      "learning_rate": 8.359801530374205e-05,
+      "loss": 2.9678,
+      "step": 4777
+    },
+    {
+      "epoch": 1.016972599095504,
+      "grad_norm": 0.9841281901484763,
+      "learning_rate": 8.358884278482015e-05,
+      "loss": 3.0013,
+      "step": 4778
+    },
+    {
+      "epoch": 1.0171854216546954,
+      "grad_norm": 1.054159829950191,
+      "learning_rate": 8.35796682053522e-05,
+      "loss": 2.9443,
+      "step": 4779
+    },
+    {
+      "epoch": 1.0173982442138867,
+      "grad_norm": 0.9808913975385753,
+      "learning_rate": 8.357049156590106e-05,
+      "loss": 3.0137,
+      "step": 4780
+    },
+    {
+      "epoch": 1.017611066773078,
+      "grad_norm": 1.0538880967743396,
+      "learning_rate": 8.356131286702971e-05,
+      "loss": 3.0176,
+      "step": 4781
+    },
+    {
+      "epoch": 1.0178238893322693,
+      "grad_norm": 1.0752187992821256,
+      "learning_rate": 8.355213210930118e-05,
+      "loss": 2.956,
+      "step": 4782
+    },
+    {
+      "epoch": 1.0180367118914604,
+      "grad_norm": 0.9365302153771,
+      "learning_rate": 8.354294929327868e-05,
+      "loss": 2.9666,
+      "step": 4783
+    },
+    {
+      "epoch": 1.0182495344506517,
+      "grad_norm": 0.8653275521326327,
+      "learning_rate": 8.353376441952554e-05,
+      "loss": 2.9271,
+      "step": 4784
+    },
+    {
+      "epoch": 1.018462357009843,
+      "grad_norm": 1.0580619342733102,
+      "learning_rate": 8.352457748860524e-05,
+      "loss": 3.0008,
+      "step": 4785
+    },
+    {
+      "epoch": 1.0186751795690343,
+      "grad_norm": 1.0248889079737993,
+      "learning_rate": 8.351538850108136e-05,
+      "loss": 2.9974,
+      "step": 4786
+    },
+    {
+      "epoch": 1.0188880021282256,
+      "grad_norm": 0.9705534919070661,
+      "learning_rate": 8.350619745751759e-05,
+      "loss": 2.9668,
+      "step": 4787
+    },
+    {
+      "epoch": 1.019100824687417,
+      "grad_norm": 0.9441173414768967,
+      "learning_rate": 8.349700435847778e-05,
+      "loss": 2.9816,
+      "step": 4788
+    },
+    {
+      "epoch": 1.0193136472466082,
+      "grad_norm": 0.8287045904615008,
+      "learning_rate": 8.348780920452588e-05,
+      "loss": 2.9601,
+      "step": 4789
+    },
+    {
+      "epoch": 1.0195264698057993,
+      "grad_norm": 0.831053922331076,
+      "learning_rate": 8.347861199622597e-05,
+      "loss": 2.9852,
+      "step": 4790
+    },
+    {
+      "epoch": 1.0197392923649906,
+      "grad_norm": 0.9100136239858315,
+      "learning_rate": 8.346941273414231e-05,
+      "loss": 2.9822,
+      "step": 4791
+    },
+    {
+      "epoch": 1.019952114924182,
+      "grad_norm": 1.002782726880192,
+      "learning_rate": 8.346021141883917e-05,
+      "loss": 2.9864,
+      "step": 4792
+    },
+    {
+      "epoch": 1.0201649374833732,
+      "grad_norm": 0.8429450326904975,
+      "learning_rate": 8.345100805088106e-05,
+      "loss": 2.987,
+      "step": 4793
+    },
+    {
+      "epoch": 1.0203777600425645,
+      "grad_norm": 0.943713598358578,
+      "learning_rate": 8.344180263083257e-05,
+      "loss": 3.017,
+      "step": 4794
+    },
+    {
+      "epoch": 1.0205905826017558,
+      "grad_norm": 1.1536797267102779,
+      "learning_rate": 8.343259515925841e-05,
+      "loss": 2.974,
+      "step": 4795
+    },
+    {
+      "epoch": 1.0208034051609471,
+      "grad_norm": 0.9742556542182034,
+      "learning_rate": 8.34233856367234e-05,
+      "loss": 2.997,
+      "step": 4796
+    },
+    {
+      "epoch": 1.0210162277201382,
+      "grad_norm": 1.0891089084864827,
+      "learning_rate": 8.341417406379255e-05,
+      "loss": 2.9828,
+      "step": 4797
+    },
+    {
+      "epoch": 1.0212290502793295,
+      "grad_norm": 0.8067130908775743,
+      "learning_rate": 8.340496044103095e-05,
+      "loss": 2.9726,
+      "step": 4798
+    },
+    {
+      "epoch": 1.0214418728385208,
+      "grad_norm": 1.0146168996978986,
+      "learning_rate": 8.339574476900378e-05,
+      "loss": 2.9503,
+      "step": 4799
+    },
+    {
+      "epoch": 1.0216546953977121,
+      "grad_norm": 1.2219148012477201,
+      "learning_rate": 8.33865270482764e-05,
+      "loss": 2.9624,
+      "step": 4800
+    },
+    {
+      "epoch": 1.0218675179569034,
+      "grad_norm": 0.7533928857069948,
+      "learning_rate": 8.337730727941431e-05,
+      "loss": 2.9742,
+      "step": 4801
+    },
+    {
+      "epoch": 1.0220803405160948,
+      "grad_norm": 0.8713937714996606,
+      "learning_rate": 8.336808546298309e-05,
+      "loss": 2.9385,
+      "step": 4802
+    },
+    {
+      "epoch": 1.022293163075286,
+      "grad_norm": 1.0572074625405727,
+      "learning_rate": 8.335886159954846e-05,
+      "loss": 2.977,
+      "step": 4803
+    },
+    {
+      "epoch": 1.0225059856344774,
+      "grad_norm": 1.0229133821864906,
+      "learning_rate": 8.334963568967626e-05,
+      "loss": 2.9928,
+      "step": 4804
+    },
+    {
+      "epoch": 1.0227188081936684,
+      "grad_norm": 0.9977934025196294,
+      "learning_rate": 8.334040773393249e-05,
+      "loss": 3.0027,
+      "step": 4805
+    },
+    {
+      "epoch": 1.0229316307528598,
+      "grad_norm": 1.0164845955278787,
+      "learning_rate": 8.333117773288324e-05,
+      "loss": 2.9734,
+      "step": 4806
+    },
+    {
+      "epoch": 1.023144453312051,
+      "grad_norm": 1.0734384212419075,
+      "learning_rate": 8.332194568709471e-05,
+      "loss": 2.9454,
+      "step": 4807
+    },
+    {
+      "epoch": 1.0233572758712424,
+      "grad_norm": 0.9207968997366578,
+      "learning_rate": 8.331271159713327e-05,
+      "loss": 2.9863,
+      "step": 4808
+    },
+    {
+      "epoch": 1.0235700984304337,
+      "grad_norm": 0.9091679480925638,
+      "learning_rate": 8.330347546356541e-05,
+      "loss": 2.9617,
+      "step": 4809
+    },
+    {
+      "epoch": 1.023782920989625,
+      "grad_norm": 0.9544631708980155,
+      "learning_rate": 8.329423728695771e-05,
+      "loss": 3.0227,
+      "step": 4810
+    },
+    {
+      "epoch": 1.023995743548816,
+      "grad_norm": 1.142987988935605,
+      "learning_rate": 8.32849970678769e-05,
+      "loss": 2.972,
+      "step": 4811
+    },
+    {
+      "epoch": 1.0242085661080074,
+      "grad_norm": 1.0932916107947868,
+      "learning_rate": 8.327575480688985e-05,
+      "loss": 2.988,
+      "step": 4812
+    },
+    {
+      "epoch": 1.0244213886671987,
+      "grad_norm": 1.020064551159225,
+      "learning_rate": 8.32665105045635e-05,
+      "loss": 3.015,
+      "step": 4813
+    },
+    {
+      "epoch": 1.02463421122639,
+      "grad_norm": 0.9478135986847697,
+      "learning_rate": 8.325726416146499e-05,
+      "loss": 2.9945,
+      "step": 4814
+    },
+    {
+      "epoch": 1.0248470337855813,
+      "grad_norm": 0.9951701607824122,
+      "learning_rate": 8.324801577816152e-05,
+      "loss": 2.9624,
+      "step": 4815
+    },
+    {
+      "epoch": 1.0250598563447726,
+      "grad_norm": 1.1263944601416656,
+      "learning_rate": 8.323876535522048e-05,
+      "loss": 2.9619,
+      "step": 4816
+    },
+    {
+      "epoch": 1.025272678903964,
+      "grad_norm": 0.9637266663056305,
+      "learning_rate": 8.322951289320928e-05,
+      "loss": 2.971,
+      "step": 4817
+    },
+    {
+      "epoch": 1.0254855014631552,
+      "grad_norm": 1.0641545054298298,
+      "learning_rate": 8.32202583926956e-05,
+      "loss": 2.9872,
+      "step": 4818
+    },
+    {
+      "epoch": 1.0256983240223463,
+      "grad_norm": 1.0320239769501969,
+      "learning_rate": 8.321100185424713e-05,
+      "loss": 3.002,
+      "step": 4819
+    },
+    {
+      "epoch": 1.0259111465815376,
+      "grad_norm": 1.0913317808187046,
+      "learning_rate": 8.320174327843171e-05,
+      "loss": 2.9828,
+      "step": 4820
+    },
+    {
+      "epoch": 1.026123969140729,
+      "grad_norm": 1.1273570299454598,
+      "learning_rate": 8.319248266581735e-05,
+      "loss": 2.9744,
+      "step": 4821
+    },
+    {
+      "epoch": 1.0263367916999202,
+      "grad_norm": 1.0158995497922967,
+      "learning_rate": 8.318322001697214e-05,
+      "loss": 2.994,
+      "step": 4822
+    },
+    {
+      "epoch": 1.0265496142591115,
+      "grad_norm": 1.171049490609529,
+      "learning_rate": 8.317395533246428e-05,
+      "loss": 2.9475,
+      "step": 4823
+    },
+    {
+      "epoch": 1.0267624368183028,
+      "grad_norm": 0.8700727717424727,
+      "learning_rate": 8.316468861286217e-05,
+      "loss": 2.9629,
+      "step": 4824
+    },
+    {
+      "epoch": 1.0269752593774941,
+      "grad_norm": 0.9194389248445234,
+      "learning_rate": 8.315541985873425e-05,
+      "loss": 2.9714,
+      "step": 4825
+    },
+    {
+      "epoch": 1.0271880819366852,
+      "grad_norm": 1.0730695839266937,
+      "learning_rate": 8.314614907064914e-05,
+      "loss": 2.9906,
+      "step": 4826
+    },
+    {
+      "epoch": 1.0274009044958765,
+      "grad_norm": 1.01827006995946,
+      "learning_rate": 8.313687624917559e-05,
+      "loss": 2.9697,
+      "step": 4827
+    },
+    {
+      "epoch": 1.0276137270550678,
+      "grad_norm": 1.0631563118054277,
+      "learning_rate": 8.31276013948824e-05,
+      "loss": 2.9461,
+      "step": 4828
+    },
+    {
+      "epoch": 1.0278265496142591,
+      "grad_norm": 0.9921343242611936,
+      "learning_rate": 8.311832450833859e-05,
+      "loss": 2.9885,
+      "step": 4829
+    },
+    {
+      "epoch": 1.0280393721734504,
+      "grad_norm": 1.0622547199565595,
+      "learning_rate": 8.310904559011323e-05,
+      "loss": 2.9736,
+      "step": 4830
+    },
+    {
+      "epoch": 1.0282521947326417,
+      "grad_norm": 0.8592371159666642,
+      "learning_rate": 8.309976464077555e-05,
+      "loss": 2.9408,
+      "step": 4831
+    },
+    {
+      "epoch": 1.028465017291833,
+      "grad_norm": 0.9511322169717129,
+      "learning_rate": 8.309048166089493e-05,
+      "loss": 2.9846,
+      "step": 4832
+    },
+    {
+      "epoch": 1.0286778398510241,
+      "grad_norm": 1.1928682410389189,
+      "learning_rate": 8.308119665104081e-05,
+      "loss": 3.0019,
+      "step": 4833
+    },
+    {
+      "epoch": 1.0288906624102154,
+      "grad_norm": 0.9746970039552073,
+      "learning_rate": 8.307190961178282e-05,
+      "loss": 2.9578,
+      "step": 4834
+    },
+    {
+      "epoch": 1.0291034849694067,
+      "grad_norm": 1.106851368055745,
+      "learning_rate": 8.306262054369068e-05,
+      "loss": 2.9542,
+      "step": 4835
+    },
+    {
+      "epoch": 1.029316307528598,
+      "grad_norm": 0.9108538598750002,
+      "learning_rate": 8.305332944733419e-05,
+      "loss": 2.9618,
+      "step": 4836
+    },
+    {
+      "epoch": 1.0295291300877893,
+      "grad_norm": 0.9554401452330523,
+      "learning_rate": 8.30440363232834e-05,
+      "loss": 2.9818,
+      "step": 4837
+    },
+    {
+      "epoch": 1.0297419526469807,
+      "grad_norm": 0.9159769790372949,
+      "learning_rate": 8.303474117210834e-05,
+      "loss": 2.9741,
+      "step": 4838
+    },
+    {
+      "epoch": 1.029954775206172,
+      "grad_norm": 0.9111695239212317,
+      "learning_rate": 8.302544399437928e-05,
+      "loss": 2.9377,
+      "step": 4839
+    },
+    {
+      "epoch": 1.030167597765363,
+      "grad_norm": 0.980689951421543,
+      "learning_rate": 8.301614479066652e-05,
+      "loss": 2.9693,
+      "step": 4840
+    },
+    {
+      "epoch": 1.0303804203245543,
+      "grad_norm": 1.0634764840010458,
+      "learning_rate": 8.300684356154058e-05,
+      "loss": 2.9994,
+      "step": 4841
+    },
+    {
+      "epoch": 1.0305932428837457,
+      "grad_norm": 0.9555908555420524,
+      "learning_rate": 8.299754030757202e-05,
+      "loss": 2.9722,
+      "step": 4842
+    },
+    {
+      "epoch": 1.030806065442937,
+      "grad_norm": 1.0442857855629875,
+      "learning_rate": 8.298823502933157e-05,
+      "loss": 2.971,
+      "step": 4843
+    },
+    {
+      "epoch": 1.0310188880021283,
+      "grad_norm": 1.232021097010998,
+      "learning_rate": 8.297892772739007e-05,
+      "loss": 2.9955,
+      "step": 4844
+    },
+    {
+      "epoch": 1.0312317105613196,
+      "grad_norm": 0.9959106318339175,
+      "learning_rate": 8.29696184023185e-05,
+      "loss": 3.006,
+      "step": 4845
+    },
+    {
+      "epoch": 1.0314445331205109,
+      "grad_norm": 1.1917474666338,
+      "learning_rate": 8.296030705468794e-05,
+      "loss": 2.9474,
+      "step": 4846
+    },
+    {
+      "epoch": 1.031657355679702,
+      "grad_norm": 0.892065218413995,
+      "learning_rate": 8.29509936850696e-05,
+      "loss": 2.924,
+      "step": 4847
+    },
+    {
+      "epoch": 1.0318701782388933,
+      "grad_norm": 1.0110206744888122,
+      "learning_rate": 8.294167829403481e-05,
+      "loss": 2.9916,
+      "step": 4848
+    },
+    {
+      "epoch": 1.0320830007980846,
+      "grad_norm": 1.1167829964616025,
+      "learning_rate": 8.293236088215506e-05,
+      "loss": 2.9989,
+      "step": 4849
+    },
+    {
+      "epoch": 1.0322958233572759,
+      "grad_norm": 0.9666756501401146,
+      "learning_rate": 8.292304145000194e-05,
+      "loss": 3.0075,
+      "step": 4850
+    },
+    {
+      "epoch": 1.0325086459164672,
+      "grad_norm": 1.0564845273006533,
+      "learning_rate": 8.291371999814712e-05,
+      "loss": 2.9986,
+      "step": 4851
+    },
+    {
+      "epoch": 1.0327214684756585,
+      "grad_norm": 1.0595373664238603,
+      "learning_rate": 8.290439652716248e-05,
+      "loss": 2.9742,
+      "step": 4852
+    },
+    {
+      "epoch": 1.0329342910348498,
+      "grad_norm": 1.012420816717979,
+      "learning_rate": 8.289507103761996e-05,
+      "loss": 2.9553,
+      "step": 4853
+    },
+    {
+      "epoch": 1.0331471135940409,
+      "grad_norm": 1.041532223998427,
+      "learning_rate": 8.288574353009164e-05,
+      "loss": 2.9811,
+      "step": 4854
+    },
+    {
+      "epoch": 1.0333599361532322,
+      "grad_norm": 0.983090813884713,
+      "learning_rate": 8.287641400514974e-05,
+      "loss": 2.9916,
+      "step": 4855
+    },
+    {
+      "epoch": 1.0335727587124235,
+      "grad_norm": 1.0276379476194222,
+      "learning_rate": 8.286708246336656e-05,
+      "loss": 2.9868,
+      "step": 4856
+    },
+    {
+      "epoch": 1.0337855812716148,
+      "grad_norm": 1.0094406479325408,
+      "learning_rate": 8.285774890531459e-05,
+      "loss": 2.9977,
+      "step": 4857
+    },
+    {
+      "epoch": 1.033998403830806,
+      "grad_norm": 1.02490954066468,
+      "learning_rate": 8.284841333156639e-05,
+      "loss": 2.9596,
+      "step": 4858
+    },
+    {
+      "epoch": 1.0342112263899974,
+      "grad_norm": 1.0855480883420114,
+      "learning_rate": 8.283907574269467e-05,
+      "loss": 2.9764,
+      "step": 4859
+    },
+    {
+      "epoch": 1.0344240489491887,
+      "grad_norm": 1.0879461465327858,
+      "learning_rate": 8.282973613927225e-05,
+      "loss": 2.9968,
+      "step": 4860
+    },
+    {
+      "epoch": 1.0346368715083798,
+      "grad_norm": 1.0568011223098057,
+      "learning_rate": 8.282039452187206e-05,
+      "loss": 2.9778,
+      "step": 4861
+    },
+    {
+      "epoch": 1.034849694067571,
+      "grad_norm": 1.0376961921969776,
+      "learning_rate": 8.281105089106722e-05,
+      "loss": 2.952,
+      "step": 4862
+    },
+    {
+      "epoch": 1.0350625166267624,
+      "grad_norm": 1.0611040780265215,
+      "learning_rate": 8.280170524743087e-05,
+      "loss": 2.9611,
+      "step": 4863
+    },
+    {
+      "epoch": 1.0352753391859537,
+      "grad_norm": 1.0060145576880444,
+      "learning_rate": 8.279235759153637e-05,
+      "loss": 2.9747,
+      "step": 4864
+    },
+    {
+      "epoch": 1.035488161745145,
+      "grad_norm": 1.128573736369962,
+      "learning_rate": 8.278300792395713e-05,
+      "loss": 2.9583,
+      "step": 4865
+    },
+    {
+      "epoch": 1.0357009843043363,
+      "grad_norm": 0.956010743997867,
+      "learning_rate": 8.277365624526675e-05,
+      "loss": 2.9846,
+      "step": 4866
+    },
+    {
+      "epoch": 1.0359138068635276,
+      "grad_norm": 1.1045783170128627,
+      "learning_rate": 8.276430255603891e-05,
+      "loss": 2.9756,
+      "step": 4867
+    },
+    {
+      "epoch": 1.0361266294227187,
+      "grad_norm": 1.0171823530174755,
+      "learning_rate": 8.275494685684739e-05,
+      "loss": 2.9468,
+      "step": 4868
+    },
+    {
+      "epoch": 1.03633945198191,
+      "grad_norm": 1.138198538360139,
+      "learning_rate": 8.274558914826618e-05,
+      "loss": 3.0055,
+      "step": 4869
+    },
+    {
+      "epoch": 1.0365522745411013,
+      "grad_norm": 0.8089635648557334,
+      "learning_rate": 8.27362294308693e-05,
+      "loss": 2.9702,
+      "step": 4870
+    },
+    {
+      "epoch": 1.0367650971002926,
+      "grad_norm": 0.917265213471669,
+      "learning_rate": 8.272686770523094e-05,
+      "loss": 3.0315,
+      "step": 4871
+    },
+    {
+      "epoch": 1.036977919659484,
+      "grad_norm": 1.0225487842439198,
+      "learning_rate": 8.271750397192542e-05,
+      "loss": 2.9659,
+      "step": 4872
+    },
+    {
+      "epoch": 1.0371907422186752,
+      "grad_norm": 1.0476628369900824,
+      "learning_rate": 8.270813823152714e-05,
+      "loss": 2.9568,
+      "step": 4873
+    },
+    {
+      "epoch": 1.0374035647778665,
+      "grad_norm": 0.9275347567607982,
+      "learning_rate": 8.269877048461068e-05,
+      "loss": 2.9292,
+      "step": 4874
+    },
+    {
+      "epoch": 1.0376163873370576,
+      "grad_norm": 0.9475344895418667,
+      "learning_rate": 8.268940073175071e-05,
+      "loss": 2.969,
+      "step": 4875
+    },
+    {
+      "epoch": 1.037829209896249,
+      "grad_norm": 0.920570114920121,
+      "learning_rate": 8.268002897352203e-05,
+      "loss": 2.9491,
+      "step": 4876
+    },
+    {
+      "epoch": 1.0380420324554402,
+      "grad_norm": 0.9794631128530843,
+      "learning_rate": 8.267065521049955e-05,
+      "loss": 3.0021,
+      "step": 4877
+    },
+    {
+      "epoch": 1.0382548550146316,
+      "grad_norm": 1.0420500223408273,
+      "learning_rate": 8.266127944325833e-05,
+      "loss": 2.9882,
+      "step": 4878
+    },
+    {
+      "epoch": 1.0384676775738229,
+      "grad_norm": 0.9892311720517694,
+      "learning_rate": 8.26519016723735e-05,
+      "loss": 2.9474,
+      "step": 4879
+    },
+    {
+      "epoch": 1.0386805001330142,
+      "grad_norm": 1.1035450909199411,
+      "learning_rate": 8.26425218984204e-05,
+      "loss": 3.0129,
+      "step": 4880
+    },
+    {
+      "epoch": 1.0388933226922055,
+      "grad_norm": 0.989606718924138,
+      "learning_rate": 8.26331401219744e-05,
+      "loss": 2.9957,
+      "step": 4881
+    },
+    {
+      "epoch": 1.0391061452513966,
+      "grad_norm": 0.9680602015405866,
+      "learning_rate": 8.262375634361108e-05,
+      "loss": 3.0033,
+      "step": 4882
+    },
+    {
+      "epoch": 1.0393189678105879,
+      "grad_norm": 1.0757718081295675,
+      "learning_rate": 8.261437056390606e-05,
+      "loss": 2.9745,
+      "step": 4883
+    },
+    {
+      "epoch": 1.0395317903697792,
+      "grad_norm": 1.0885444670573299,
+      "learning_rate": 8.260498278343514e-05,
+      "loss": 2.9596,
+      "step": 4884
+    },
+    {
+      "epoch": 1.0397446129289705,
+      "grad_norm": 0.9978418352488602,
+      "learning_rate": 8.259559300277422e-05,
+      "loss": 2.944,
+      "step": 4885
+    },
+    {
+      "epoch": 1.0399574354881618,
+      "grad_norm": 0.9760631890040407,
+      "learning_rate": 8.258620122249933e-05,
+      "loss": 2.959,
+      "step": 4886
+    },
+    {
+      "epoch": 1.040170258047353,
+      "grad_norm": 1.1204955443297362,
+      "learning_rate": 8.257680744318662e-05,
+      "loss": 2.9838,
+      "step": 4887
+    },
+    {
+      "epoch": 1.0403830806065444,
+      "grad_norm": 0.9674986874793395,
+      "learning_rate": 8.256741166541234e-05,
+      "loss": 2.9945,
+      "step": 4888
+    },
+    {
+      "epoch": 1.0405959031657355,
+      "grad_norm": 1.0305294615594933,
+      "learning_rate": 8.255801388975292e-05,
+      "loss": 2.948,
+      "step": 4889
+    },
+    {
+      "epoch": 1.0408087257249268,
+      "grad_norm": 1.0192730788308153,
+      "learning_rate": 8.254861411678485e-05,
+      "loss": 2.9418,
+      "step": 4890
+    },
+    {
+      "epoch": 1.041021548284118,
+      "grad_norm": 1.00270251963147,
+      "learning_rate": 8.25392123470848e-05,
+      "loss": 2.9923,
+      "step": 4891
+    },
+    {
+      "epoch": 1.0412343708433094,
+      "grad_norm": 0.9755228482183645,
+      "learning_rate": 8.252980858122949e-05,
+      "loss": 2.9497,
+      "step": 4892
+    },
+    {
+      "epoch": 1.0414471934025007,
+      "grad_norm": 0.9910528297879753,
+      "learning_rate": 8.252040281979583e-05,
+      "loss": 2.93,
+      "step": 4893
+    },
+    {
+      "epoch": 1.041660015961692,
+      "grad_norm": 1.1377130416810728,
+      "learning_rate": 8.251099506336083e-05,
+      "loss": 2.9746,
+      "step": 4894
+    },
+    {
+      "epoch": 1.0418728385208833,
+      "grad_norm": 0.9145183535368198,
+      "learning_rate": 8.25015853125016e-05,
+      "loss": 2.9372,
+      "step": 4895
+    },
+    {
+      "epoch": 1.0420856610800744,
+      "grad_norm": 0.9361545734653351,
+      "learning_rate": 8.249217356779544e-05,
+      "loss": 2.9951,
+      "step": 4896
+    },
+    {
+      "epoch": 1.0422984836392657,
+      "grad_norm": 0.9398746694544126,
+      "learning_rate": 8.248275982981967e-05,
+      "loss": 2.9295,
+      "step": 4897
+    },
+    {
+      "epoch": 1.042511306198457,
+      "grad_norm": 0.9199395404735503,
+      "learning_rate": 8.247334409915181e-05,
+      "loss": 2.9217,
+      "step": 4898
+    },
+    {
+      "epoch": 1.0427241287576483,
+      "grad_norm": 1.0415947833880799,
+      "learning_rate": 8.246392637636945e-05,
+      "loss": 2.9565,
+      "step": 4899
+    },
+    {
+      "epoch": 1.0429369513168396,
+      "grad_norm": 1.009656021241841,
+      "learning_rate": 8.245450666205039e-05,
+      "loss": 2.98,
+      "step": 4900
+    },
+    {
+      "epoch": 1.043149773876031,
+      "grad_norm": 1.0636654438793407,
+      "learning_rate": 8.244508495677245e-05,
+      "loss": 2.9648,
+      "step": 4901
+    },
+    {
+      "epoch": 1.0433625964352222,
+      "grad_norm": 1.0087380235677021,
+      "learning_rate": 8.243566126111363e-05,
+      "loss": 2.9459,
+      "step": 4902
+    },
+    {
+      "epoch": 1.0435754189944133,
+      "grad_norm": 1.0393960733494472,
+      "learning_rate": 8.242623557565203e-05,
+      "loss": 2.9416,
+      "step": 4903
+    },
+    {
+      "epoch": 1.0437882415536046,
+      "grad_norm": 0.8499464945523952,
+      "learning_rate": 8.241680790096589e-05,
+      "loss": 2.991,
+      "step": 4904
+    },
+    {
+      "epoch": 1.044001064112796,
+      "grad_norm": 0.9163707166231363,
+      "learning_rate": 8.240737823763354e-05,
+      "loss": 2.9904,
+      "step": 4905
+    },
+    {
+      "epoch": 1.0442138866719872,
+      "grad_norm": 1.093273159422453,
+      "learning_rate": 8.239794658623346e-05,
+      "loss": 2.9748,
+      "step": 4906
+    },
+    {
+      "epoch": 1.0444267092311785,
+      "grad_norm": 1.0333683792102288,
+      "learning_rate": 8.238851294734428e-05,
+      "loss": 2.9958,
+      "step": 4907
+    },
+    {
+      "epoch": 1.0446395317903698,
+      "grad_norm": 1.1661904336597742,
+      "learning_rate": 8.237907732154467e-05,
+      "loss": 2.9297,
+      "step": 4908
+    },
+    {
+      "epoch": 1.0448523543495611,
+      "grad_norm": 0.9414637450590264,
+      "learning_rate": 8.236963970941349e-05,
+      "loss": 2.9342,
+      "step": 4909
+    },
+    {
+      "epoch": 1.0450651769087522,
+      "grad_norm": 0.9839276953494256,
+      "learning_rate": 8.23602001115297e-05,
+      "loss": 2.9568,
+      "step": 4910
+    },
+    {
+      "epoch": 1.0452779994679435,
+      "grad_norm": 0.9183587753122485,
+      "learning_rate": 8.23507585284724e-05,
+      "loss": 2.8992,
+      "step": 4911
+    },
+    {
+      "epoch": 1.0454908220271348,
+      "grad_norm": 0.8654564666130243,
+      "learning_rate": 8.234131496082076e-05,
+      "loss": 2.9675,
+      "step": 4912
+    },
+    {
+      "epoch": 1.0457036445863261,
+      "grad_norm": 0.8580109097532509,
+      "learning_rate": 8.233186940915412e-05,
+      "loss": 2.9552,
+      "step": 4913
+    },
+    {
+      "epoch": 1.0459164671455174,
+      "grad_norm": 0.9291013482587233,
+      "learning_rate": 8.232242187405194e-05,
+      "loss": 2.9838,
+      "step": 4914
+    },
+    {
+      "epoch": 1.0461292897047088,
+      "grad_norm": 1.0236623871552268,
+      "learning_rate": 8.231297235609379e-05,
+      "loss": 2.9729,
+      "step": 4915
+    },
+    {
+      "epoch": 1.0463421122639,
+      "grad_norm": 1.0423877622821076,
+      "learning_rate": 8.230352085585935e-05,
+      "loss": 2.9552,
+      "step": 4916
+    },
+    {
+      "epoch": 1.0465549348230911,
+      "grad_norm": 1.037972919341196,
+      "learning_rate": 8.229406737392843e-05,
+      "loss": 2.9728,
+      "step": 4917
+    },
+    {
+      "epoch": 1.0467677573822824,
+      "grad_norm": 0.9604785639071095,
+      "learning_rate": 8.228461191088099e-05,
+      "loss": 3.0122,
+      "step": 4918
+    },
+    {
+      "epoch": 1.0469805799414738,
+      "grad_norm": 1.0334734541641162,
+      "learning_rate": 8.227515446729705e-05,
+      "loss": 2.9355,
+      "step": 4919
+    },
+    {
+      "epoch": 1.047193402500665,
+      "grad_norm": 0.9099578860986632,
+      "learning_rate": 8.22656950437568e-05,
+      "loss": 2.9788,
+      "step": 4920
+    },
+    {
+      "epoch": 1.0474062250598564,
+      "grad_norm": 0.9821490665807946,
+      "learning_rate": 8.225623364084057e-05,
+      "loss": 2.9735,
+      "step": 4921
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 1.2432304832228198,
+      "learning_rate": 8.224677025912874e-05,
+      "loss": 3.0024,
+      "step": 4922
+    },
+    {
+      "epoch": 1.047831870178239,
+      "grad_norm": 0.7881869435819551,
+      "learning_rate": 8.223730489920187e-05,
+      "loss": 2.9308,
+      "step": 4923
+    },
+    {
+      "epoch": 1.04804469273743,
+      "grad_norm": 0.9611575111841762,
+      "learning_rate": 8.22278375616406e-05,
+      "loss": 2.9312,
+      "step": 4924
+    },
+    {
+      "epoch": 1.0482575152966214,
+      "grad_norm": 1.1883859756369897,
+      "learning_rate": 8.221836824702577e-05,
+      "loss": 2.9809,
+      "step": 4925
+    },
+    {
+      "epoch": 1.0484703378558127,
+      "grad_norm": 0.8089183896639949,
+      "learning_rate": 8.220889695593823e-05,
+      "loss": 2.9553,
+      "step": 4926
+    },
+    {
+      "epoch": 1.048683160415004,
+      "grad_norm": 0.9509617623792441,
+      "learning_rate": 8.219942368895904e-05,
+      "loss": 2.9737,
+      "step": 4927
+    },
+    {
+      "epoch": 1.0488959829741953,
+      "grad_norm": 0.9859159748363463,
+      "learning_rate": 8.218994844666934e-05,
+      "loss": 2.9527,
+      "step": 4928
+    },
+    {
+      "epoch": 1.0491088055333866,
+      "grad_norm": 1.078750105794065,
+      "learning_rate": 8.218047122965038e-05,
+      "loss": 2.9712,
+      "step": 4929
+    },
+    {
+      "epoch": 1.049321628092578,
+      "grad_norm": 1.003733502317901,
+      "learning_rate": 8.217099203848357e-05,
+      "loss": 2.9826,
+      "step": 4930
+    },
+    {
+      "epoch": 1.0495344506517692,
+      "grad_norm": 1.050323771436913,
+      "learning_rate": 8.216151087375044e-05,
+      "loss": 2.98,
+      "step": 4931
+    },
+    {
+      "epoch": 1.0497472732109603,
+      "grad_norm": 1.015474264214342,
+      "learning_rate": 8.215202773603258e-05,
+      "loss": 2.9393,
+      "step": 4932
+    },
+    {
+      "epoch": 1.0499600957701516,
+      "grad_norm": 1.07778029731386,
+      "learning_rate": 8.214254262591176e-05,
+      "loss": 2.9355,
+      "step": 4933
+    },
+    {
+      "epoch": 1.050172918329343,
+      "grad_norm": 0.8551875538072051,
+      "learning_rate": 8.213305554396989e-05,
+      "loss": 2.9645,
+      "step": 4934
+    },
+    {
+      "epoch": 1.0503857408885342,
+      "grad_norm": 0.945355479867686,
+      "learning_rate": 8.212356649078892e-05,
+      "loss": 2.9773,
+      "step": 4935
+    },
+    {
+      "epoch": 1.0505985634477255,
+      "grad_norm": 0.9560619176174896,
+      "learning_rate": 8.211407546695097e-05,
+      "loss": 2.9863,
+      "step": 4936
+    },
+    {
+      "epoch": 1.0508113860069168,
+      "grad_norm": 1.077285964851622,
+      "learning_rate": 8.210458247303832e-05,
+      "loss": 2.9526,
+      "step": 4937
+    },
+    {
+      "epoch": 1.051024208566108,
+      "grad_norm": 1.0529076455015283,
+      "learning_rate": 8.209508750963328e-05,
+      "loss": 2.9704,
+      "step": 4938
+    },
+    {
+      "epoch": 1.0512370311252992,
+      "grad_norm": 1.0091592195118324,
+      "learning_rate": 8.208559057731836e-05,
+      "loss": 2.9283,
+      "step": 4939
+    },
+    {
+      "epoch": 1.0514498536844905,
+      "grad_norm": 0.9087472039125373,
+      "learning_rate": 8.207609167667614e-05,
+      "loss": 3.0081,
+      "step": 4940
+    },
+    {
+      "epoch": 1.0516626762436818,
+      "grad_norm": 0.8710923676902543,
+      "learning_rate": 8.206659080828936e-05,
+      "loss": 2.9681,
+      "step": 4941
+    },
+    {
+      "epoch": 1.0518754988028731,
+      "grad_norm": 0.9533968838849503,
+      "learning_rate": 8.205708797274086e-05,
+      "loss": 2.9827,
+      "step": 4942
+    },
+    {
+      "epoch": 1.0520883213620644,
+      "grad_norm": 1.1733770426187737,
+      "learning_rate": 8.204758317061357e-05,
+      "loss": 2.9824,
+      "step": 4943
+    },
+    {
+      "epoch": 1.0523011439212557,
+      "grad_norm": 0.9465904683108032,
+      "learning_rate": 8.203807640249062e-05,
+      "loss": 2.9615,
+      "step": 4944
+    },
+    {
+      "epoch": 1.052513966480447,
+      "grad_norm": 0.9697838528562955,
+      "learning_rate": 8.202856766895517e-05,
+      "loss": 3.0112,
+      "step": 4945
+    },
+    {
+      "epoch": 1.0527267890396381,
+      "grad_norm": 1.1471448135309121,
+      "learning_rate": 8.201905697059057e-05,
+      "loss": 2.9665,
+      "step": 4946
+    },
+    {
+      "epoch": 1.0529396115988294,
+      "grad_norm": 0.9077690076748137,
+      "learning_rate": 8.200954430798027e-05,
+      "loss": 2.9827,
+      "step": 4947
+    },
+    {
+      "epoch": 1.0531524341580207,
+      "grad_norm": 0.8945690727724476,
+      "learning_rate": 8.200002968170783e-05,
+      "loss": 2.9735,
+      "step": 4948
+    },
+    {
+      "epoch": 1.053365256717212,
+      "grad_norm": 0.9326300790261497,
+      "learning_rate": 8.199051309235691e-05,
+      "loss": 2.9628,
+      "step": 4949
+    },
+    {
+      "epoch": 1.0535780792764033,
+      "grad_norm": 0.8861291995597869,
+      "learning_rate": 8.198099454051136e-05,
+      "loss": 2.9919,
+      "step": 4950
+    },
+    {
+      "epoch": 1.0537909018355947,
+      "grad_norm": 0.945898826245514,
+      "learning_rate": 8.197147402675507e-05,
+      "loss": 2.9527,
+      "step": 4951
+    },
+    {
+      "epoch": 1.0540037243947857,
+      "grad_norm": 1.0881003246496117,
+      "learning_rate": 8.19619515516721e-05,
+      "loss": 2.9651,
+      "step": 4952
+    },
+    {
+      "epoch": 1.054216546953977,
+      "grad_norm": 0.9574249413104418,
+      "learning_rate": 8.195242711584661e-05,
+      "loss": 2.9768,
+      "step": 4953
+    },
+    {
+      "epoch": 1.0544293695131683,
+      "grad_norm": 1.0284651234122062,
+      "learning_rate": 8.194290071986291e-05,
+      "loss": 2.9494,
+      "step": 4954
+    },
+    {
+      "epoch": 1.0546421920723597,
+      "grad_norm": 0.9691314411179678,
+      "learning_rate": 8.193337236430538e-05,
+      "loss": 2.97,
+      "step": 4955
+    },
+    {
+      "epoch": 1.054855014631551,
+      "grad_norm": 0.9288571361116604,
+      "learning_rate": 8.192384204975857e-05,
+      "loss": 2.9875,
+      "step": 4956
+    },
+    {
+      "epoch": 1.0550678371907423,
+      "grad_norm": 0.9017046797412063,
+      "learning_rate": 8.191430977680711e-05,
+      "loss": 2.9837,
+      "step": 4957
+    },
+    {
+      "epoch": 1.0552806597499336,
+      "grad_norm": 0.9326220611856112,
+      "learning_rate": 8.190477554603578e-05,
+      "loss": 2.9597,
+      "step": 4958
+    },
+    {
+      "epoch": 1.0554934823091249,
+      "grad_norm": 0.9568533068626743,
+      "learning_rate": 8.189523935802948e-05,
+      "loss": 3.0108,
+      "step": 4959
+    },
+    {
+      "epoch": 1.055706304868316,
+      "grad_norm": 1.0199041040081003,
+      "learning_rate": 8.188570121337318e-05,
+      "loss": 2.9409,
+      "step": 4960
+    },
+    {
+      "epoch": 1.0559191274275073,
+      "grad_norm": 1.0101013708579352,
+      "learning_rate": 8.187616111265203e-05,
+      "loss": 2.966,
+      "step": 4961
+    },
+    {
+      "epoch": 1.0561319499866986,
+      "grad_norm": 1.0699621837994833,
+      "learning_rate": 8.18666190564513e-05,
+      "loss": 2.9643,
+      "step": 4962
+    },
+    {
+      "epoch": 1.0563447725458899,
+      "grad_norm": 1.0381097134386859,
+      "learning_rate": 8.185707504535631e-05,
+      "loss": 2.9559,
+      "step": 4963
+    },
+    {
+      "epoch": 1.0565575951050812,
+      "grad_norm": 1.1605147294126423,
+      "learning_rate": 8.18475290799526e-05,
+      "loss": 2.9339,
+      "step": 4964
+    },
+    {
+      "epoch": 1.0567704176642725,
+      "grad_norm": 0.957321056953365,
+      "learning_rate": 8.183798116082575e-05,
+      "loss": 2.9693,
+      "step": 4965
+    },
+    {
+      "epoch": 1.0569832402234638,
+      "grad_norm": 1.0584791316481703,
+      "learning_rate": 8.182843128856151e-05,
+      "loss": 3.0114,
+      "step": 4966
+    },
+    {
+      "epoch": 1.0571960627826549,
+      "grad_norm": 0.9074003943763991,
+      "learning_rate": 8.181887946374567e-05,
+      "loss": 2.9551,
+      "step": 4967
+    },
+    {
+      "epoch": 1.0574088853418462,
+      "grad_norm": 0.9095666870814506,
+      "learning_rate": 8.180932568696427e-05,
+      "loss": 2.9586,
+      "step": 4968
+    },
+    {
+      "epoch": 1.0576217079010375,
+      "grad_norm": 0.954612801795177,
+      "learning_rate": 8.179976995880335e-05,
+      "loss": 2.9424,
+      "step": 4969
+    },
+    {
+      "epoch": 1.0578345304602288,
+      "grad_norm": 1.0800499067515912,
+      "learning_rate": 8.179021227984914e-05,
+      "loss": 2.9744,
+      "step": 4970
+    },
+    {
+      "epoch": 1.05804735301942,
+      "grad_norm": 0.8599396105331125,
+      "learning_rate": 8.178065265068795e-05,
+      "loss": 2.954,
+      "step": 4971
+    },
+    {
+      "epoch": 1.0582601755786114,
+      "grad_norm": 0.970485328527493,
+      "learning_rate": 8.177109107190626e-05,
+      "loss": 2.9524,
+      "step": 4972
+    },
+    {
+      "epoch": 1.0584729981378027,
+      "grad_norm": 1.014843079945405,
+      "learning_rate": 8.176152754409058e-05,
+      "loss": 2.9723,
+      "step": 4973
+    },
+    {
+      "epoch": 1.0586858206969938,
+      "grad_norm": 1.0763354167056582,
+      "learning_rate": 8.175196206782765e-05,
+      "loss": 2.9487,
+      "step": 4974
+    },
+    {
+      "epoch": 1.058898643256185,
+      "grad_norm": 1.086470377933148,
+      "learning_rate": 8.174239464370424e-05,
+      "loss": 2.9579,
+      "step": 4975
+    },
+    {
+      "epoch": 1.0591114658153764,
+      "grad_norm": 0.8621994559590002,
+      "learning_rate": 8.17328252723073e-05,
+      "loss": 2.9951,
+      "step": 4976
+    },
+    {
+      "epoch": 1.0593242883745677,
+      "grad_norm": 0.9349701599435175,
+      "learning_rate": 8.172325395422384e-05,
+      "loss": 2.9403,
+      "step": 4977
+    },
+    {
+      "epoch": 1.059537110933759,
+      "grad_norm": 0.9484770756408717,
+      "learning_rate": 8.171368069004106e-05,
+      "loss": 2.9329,
+      "step": 4978
+    },
+    {
+      "epoch": 1.0597499334929503,
+      "grad_norm": 0.8646388254212829,
+      "learning_rate": 8.170410548034623e-05,
+      "loss": 2.971,
+      "step": 4979
+    },
+    {
+      "epoch": 1.0599627560521416,
+      "grad_norm": 0.8108873622843508,
+      "learning_rate": 8.169452832572675e-05,
+      "loss": 2.9571,
+      "step": 4980
+    },
+    {
+      "epoch": 1.0601755786113327,
+      "grad_norm": 0.7795012387710463,
+      "learning_rate": 8.168494922677013e-05,
+      "loss": 2.9473,
+      "step": 4981
+    },
+    {
+      "epoch": 1.060388401170524,
+      "grad_norm": 0.8340474911388237,
+      "learning_rate": 8.167536818406403e-05,
+      "loss": 2.9764,
+      "step": 4982
+    },
+    {
+      "epoch": 1.0606012237297153,
+      "grad_norm": 0.992168878164364,
+      "learning_rate": 8.166578519819622e-05,
+      "loss": 2.9561,
+      "step": 4983
+    },
+    {
+      "epoch": 1.0608140462889066,
+      "grad_norm": 1.1104243435749395,
+      "learning_rate": 8.165620026975452e-05,
+      "loss": 2.9987,
+      "step": 4984
+    },
+    {
+      "epoch": 1.061026868848098,
+      "grad_norm": 0.9889490216391454,
+      "learning_rate": 8.164661339932699e-05,
+      "loss": 2.9197,
+      "step": 4985
+    },
+    {
+      "epoch": 1.0612396914072892,
+      "grad_norm": 1.0363461978164996,
+      "learning_rate": 8.163702458750173e-05,
+      "loss": 2.9289,
+      "step": 4986
+    },
+    {
+      "epoch": 1.0614525139664805,
+      "grad_norm": 1.0393807503067696,
+      "learning_rate": 8.162743383486698e-05,
+      "loss": 2.9779,
+      "step": 4987
+    },
+    {
+      "epoch": 1.0616653365256716,
+      "grad_norm": 1.1026241806793637,
+      "learning_rate": 8.161784114201108e-05,
+      "loss": 2.9539,
+      "step": 4988
+    },
+    {
+      "epoch": 1.061878159084863,
+      "grad_norm": 0.9300117044151984,
+      "learning_rate": 8.16082465095225e-05,
+      "loss": 2.9368,
+      "step": 4989
+    },
+    {
+      "epoch": 1.0620909816440542,
+      "grad_norm": 0.9148938192125373,
+      "learning_rate": 8.159864993798988e-05,
+      "loss": 2.9634,
+      "step": 4990
+    },
+    {
+      "epoch": 1.0623038042032455,
+      "grad_norm": 1.0127753783269746,
+      "learning_rate": 8.158905142800188e-05,
+      "loss": 2.947,
+      "step": 4991
+    },
+    {
+      "epoch": 1.0625166267624369,
+      "grad_norm": 1.008567791087387,
+      "learning_rate": 8.157945098014734e-05,
+      "loss": 2.9126,
+      "step": 4992
+    },
+    {
+      "epoch": 1.0627294493216282,
+      "grad_norm": 0.9719761708937191,
+      "learning_rate": 8.156984859501525e-05,
+      "loss": 2.9416,
+      "step": 4993
+    },
+    {
+      "epoch": 1.0629422718808195,
+      "grad_norm": 1.0518047418325256,
+      "learning_rate": 8.156024427319465e-05,
+      "loss": 2.9788,
+      "step": 4994
+    },
+    {
+      "epoch": 1.0631550944400106,
+      "grad_norm": 0.9323323022242621,
+      "learning_rate": 8.155063801527472e-05,
+      "loss": 2.964,
+      "step": 4995
+    },
+    {
+      "epoch": 1.0633679169992019,
+      "grad_norm": 0.9216386005958351,
+      "learning_rate": 8.154102982184477e-05,
+      "loss": 2.9827,
+      "step": 4996
+    },
+    {
+      "epoch": 1.0635807395583932,
+      "grad_norm": 1.0145543763993836,
+      "learning_rate": 8.153141969349425e-05,
+      "loss": 3.0014,
+      "step": 4997
+    },
+    {
+      "epoch": 1.0637935621175845,
+      "grad_norm": 1.02974342259869,
+      "learning_rate": 8.152180763081267e-05,
+      "loss": 2.9804,
+      "step": 4998
+    },
+    {
+      "epoch": 1.0640063846767758,
+      "grad_norm": 1.0295916614720773,
+      "learning_rate": 8.15121936343897e-05,
+      "loss": 2.9814,
+      "step": 4999
+    },
+    {
+      "epoch": 1.064219207235967,
+      "grad_norm": 1.077072651604443,
+      "learning_rate": 8.150257770481517e-05,
+      "loss": 2.9915,
+      "step": 5000
+    },
+    {
+      "epoch": 1.0644320297951584,
+      "grad_norm": 1.0246115726817098,
+      "learning_rate": 8.149295984267892e-05,
+      "loss": 3.0118,
+      "step": 5001
+    },
+    {
+      "epoch": 1.0646448523543495,
+      "grad_norm": 0.9657791160060734,
+      "learning_rate": 8.148334004857099e-05,
+      "loss": 2.9558,
+      "step": 5002
+    },
+    {
+      "epoch": 1.0648576749135408,
+      "grad_norm": 0.9960690073986881,
+      "learning_rate": 8.147371832308152e-05,
+      "loss": 2.9739,
+      "step": 5003
+    },
+    {
+      "epoch": 1.065070497472732,
+      "grad_norm": 1.164755015262347,
+      "learning_rate": 8.146409466680076e-05,
+      "loss": 2.9881,
+      "step": 5004
+    },
+    {
+      "epoch": 1.0652833200319234,
+      "grad_norm": 0.777209033887991,
+      "learning_rate": 8.145446908031909e-05,
+      "loss": 2.9401,
+      "step": 5005
+    },
+    {
+      "epoch": 1.0654961425911147,
+      "grad_norm": 0.8880749380645723,
+      "learning_rate": 8.1444841564227e-05,
+      "loss": 2.959,
+      "step": 5006
+    },
+    {
+      "epoch": 1.065708965150306,
+      "grad_norm": 0.9399280905913681,
+      "learning_rate": 8.143521211911511e-05,
+      "loss": 2.9433,
+      "step": 5007
+    },
+    {
+      "epoch": 1.0659217877094973,
+      "grad_norm": 0.9345720903301782,
+      "learning_rate": 8.142558074557412e-05,
+      "loss": 2.9864,
+      "step": 5008
+    },
+    {
+      "epoch": 1.0661346102686884,
+      "grad_norm": 1.060358294003073,
+      "learning_rate": 8.141594744419492e-05,
+      "loss": 2.9561,
+      "step": 5009
+    },
+    {
+      "epoch": 1.0663474328278797,
+      "grad_norm": 1.103841223540103,
+      "learning_rate": 8.140631221556845e-05,
+      "loss": 2.9633,
+      "step": 5010
+    },
+    {
+      "epoch": 1.066560255387071,
+      "grad_norm": 0.8425451718033276,
+      "learning_rate": 8.139667506028579e-05,
+      "loss": 2.9596,
+      "step": 5011
+    },
+    {
+      "epoch": 1.0667730779462623,
+      "grad_norm": 0.815971235494528,
+      "learning_rate": 8.138703597893816e-05,
+      "loss": 2.9507,
+      "step": 5012
+    },
+    {
+      "epoch": 1.0669859005054536,
+      "grad_norm": 0.9629721548660249,
+      "learning_rate": 8.137739497211687e-05,
+      "loss": 2.9081,
+      "step": 5013
+    },
+    {
+      "epoch": 1.067198723064645,
+      "grad_norm": 1.053388964229547,
+      "learning_rate": 8.136775204041335e-05,
+      "loss": 2.9904,
+      "step": 5014
+    },
+    {
+      "epoch": 1.0674115456238362,
+      "grad_norm": 1.0816778782332082,
+      "learning_rate": 8.135810718441917e-05,
+      "loss": 2.9252,
+      "step": 5015
+    },
+    {
+      "epoch": 1.0676243681830273,
+      "grad_norm": 0.8530636952840202,
+      "learning_rate": 8.134846040472599e-05,
+      "loss": 3.0025,
+      "step": 5016
+    },
+    {
+      "epoch": 1.0678371907422186,
+      "grad_norm": 0.8800419462408579,
+      "learning_rate": 8.133881170192563e-05,
+      "loss": 2.9988,
+      "step": 5017
+    },
+    {
+      "epoch": 1.06805001330141,
+      "grad_norm": 1.0210618252072299,
+      "learning_rate": 8.132916107661e-05,
+      "loss": 2.9772,
+      "step": 5018
+    },
+    {
+      "epoch": 1.0682628358606012,
+      "grad_norm": 1.083506628671271,
+      "learning_rate": 8.131950852937109e-05,
+      "loss": 2.9517,
+      "step": 5019
+    },
+    {
+      "epoch": 1.0684756584197925,
+      "grad_norm": 0.9441021962023476,
+      "learning_rate": 8.130985406080107e-05,
+      "loss": 2.9275,
+      "step": 5020
+    },
+    {
+      "epoch": 1.0686884809789838,
+      "grad_norm": 0.9857788780314626,
+      "learning_rate": 8.130019767149222e-05,
+      "loss": 2.9244,
+      "step": 5021
+    },
+    {
+      "epoch": 1.0689013035381751,
+      "grad_norm": 1.0299789845878153,
+      "learning_rate": 8.129053936203687e-05,
+      "loss": 2.9762,
+      "step": 5022
+    },
+    {
+      "epoch": 1.0691141260973662,
+      "grad_norm": 0.9221779691695783,
+      "learning_rate": 8.12808791330276e-05,
+      "loss": 2.9705,
+      "step": 5023
+    },
+    {
+      "epoch": 1.0693269486565575,
+      "grad_norm": 0.845254261510299,
+      "learning_rate": 8.127121698505697e-05,
+      "loss": 2.9245,
+      "step": 5024
+    },
+    {
+      "epoch": 1.0695397712157488,
+      "grad_norm": 0.8298079335193072,
+      "learning_rate": 8.126155291871772e-05,
+      "loss": 2.9131,
+      "step": 5025
+    },
+    {
+      "epoch": 1.0697525937749401,
+      "grad_norm": 0.8977678868861689,
+      "learning_rate": 8.125188693460273e-05,
+      "loss": 2.9379,
+      "step": 5026
+    },
+    {
+      "epoch": 1.0699654163341314,
+      "grad_norm": 0.9406703180124368,
+      "learning_rate": 8.124221903330494e-05,
+      "loss": 2.9591,
+      "step": 5027
+    },
+    {
+      "epoch": 1.0701782388933228,
+      "grad_norm": 1.0498341334358663,
+      "learning_rate": 8.123254921541746e-05,
+      "loss": 2.9593,
+      "step": 5028
+    },
+    {
+      "epoch": 1.070391061452514,
+      "grad_norm": 0.8874898250746517,
+      "learning_rate": 8.122287748153348e-05,
+      "loss": 2.9723,
+      "step": 5029
+    },
+    {
+      "epoch": 1.0706038840117051,
+      "grad_norm": 0.8443363200948446,
+      "learning_rate": 8.121320383224635e-05,
+      "loss": 2.9565,
+      "step": 5030
+    },
+    {
+      "epoch": 1.0708167065708964,
+      "grad_norm": 0.9188061179702666,
+      "learning_rate": 8.120352826814948e-05,
+      "loss": 2.9369,
+      "step": 5031
+    },
+    {
+      "epoch": 1.0710295291300878,
+      "grad_norm": 0.8881555720317561,
+      "learning_rate": 8.119385078983643e-05,
+      "loss": 3.0077,
+      "step": 5032
+    },
+    {
+      "epoch": 1.071242351689279,
+      "grad_norm": 0.8685425492846617,
+      "learning_rate": 8.11841713979009e-05,
+      "loss": 2.9626,
+      "step": 5033
+    },
+    {
+      "epoch": 1.0714551742484704,
+      "grad_norm": 0.953466797064662,
+      "learning_rate": 8.117449009293668e-05,
+      "loss": 2.9549,
+      "step": 5034
+    },
+    {
+      "epoch": 1.0716679968076617,
+      "grad_norm": 1.0975964336584902,
+      "learning_rate": 8.116480687553767e-05,
+      "loss": 2.9255,
+      "step": 5035
+    },
+    {
+      "epoch": 1.071880819366853,
+      "grad_norm": 0.8401657408711997,
+      "learning_rate": 8.115512174629789e-05,
+      "loss": 2.9656,
+      "step": 5036
+    },
+    {
+      "epoch": 1.072093641926044,
+      "grad_norm": 0.8735722665110698,
+      "learning_rate": 8.11454347058115e-05,
+      "loss": 2.9555,
+      "step": 5037
+    },
+    {
+      "epoch": 1.0723064644852354,
+      "grad_norm": 0.8910969644075982,
+      "learning_rate": 8.113574575467275e-05,
+      "loss": 2.9377,
+      "step": 5038
+    },
+    {
+      "epoch": 1.0725192870444267,
+      "grad_norm": 0.9554501911906712,
+      "learning_rate": 8.112605489347604e-05,
+      "loss": 2.92,
+      "step": 5039
+    },
+    {
+      "epoch": 1.072732109603618,
+      "grad_norm": 1.147724947394751,
+      "learning_rate": 8.111636212281586e-05,
+      "loss": 2.9775,
+      "step": 5040
+    },
+    {
+      "epoch": 1.0729449321628093,
+      "grad_norm": 0.9584720697397594,
+      "learning_rate": 8.110666744328682e-05,
+      "loss": 2.9554,
+      "step": 5041
+    },
+    {
+      "epoch": 1.0731577547220006,
+      "grad_norm": 1.0097699118552093,
+      "learning_rate": 8.109697085548362e-05,
+      "loss": 2.9916,
+      "step": 5042
+    },
+    {
+      "epoch": 1.073370577281192,
+      "grad_norm": 0.9724314342528693,
+      "learning_rate": 8.108727236000115e-05,
+      "loss": 2.9343,
+      "step": 5043
+    },
+    {
+      "epoch": 1.073583399840383,
+      "grad_norm": 1.0279772756864227,
+      "learning_rate": 8.107757195743439e-05,
+      "loss": 2.9836,
+      "step": 5044
+    },
+    {
+      "epoch": 1.0737962223995743,
+      "grad_norm": 1.0153520022611389,
+      "learning_rate": 8.106786964837836e-05,
+      "loss": 2.9726,
+      "step": 5045
+    },
+    {
+      "epoch": 1.0740090449587656,
+      "grad_norm": 1.0484838802193135,
+      "learning_rate": 8.105816543342832e-05,
+      "loss": 2.9752,
+      "step": 5046
+    },
+    {
+      "epoch": 1.074221867517957,
+      "grad_norm": 1.0526056696251656,
+      "learning_rate": 8.104845931317955e-05,
+      "loss": 2.954,
+      "step": 5047
+    },
+    {
+      "epoch": 1.0744346900771482,
+      "grad_norm": 0.988081705073553,
+      "learning_rate": 8.10387512882275e-05,
+      "loss": 3.0011,
+      "step": 5048
+    },
+    {
+      "epoch": 1.0746475126363395,
+      "grad_norm": 1.0885798903998236,
+      "learning_rate": 8.102904135916772e-05,
+      "loss": 2.963,
+      "step": 5049
+    },
+    {
+      "epoch": 1.0748603351955308,
+      "grad_norm": 0.9351494604810137,
+      "learning_rate": 8.101932952659586e-05,
+      "loss": 2.9841,
+      "step": 5050
+    },
+    {
+      "epoch": 1.075073157754722,
+      "grad_norm": 0.9140125570437961,
+      "learning_rate": 8.100961579110772e-05,
+      "loss": 2.9721,
+      "step": 5051
+    },
+    {
+      "epoch": 1.0752859803139132,
+      "grad_norm": 0.8299807216694707,
+      "learning_rate": 8.099990015329919e-05,
+      "loss": 2.9671,
+      "step": 5052
+    },
+    {
+      "epoch": 1.0754988028731045,
+      "grad_norm": 0.741037786705879,
+      "learning_rate": 8.09901826137663e-05,
+      "loss": 2.9538,
+      "step": 5053
+    },
+    {
+      "epoch": 1.0757116254322958,
+      "grad_norm": 0.8096251613474287,
+      "learning_rate": 8.098046317310518e-05,
+      "loss": 2.9539,
+      "step": 5054
+    },
+    {
+      "epoch": 1.0759244479914871,
+      "grad_norm": 0.8651839875301977,
+      "learning_rate": 8.097074183191204e-05,
+      "loss": 2.9601,
+      "step": 5055
+    },
+    {
+      "epoch": 1.0761372705506784,
+      "grad_norm": 0.8943423364899147,
+      "learning_rate": 8.09610185907833e-05,
+      "loss": 2.9523,
+      "step": 5056
+    },
+    {
+      "epoch": 1.0763500931098697,
+      "grad_norm": 0.9217303197802662,
+      "learning_rate": 8.095129345031544e-05,
+      "loss": 2.9669,
+      "step": 5057
+    },
+    {
+      "epoch": 1.076562915669061,
+      "grad_norm": 0.9877709234880371,
+      "learning_rate": 8.094156641110504e-05,
+      "loss": 2.9461,
+      "step": 5058
+    },
+    {
+      "epoch": 1.0767757382282521,
+      "grad_norm": 1.1797784836103207,
+      "learning_rate": 8.093183747374883e-05,
+      "loss": 2.958,
+      "step": 5059
+    },
+    {
+      "epoch": 1.0769885607874434,
+      "grad_norm": 0.9495183313513028,
+      "learning_rate": 8.092210663884362e-05,
+      "loss": 2.9502,
+      "step": 5060
+    },
+    {
+      "epoch": 1.0772013833466347,
+      "grad_norm": 1.0851148071766774,
+      "learning_rate": 8.091237390698639e-05,
+      "loss": 2.9333,
+      "step": 5061
+    },
+    {
+      "epoch": 1.077414205905826,
+      "grad_norm": 1.0593793037374497,
+      "learning_rate": 8.090263927877417e-05,
+      "loss": 2.9754,
+      "step": 5062
+    },
+    {
+      "epoch": 1.0776270284650173,
+      "grad_norm": 0.9685155211231746,
+      "learning_rate": 8.089290275480416e-05,
+      "loss": 2.9346,
+      "step": 5063
+    },
+    {
+      "epoch": 1.0778398510242087,
+      "grad_norm": 1.01053638196176,
+      "learning_rate": 8.08831643356737e-05,
+      "loss": 2.9191,
+      "step": 5064
+    },
+    {
+      "epoch": 1.0780526735833997,
+      "grad_norm": 1.0565359339848326,
+      "learning_rate": 8.087342402198012e-05,
+      "loss": 2.9908,
+      "step": 5065
+    },
+    {
+      "epoch": 1.078265496142591,
+      "grad_norm": 0.9430707196667134,
+      "learning_rate": 8.0863681814321e-05,
+      "loss": 3.0028,
+      "step": 5066
+    },
+    {
+      "epoch": 1.0784783187017823,
+      "grad_norm": 0.9008913018193682,
+      "learning_rate": 8.0853937713294e-05,
+      "loss": 2.9474,
+      "step": 5067
+    },
+    {
+      "epoch": 1.0786911412609737,
+      "grad_norm": 0.8169119998421475,
+      "learning_rate": 8.084419171949685e-05,
+      "loss": 2.9603,
+      "step": 5068
+    },
+    {
+      "epoch": 1.078903963820165,
+      "grad_norm": 0.7711458360465393,
+      "learning_rate": 8.083444383352744e-05,
+      "loss": 2.9065,
+      "step": 5069
+    },
+    {
+      "epoch": 1.0791167863793563,
+      "grad_norm": 0.8797999676004048,
+      "learning_rate": 8.082469405598378e-05,
+      "loss": 2.9791,
+      "step": 5070
+    },
+    {
+      "epoch": 1.0793296089385476,
+      "grad_norm": 0.9168098493025659,
+      "learning_rate": 8.081494238746397e-05,
+      "loss": 2.9509,
+      "step": 5071
+    },
+    {
+      "epoch": 1.0795424314977389,
+      "grad_norm": 0.8877118702915239,
+      "learning_rate": 8.080518882856625e-05,
+      "loss": 3.0197,
+      "step": 5072
+    },
+    {
+      "epoch": 1.07975525405693,
+      "grad_norm": 0.9284525963821535,
+      "learning_rate": 8.079543337988893e-05,
+      "loss": 2.975,
+      "step": 5073
+    },
+    {
+      "epoch": 1.0799680766161213,
+      "grad_norm": 1.0351628817540581,
+      "learning_rate": 8.07856760420305e-05,
+      "loss": 2.9634,
+      "step": 5074
+    },
+    {
+      "epoch": 1.0801808991753126,
+      "grad_norm": 1.0449763927333993,
+      "learning_rate": 8.077591681558951e-05,
+      "loss": 2.9433,
+      "step": 5075
+    },
+    {
+      "epoch": 1.0803937217345039,
+      "grad_norm": 0.9863728287990974,
+      "learning_rate": 8.076615570116468e-05,
+      "loss": 2.9541,
+      "step": 5076
+    },
+    {
+      "epoch": 1.0806065442936952,
+      "grad_norm": 0.9121959906968585,
+      "learning_rate": 8.07563926993548e-05,
+      "loss": 2.976,
+      "step": 5077
+    },
+    {
+      "epoch": 1.0808193668528865,
+      "grad_norm": 0.8907418357839556,
+      "learning_rate": 8.074662781075879e-05,
+      "loss": 2.9673,
+      "step": 5078
+    },
+    {
+      "epoch": 1.0810321894120776,
+      "grad_norm": 0.9424159945365818,
+      "learning_rate": 8.073686103597571e-05,
+      "loss": 2.9459,
+      "step": 5079
+    },
+    {
+      "epoch": 1.0812450119712689,
+      "grad_norm": 1.0425313786906825,
+      "learning_rate": 8.072709237560467e-05,
+      "loss": 2.9639,
+      "step": 5080
+    },
+    {
+      "epoch": 1.0814578345304602,
+      "grad_norm": 1.029842544090139,
+      "learning_rate": 8.071732183024498e-05,
+      "loss": 2.969,
+      "step": 5081
+    },
+    {
+      "epoch": 1.0816706570896515,
+      "grad_norm": 0.9120693770332322,
+      "learning_rate": 8.070754940049603e-05,
+      "loss": 2.9898,
+      "step": 5082
+    },
+    {
+      "epoch": 1.0818834796488428,
+      "grad_norm": 0.9875947224144006,
+      "learning_rate": 8.069777508695728e-05,
+      "loss": 2.9533,
+      "step": 5083
+    },
+    {
+      "epoch": 1.082096302208034,
+      "grad_norm": 1.2220769326906054,
+      "learning_rate": 8.06879988902284e-05,
+      "loss": 2.9463,
+      "step": 5084
+    },
+    {
+      "epoch": 1.0823091247672254,
+      "grad_norm": 0.9540050110595114,
+      "learning_rate": 8.067822081090907e-05,
+      "loss": 2.9448,
+      "step": 5085
+    },
+    {
+      "epoch": 1.0825219473264167,
+      "grad_norm": 0.9288825788413246,
+      "learning_rate": 8.066844084959914e-05,
+      "loss": 2.9061,
+      "step": 5086
+    },
+    {
+      "epoch": 1.0827347698856078,
+      "grad_norm": 0.9778107635333925,
+      "learning_rate": 8.065865900689862e-05,
+      "loss": 2.9612,
+      "step": 5087
+    },
+    {
+      "epoch": 1.082947592444799,
+      "grad_norm": 1.018448886252954,
+      "learning_rate": 8.064887528340757e-05,
+      "loss": 2.9308,
+      "step": 5088
+    },
+    {
+      "epoch": 1.0831604150039904,
+      "grad_norm": 1.100035150588228,
+      "learning_rate": 8.063908967972615e-05,
+      "loss": 2.9523,
+      "step": 5089
+    },
+    {
+      "epoch": 1.0833732375631817,
+      "grad_norm": 0.920511848986071,
+      "learning_rate": 8.06293021964547e-05,
+      "loss": 2.981,
+      "step": 5090
+    },
+    {
+      "epoch": 1.083586060122373,
+      "grad_norm": 0.8687779940188517,
+      "learning_rate": 8.061951283419363e-05,
+      "loss": 2.9644,
+      "step": 5091
+    },
+    {
+      "epoch": 1.0837988826815643,
+      "grad_norm": 0.8863145645137477,
+      "learning_rate": 8.060972159354349e-05,
+      "loss": 2.9413,
+      "step": 5092
+    },
+    {
+      "epoch": 1.0840117052407554,
+      "grad_norm": 0.9764567389616488,
+      "learning_rate": 8.059992847510495e-05,
+      "loss": 2.9576,
+      "step": 5093
+    },
+    {
+      "epoch": 1.0842245277999467,
+      "grad_norm": 0.9378306138247123,
+      "learning_rate": 8.059013347947874e-05,
+      "loss": 2.9319,
+      "step": 5094
+    },
+    {
+      "epoch": 1.084437350359138,
+      "grad_norm": 0.9556961348627985,
+      "learning_rate": 8.058033660726579e-05,
+      "loss": 2.9711,
+      "step": 5095
+    },
+    {
+      "epoch": 1.0846501729183293,
+      "grad_norm": 1.1433058748494835,
+      "learning_rate": 8.057053785906707e-05,
+      "loss": 2.9517,
+      "step": 5096
+    },
+    {
+      "epoch": 1.0848629954775206,
+      "grad_norm": 0.8440132014308314,
+      "learning_rate": 8.05607372354837e-05,
+      "loss": 2.9629,
+      "step": 5097
+    },
+    {
+      "epoch": 1.085075818036712,
+      "grad_norm": 0.8879003203334376,
+      "learning_rate": 8.05509347371169e-05,
+      "loss": 2.9399,
+      "step": 5098
+    },
+    {
+      "epoch": 1.0852886405959032,
+      "grad_norm": 1.013980462279543,
+      "learning_rate": 8.054113036456804e-05,
+      "loss": 2.9394,
+      "step": 5099
+    },
+    {
+      "epoch": 1.0855014631550945,
+      "grad_norm": 1.1352112002066885,
+      "learning_rate": 8.053132411843858e-05,
+      "loss": 2.9392,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.8560130608278299,
+      "learning_rate": 8.052151599933007e-05,
+      "loss": 2.9618,
+      "step": 5101
+    },
+    {
+      "epoch": 1.085927108273477,
+      "grad_norm": 0.8911866344152223,
+      "learning_rate": 8.051170600784423e-05,
+      "loss": 2.9573,
+      "step": 5102
+    },
+    {
+      "epoch": 1.0861399308326682,
+      "grad_norm": 1.0833663567282366,
+      "learning_rate": 8.050189414458284e-05,
+      "loss": 2.9665,
+      "step": 5103
+    },
+    {
+      "epoch": 1.0863527533918595,
+      "grad_norm": 1.0977930680909906,
+      "learning_rate": 8.049208041014783e-05,
+      "loss": 2.9748,
+      "step": 5104
+    },
+    {
+      "epoch": 1.0865655759510509,
+      "grad_norm": 0.8926528849872596,
+      "learning_rate": 8.048226480514125e-05,
+      "loss": 2.9597,
+      "step": 5105
+    },
+    {
+      "epoch": 1.0867783985102422,
+      "grad_norm": 0.9959555252770979,
+      "learning_rate": 8.047244733016522e-05,
+      "loss": 2.9472,
+      "step": 5106
+    },
+    {
+      "epoch": 1.0869912210694332,
+      "grad_norm": 1.122417080731339,
+      "learning_rate": 8.0462627985822e-05,
+      "loss": 2.9218,
+      "step": 5107
+    },
+    {
+      "epoch": 1.0872040436286246,
+      "grad_norm": 0.9869021281285547,
+      "learning_rate": 8.045280677271402e-05,
+      "loss": 2.9893,
+      "step": 5108
+    },
+    {
+      "epoch": 1.0874168661878159,
+      "grad_norm": 1.0077657322248625,
+      "learning_rate": 8.044298369144374e-05,
+      "loss": 2.9512,
+      "step": 5109
+    },
+    {
+      "epoch": 1.0876296887470072,
+      "grad_norm": 1.150543180070037,
+      "learning_rate": 8.043315874261376e-05,
+      "loss": 2.9615,
+      "step": 5110
+    },
+    {
+      "epoch": 1.0878425113061985,
+      "grad_norm": 0.8799502776094149,
+      "learning_rate": 8.04233319268268e-05,
+      "loss": 2.9595,
+      "step": 5111
+    },
+    {
+      "epoch": 1.0880553338653898,
+      "grad_norm": 0.8617199651547466,
+      "learning_rate": 8.041350324468573e-05,
+      "loss": 2.9567,
+      "step": 5112
+    },
+    {
+      "epoch": 1.088268156424581,
+      "grad_norm": 0.8897625896902239,
+      "learning_rate": 8.040367269679347e-05,
+      "loss": 2.9359,
+      "step": 5113
+    },
+    {
+      "epoch": 1.0884809789837724,
+      "grad_norm": 0.9617127733479324,
+      "learning_rate": 8.03938402837531e-05,
+      "loss": 2.9605,
+      "step": 5114
+    },
+    {
+      "epoch": 1.0886938015429635,
+      "grad_norm": 0.9957090688145036,
+      "learning_rate": 8.03840060061678e-05,
+      "loss": 2.9716,
+      "step": 5115
+    },
+    {
+      "epoch": 1.0889066241021548,
+      "grad_norm": 0.9727999884380444,
+      "learning_rate": 8.037416986464088e-05,
+      "loss": 2.9582,
+      "step": 5116
+    },
+    {
+      "epoch": 1.089119446661346,
+      "grad_norm": 0.894315818114433,
+      "learning_rate": 8.03643318597757e-05,
+      "loss": 2.9589,
+      "step": 5117
+    },
+    {
+      "epoch": 1.0893322692205374,
+      "grad_norm": 0.9216373829207031,
+      "learning_rate": 8.035449199217584e-05,
+      "loss": 2.9916,
+      "step": 5118
+    },
+    {
+      "epoch": 1.0895450917797287,
+      "grad_norm": 1.112781002308459,
+      "learning_rate": 8.034465026244491e-05,
+      "loss": 2.9581,
+      "step": 5119
+    },
+    {
+      "epoch": 1.08975791433892,
+      "grad_norm": 0.9021081586931834,
+      "learning_rate": 8.033480667118667e-05,
+      "loss": 2.9627,
+      "step": 5120
+    },
+    {
+      "epoch": 1.0899707368981113,
+      "grad_norm": 0.9549190089119266,
+      "learning_rate": 8.032496121900497e-05,
+      "loss": 2.9641,
+      "step": 5121
+    },
+    {
+      "epoch": 1.0901835594573024,
+      "grad_norm": 1.1442829026277148,
+      "learning_rate": 8.03151139065038e-05,
+      "loss": 2.9489,
+      "step": 5122
+    },
+    {
+      "epoch": 1.0903963820164937,
+      "grad_norm": 0.9699940274457934,
+      "learning_rate": 8.030526473428728e-05,
+      "loss": 2.9542,
+      "step": 5123
+    },
+    {
+      "epoch": 1.090609204575685,
+      "grad_norm": 1.0598932203194045,
+      "learning_rate": 8.029541370295957e-05,
+      "loss": 2.937,
+      "step": 5124
+    },
+    {
+      "epoch": 1.0908220271348763,
+      "grad_norm": 1.0300375618787978,
+      "learning_rate": 8.028556081312503e-05,
+      "loss": 2.9736,
+      "step": 5125
+    },
+    {
+      "epoch": 1.0910348496940676,
+      "grad_norm": 1.0381824512581543,
+      "learning_rate": 8.027570606538811e-05,
+      "loss": 2.9734,
+      "step": 5126
+    },
+    {
+      "epoch": 1.091247672253259,
+      "grad_norm": 0.9949653384676503,
+      "learning_rate": 8.026584946035331e-05,
+      "loss": 2.9367,
+      "step": 5127
+    },
+    {
+      "epoch": 1.0914604948124502,
+      "grad_norm": 0.9362417060568804,
+      "learning_rate": 8.025599099862533e-05,
+      "loss": 2.9826,
+      "step": 5128
+    },
+    {
+      "epoch": 1.0916733173716413,
+      "grad_norm": 1.0488758087533767,
+      "learning_rate": 8.024613068080895e-05,
+      "loss": 2.9786,
+      "step": 5129
+    },
+    {
+      "epoch": 1.0918861399308326,
+      "grad_norm": 1.0032654650702983,
+      "learning_rate": 8.023626850750904e-05,
+      "loss": 2.9286,
+      "step": 5130
+    },
+    {
+      "epoch": 1.092098962490024,
+      "grad_norm": 1.013490319008311,
+      "learning_rate": 8.022640447933062e-05,
+      "loss": 2.9565,
+      "step": 5131
+    },
+    {
+      "epoch": 1.0923117850492152,
+      "grad_norm": 1.0182724144481814,
+      "learning_rate": 8.021653859687881e-05,
+      "loss": 2.9781,
+      "step": 5132
+    },
+    {
+      "epoch": 1.0925246076084065,
+      "grad_norm": 0.9951413210401853,
+      "learning_rate": 8.020667086075884e-05,
+      "loss": 2.9371,
+      "step": 5133
+    },
+    {
+      "epoch": 1.0927374301675978,
+      "grad_norm": 1.167452269375743,
+      "learning_rate": 8.019680127157607e-05,
+      "loss": 2.9301,
+      "step": 5134
+    },
+    {
+      "epoch": 1.0929502527267891,
+      "grad_norm": 0.8587450327784532,
+      "learning_rate": 8.018692982993594e-05,
+      "loss": 2.9366,
+      "step": 5135
+    },
+    {
+      "epoch": 1.0931630752859802,
+      "grad_norm": 1.0183510677032874,
+      "learning_rate": 8.017705653644407e-05,
+      "loss": 2.9644,
+      "step": 5136
+    },
+    {
+      "epoch": 1.0933758978451715,
+      "grad_norm": 1.0457293526696079,
+      "learning_rate": 8.016718139170607e-05,
+      "loss": 2.9352,
+      "step": 5137
+    },
+    {
+      "epoch": 1.0935887204043628,
+      "grad_norm": 0.902746298653763,
+      "learning_rate": 8.015730439632782e-05,
+      "loss": 2.9709,
+      "step": 5138
+    },
+    {
+      "epoch": 1.0938015429635541,
+      "grad_norm": 0.9813542368042297,
+      "learning_rate": 8.014742555091519e-05,
+      "loss": 2.9401,
+      "step": 5139
+    },
+    {
+      "epoch": 1.0940143655227454,
+      "grad_norm": 0.9691902756009789,
+      "learning_rate": 8.013754485607422e-05,
+      "loss": 2.9385,
+      "step": 5140
+    },
+    {
+      "epoch": 1.0942271880819368,
+      "grad_norm": 1.048039351610427,
+      "learning_rate": 8.012766231241107e-05,
+      "loss": 2.9247,
+      "step": 5141
+    },
+    {
+      "epoch": 1.094440010641128,
+      "grad_norm": 0.9853948853457242,
+      "learning_rate": 8.011777792053195e-05,
+      "loss": 2.9833,
+      "step": 5142
+    },
+    {
+      "epoch": 1.0946528332003191,
+      "grad_norm": 1.0251501443189517,
+      "learning_rate": 8.010789168104329e-05,
+      "loss": 2.9411,
+      "step": 5143
+    },
+    {
+      "epoch": 1.0948656557595104,
+      "grad_norm": 1.0128986168486627,
+      "learning_rate": 8.009800359455153e-05,
+      "loss": 2.9871,
+      "step": 5144
+    },
+    {
+      "epoch": 1.0950784783187018,
+      "grad_norm": 0.9781780192172393,
+      "learning_rate": 8.008811366166328e-05,
+      "loss": 2.9694,
+      "step": 5145
+    },
+    {
+      "epoch": 1.095291300877893,
+      "grad_norm": 1.1248597200156902,
+      "learning_rate": 8.007822188298525e-05,
+      "loss": 2.9849,
+      "step": 5146
+    },
+    {
+      "epoch": 1.0955041234370844,
+      "grad_norm": 0.9919942116660777,
+      "learning_rate": 8.006832825912426e-05,
+      "loss": 2.9689,
+      "step": 5147
+    },
+    {
+      "epoch": 1.0957169459962757,
+      "grad_norm": 1.1459453315382555,
+      "learning_rate": 8.005843279068724e-05,
+      "loss": 2.9917,
+      "step": 5148
+    },
+    {
+      "epoch": 1.095929768555467,
+      "grad_norm": 0.917566861255955,
+      "learning_rate": 8.004853547828125e-05,
+      "loss": 2.959,
+      "step": 5149
+    },
+    {
+      "epoch": 1.096142591114658,
+      "grad_norm": 0.9368748449997002,
+      "learning_rate": 8.003863632251345e-05,
+      "loss": 2.9431,
+      "step": 5150
+    },
+    {
+      "epoch": 1.0963554136738494,
+      "grad_norm": 0.9265531301074414,
+      "learning_rate": 8.002873532399111e-05,
+      "loss": 2.9644,
+      "step": 5151
+    },
+    {
+      "epoch": 1.0965682362330407,
+      "grad_norm": 0.9554894758019343,
+      "learning_rate": 8.001883248332161e-05,
+      "loss": 2.9715,
+      "step": 5152
+    },
+    {
+      "epoch": 1.096781058792232,
+      "grad_norm": 0.9979662924870324,
+      "learning_rate": 8.000892780111249e-05,
+      "loss": 2.9599,
+      "step": 5153
+    },
+    {
+      "epoch": 1.0969938813514233,
+      "grad_norm": 1.0355718541706802,
+      "learning_rate": 7.999902127797129e-05,
+      "loss": 2.9312,
+      "step": 5154
+    },
+    {
+      "epoch": 1.0972067039106146,
+      "grad_norm": 0.9912916257745217,
+      "learning_rate": 7.998911291450583e-05,
+      "loss": 2.9617,
+      "step": 5155
+    },
+    {
+      "epoch": 1.097419526469806,
+      "grad_norm": 1.0411412762749666,
+      "learning_rate": 7.997920271132387e-05,
+      "loss": 2.88,
+      "step": 5156
+    },
+    {
+      "epoch": 1.097632349028997,
+      "grad_norm": 1.0066181652636275,
+      "learning_rate": 7.996929066903342e-05,
+      "loss": 2.9332,
+      "step": 5157
+    },
+    {
+      "epoch": 1.0978451715881883,
+      "grad_norm": 0.95988164062656,
+      "learning_rate": 7.99593767882425e-05,
+      "loss": 2.9567,
+      "step": 5158
+    },
+    {
+      "epoch": 1.0980579941473796,
+      "grad_norm": 0.8833555480296489,
+      "learning_rate": 7.994946106955931e-05,
+      "loss": 2.9263,
+      "step": 5159
+    },
+    {
+      "epoch": 1.098270816706571,
+      "grad_norm": 0.9163626929222254,
+      "learning_rate": 7.993954351359215e-05,
+      "loss": 2.9658,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0984836392657622,
+      "grad_norm": 1.0853192197499197,
+      "learning_rate": 7.99296241209494e-05,
+      "loss": 2.9307,
+      "step": 5161
+    },
+    {
+      "epoch": 1.0986964618249535,
+      "grad_norm": 0.9902675815268479,
+      "learning_rate": 7.99197028922396e-05,
+      "loss": 2.9596,
+      "step": 5162
+    },
+    {
+      "epoch": 1.0989092843841448,
+      "grad_norm": 0.8778902821497587,
+      "learning_rate": 7.990977982807138e-05,
+      "loss": 2.9383,
+      "step": 5163
+    },
+    {
+      "epoch": 1.099122106943336,
+      "grad_norm": 0.9024333331650014,
+      "learning_rate": 7.989985492905346e-05,
+      "loss": 2.9179,
+      "step": 5164
+    },
+    {
+      "epoch": 1.0993349295025272,
+      "grad_norm": 1.0276966927135438,
+      "learning_rate": 7.98899281957947e-05,
+      "loss": 2.9286,
+      "step": 5165
+    },
+    {
+      "epoch": 1.0995477520617185,
+      "grad_norm": 1.1075000049563422,
+      "learning_rate": 7.987999962890407e-05,
+      "loss": 2.9555,
+      "step": 5166
+    },
+    {
+      "epoch": 1.0997605746209098,
+      "grad_norm": 0.970080562952608,
+      "learning_rate": 7.987006922899066e-05,
+      "loss": 2.9795,
+      "step": 5167
+    },
+    {
+      "epoch": 1.0999733971801011,
+      "grad_norm": 1.0026318978138007,
+      "learning_rate": 7.986013699666364e-05,
+      "loss": 2.9633,
+      "step": 5168
+    },
+    {
+      "epoch": 1.1001862197392924,
+      "grad_norm": 0.8852420680773793,
+      "learning_rate": 7.985020293253235e-05,
+      "loss": 2.9279,
+      "step": 5169
+    },
+    {
+      "epoch": 1.1003990422984837,
+      "grad_norm": 0.9084134304354853,
+      "learning_rate": 7.984026703720617e-05,
+      "loss": 2.9403,
+      "step": 5170
+    },
+    {
+      "epoch": 1.1006118648576748,
+      "grad_norm": 1.0516250007838137,
+      "learning_rate": 7.983032931129465e-05,
+      "loss": 2.9253,
+      "step": 5171
+    },
+    {
+      "epoch": 1.1008246874168661,
+      "grad_norm": 1.0169422385883684,
+      "learning_rate": 7.982038975540742e-05,
+      "loss": 2.8969,
+      "step": 5172
+    },
+    {
+      "epoch": 1.1010375099760574,
+      "grad_norm": 0.9865602003263879,
+      "learning_rate": 7.981044837015423e-05,
+      "loss": 2.9133,
+      "step": 5173
+    },
+    {
+      "epoch": 1.1012503325352487,
+      "grad_norm": 1.0335519711342689,
+      "learning_rate": 7.980050515614497e-05,
+      "loss": 2.9649,
+      "step": 5174
+    },
+    {
+      "epoch": 1.10146315509444,
+      "grad_norm": 1.1176009900052213,
+      "learning_rate": 7.979056011398959e-05,
+      "loss": 2.9466,
+      "step": 5175
+    },
+    {
+      "epoch": 1.1016759776536313,
+      "grad_norm": 0.866620045050028,
+      "learning_rate": 7.97806132442982e-05,
+      "loss": 2.9654,
+      "step": 5176
+    },
+    {
+      "epoch": 1.1018888002128226,
+      "grad_norm": 0.7932758974212933,
+      "learning_rate": 7.977066454768098e-05,
+      "loss": 2.9706,
+      "step": 5177
+    },
+    {
+      "epoch": 1.1021016227720137,
+      "grad_norm": 0.8880939704915201,
+      "learning_rate": 7.976071402474826e-05,
+      "loss": 2.8915,
+      "step": 5178
+    },
+    {
+      "epoch": 1.102314445331205,
+      "grad_norm": 0.9037401273237636,
+      "learning_rate": 7.975076167611046e-05,
+      "loss": 2.9236,
+      "step": 5179
+    },
+    {
+      "epoch": 1.1025272678903963,
+      "grad_norm": 0.9633268888055424,
+      "learning_rate": 7.974080750237813e-05,
+      "loss": 2.958,
+      "step": 5180
+    },
+    {
+      "epoch": 1.1027400904495877,
+      "grad_norm": 1.0390407528895478,
+      "learning_rate": 7.973085150416192e-05,
+      "loss": 2.9276,
+      "step": 5181
+    },
+    {
+      "epoch": 1.102952913008779,
+      "grad_norm": 0.9082077901724372,
+      "learning_rate": 7.972089368207259e-05,
+      "loss": 2.9097,
+      "step": 5182
+    },
+    {
+      "epoch": 1.1031657355679703,
+      "grad_norm": 1.0684884792401947,
+      "learning_rate": 7.971093403672097e-05,
+      "loss": 2.9433,
+      "step": 5183
+    },
+    {
+      "epoch": 1.1033785581271616,
+      "grad_norm": 1.1038901075984775,
+      "learning_rate": 7.970097256871812e-05,
+      "loss": 2.9451,
+      "step": 5184
+    },
+    {
+      "epoch": 1.1035913806863527,
+      "grad_norm": 0.8406284475628424,
+      "learning_rate": 7.969100927867507e-05,
+      "loss": 2.9462,
+      "step": 5185
+    },
+    {
+      "epoch": 1.103804203245544,
+      "grad_norm": 0.8995623266987168,
+      "learning_rate": 7.96810441672031e-05,
+      "loss": 2.9488,
+      "step": 5186
+    },
+    {
+      "epoch": 1.1040170258047353,
+      "grad_norm": 0.8631474288561222,
+      "learning_rate": 7.967107723491347e-05,
+      "loss": 2.9405,
+      "step": 5187
+    },
+    {
+      "epoch": 1.1042298483639266,
+      "grad_norm": 0.8699688886619555,
+      "learning_rate": 7.966110848241764e-05,
+      "loss": 2.9491,
+      "step": 5188
+    },
+    {
+      "epoch": 1.1044426709231179,
+      "grad_norm": 0.8647839317770866,
+      "learning_rate": 7.965113791032715e-05,
+      "loss": 2.9431,
+      "step": 5189
+    },
+    {
+      "epoch": 1.1046554934823092,
+      "grad_norm": 0.8497154489943208,
+      "learning_rate": 7.964116551925365e-05,
+      "loss": 2.9642,
+      "step": 5190
+    },
+    {
+      "epoch": 1.1048683160415005,
+      "grad_norm": 0.8936064258026035,
+      "learning_rate": 7.963119130980892e-05,
+      "loss": 2.9738,
+      "step": 5191
+    },
+    {
+      "epoch": 1.1050811386006916,
+      "grad_norm": 0.8884083238194328,
+      "learning_rate": 7.962121528260483e-05,
+      "loss": 2.9469,
+      "step": 5192
+    },
+    {
+      "epoch": 1.1052939611598829,
+      "grad_norm": 0.8877762524034513,
+      "learning_rate": 7.961123743825337e-05,
+      "loss": 2.9092,
+      "step": 5193
+    },
+    {
+      "epoch": 1.1055067837190742,
+      "grad_norm": 0.9815662545968109,
+      "learning_rate": 7.960125777736664e-05,
+      "loss": 2.9475,
+      "step": 5194
+    },
+    {
+      "epoch": 1.1057196062782655,
+      "grad_norm": 1.0930466082093573,
+      "learning_rate": 7.959127630055687e-05,
+      "loss": 2.9853,
+      "step": 5195
+    },
+    {
+      "epoch": 1.1059324288374568,
+      "grad_norm": 1.0134686136242548,
+      "learning_rate": 7.958129300843637e-05,
+      "loss": 2.9565,
+      "step": 5196
+    },
+    {
+      "epoch": 1.106145251396648,
+      "grad_norm": 1.070610136868549,
+      "learning_rate": 7.957130790161757e-05,
+      "loss": 2.9703,
+      "step": 5197
+    },
+    {
+      "epoch": 1.1063580739558394,
+      "grad_norm": 1.0396501287107758,
+      "learning_rate": 7.956132098071304e-05,
+      "loss": 2.9485,
+      "step": 5198
+    },
+    {
+      "epoch": 1.1065708965150307,
+      "grad_norm": 0.9159495907693329,
+      "learning_rate": 7.955133224633544e-05,
+      "loss": 2.9422,
+      "step": 5199
+    },
+    {
+      "epoch": 1.1067837190742218,
+      "grad_norm": 0.8786289095484325,
+      "learning_rate": 7.954134169909753e-05,
+      "loss": 2.9419,
+      "step": 5200
+    },
+    {
+      "epoch": 1.106996541633413,
+      "grad_norm": 0.8416017626247279,
+      "learning_rate": 7.953134933961218e-05,
+      "loss": 2.976,
+      "step": 5201
+    },
+    {
+      "epoch": 1.1072093641926044,
+      "grad_norm": 0.9086530959852277,
+      "learning_rate": 7.95213551684924e-05,
+      "loss": 2.9027,
+      "step": 5202
+    },
+    {
+      "epoch": 1.1074221867517957,
+      "grad_norm": 0.9715733463115068,
+      "learning_rate": 7.951135918635127e-05,
+      "loss": 2.9427,
+      "step": 5203
+    },
+    {
+      "epoch": 1.107635009310987,
+      "grad_norm": 0.8862915404620684,
+      "learning_rate": 7.950136139380203e-05,
+      "loss": 2.9469,
+      "step": 5204
+    },
+    {
+      "epoch": 1.1078478318701783,
+      "grad_norm": 0.8793060065808916,
+      "learning_rate": 7.9491361791458e-05,
+      "loss": 2.9645,
+      "step": 5205
+    },
+    {
+      "epoch": 1.1080606544293694,
+      "grad_norm": 0.9247335989348299,
+      "learning_rate": 7.948136037993261e-05,
+      "loss": 2.9725,
+      "step": 5206
+    },
+    {
+      "epoch": 1.1082734769885607,
+      "grad_norm": 0.9510999944250276,
+      "learning_rate": 7.947135715983941e-05,
+      "loss": 2.9492,
+      "step": 5207
+    },
+    {
+      "epoch": 1.108486299547752,
+      "grad_norm": 0.9953243298343221,
+      "learning_rate": 7.946135213179207e-05,
+      "loss": 2.9568,
+      "step": 5208
+    },
+    {
+      "epoch": 1.1086991221069433,
+      "grad_norm": 0.9029203081000277,
+      "learning_rate": 7.945134529640434e-05,
+      "loss": 2.9552,
+      "step": 5209
+    },
+    {
+      "epoch": 1.1089119446661346,
+      "grad_norm": 0.8590185491766194,
+      "learning_rate": 7.944133665429014e-05,
+      "loss": 2.9671,
+      "step": 5210
+    },
+    {
+      "epoch": 1.109124767225326,
+      "grad_norm": 1.0169375331817365,
+      "learning_rate": 7.943132620606341e-05,
+      "loss": 2.9695,
+      "step": 5211
+    },
+    {
+      "epoch": 1.1093375897845172,
+      "grad_norm": 1.1047330203630374,
+      "learning_rate": 7.942131395233829e-05,
+      "loss": 2.9779,
+      "step": 5212
+    },
+    {
+      "epoch": 1.1095504123437085,
+      "grad_norm": 0.9649032913823316,
+      "learning_rate": 7.941129989372895e-05,
+      "loss": 2.9699,
+      "step": 5213
+    },
+    {
+      "epoch": 1.1097632349028996,
+      "grad_norm": 1.0400668791952126,
+      "learning_rate": 7.940128403084976e-05,
+      "loss": 2.9574,
+      "step": 5214
+    },
+    {
+      "epoch": 1.109976057462091,
+      "grad_norm": 0.9760119062687047,
+      "learning_rate": 7.939126636431515e-05,
+      "loss": 2.9771,
+      "step": 5215
+    },
+    {
+      "epoch": 1.1101888800212822,
+      "grad_norm": 1.0265062631292041,
+      "learning_rate": 7.938124689473966e-05,
+      "loss": 2.9876,
+      "step": 5216
+    },
+    {
+      "epoch": 1.1104017025804735,
+      "grad_norm": 1.1263222323934927,
+      "learning_rate": 7.937122562273792e-05,
+      "loss": 2.9293,
+      "step": 5217
+    },
+    {
+      "epoch": 1.1106145251396649,
+      "grad_norm": 1.0225931383577627,
+      "learning_rate": 7.936120254892472e-05,
+      "loss": 2.9184,
+      "step": 5218
+    },
+    {
+      "epoch": 1.1108273476988562,
+      "grad_norm": 1.0632989625345626,
+      "learning_rate": 7.935117767391493e-05,
+      "loss": 2.9186,
+      "step": 5219
+    },
+    {
+      "epoch": 1.1110401702580472,
+      "grad_norm": 0.9524271614509665,
+      "learning_rate": 7.934115099832356e-05,
+      "loss": 2.9292,
+      "step": 5220
+    },
+    {
+      "epoch": 1.1112529928172385,
+      "grad_norm": 1.0518406514003633,
+      "learning_rate": 7.933112252276566e-05,
+      "loss": 2.9495,
+      "step": 5221
+    },
+    {
+      "epoch": 1.1114658153764299,
+      "grad_norm": 1.0034993367300444,
+      "learning_rate": 7.932109224785647e-05,
+      "loss": 2.9455,
+      "step": 5222
+    },
+    {
+      "epoch": 1.1116786379356212,
+      "grad_norm": 0.9961571599557254,
+      "learning_rate": 7.931106017421132e-05,
+      "loss": 2.9473,
+      "step": 5223
+    },
+    {
+      "epoch": 1.1118914604948125,
+      "grad_norm": 1.023161217565719,
+      "learning_rate": 7.930102630244562e-05,
+      "loss": 2.951,
+      "step": 5224
+    },
+    {
+      "epoch": 1.1121042830540038,
+      "grad_norm": 0.9730441709554324,
+      "learning_rate": 7.92909906331749e-05,
+      "loss": 2.9213,
+      "step": 5225
+    },
+    {
+      "epoch": 1.112317105613195,
+      "grad_norm": 0.910842274893541,
+      "learning_rate": 7.928095316701484e-05,
+      "loss": 2.9312,
+      "step": 5226
+    },
+    {
+      "epoch": 1.1125299281723864,
+      "grad_norm": 0.8541811364301843,
+      "learning_rate": 7.927091390458117e-05,
+      "loss": 2.9752,
+      "step": 5227
+    },
+    {
+      "epoch": 1.1127427507315775,
+      "grad_norm": 0.9667116845416945,
+      "learning_rate": 7.926087284648977e-05,
+      "loss": 2.9364,
+      "step": 5228
+    },
+    {
+      "epoch": 1.1129555732907688,
+      "grad_norm": 1.2116933925880184,
+      "learning_rate": 7.925082999335665e-05,
+      "loss": 2.9435,
+      "step": 5229
+    },
+    {
+      "epoch": 1.11316839584996,
+      "grad_norm": 0.7899314378219914,
+      "learning_rate": 7.924078534579784e-05,
+      "loss": 2.9304,
+      "step": 5230
+    },
+    {
+      "epoch": 1.1133812184091514,
+      "grad_norm": 0.8041904177078424,
+      "learning_rate": 7.923073890442958e-05,
+      "loss": 2.9204,
+      "step": 5231
+    },
+    {
+      "epoch": 1.1135940409683427,
+      "grad_norm": 0.927875893868934,
+      "learning_rate": 7.92206906698682e-05,
+      "loss": 2.9449,
+      "step": 5232
+    },
+    {
+      "epoch": 1.113806863527534,
+      "grad_norm": 0.9199853887641224,
+      "learning_rate": 7.921064064273008e-05,
+      "loss": 2.9402,
+      "step": 5233
+    },
+    {
+      "epoch": 1.114019686086725,
+      "grad_norm": 0.8868798792378826,
+      "learning_rate": 7.920058882363175e-05,
+      "loss": 2.9259,
+      "step": 5234
+    },
+    {
+      "epoch": 1.1142325086459164,
+      "grad_norm": 0.9715252311851881,
+      "learning_rate": 7.919053521318988e-05,
+      "loss": 2.9366,
+      "step": 5235
+    },
+    {
+      "epoch": 1.1144453312051077,
+      "grad_norm": 1.0071569276395276,
+      "learning_rate": 7.91804798120212e-05,
+      "loss": 2.9756,
+      "step": 5236
+    },
+    {
+      "epoch": 1.114658153764299,
+      "grad_norm": 0.9523872861017133,
+      "learning_rate": 7.917042262074256e-05,
+      "loss": 2.9307,
+      "step": 5237
+    },
+    {
+      "epoch": 1.1148709763234903,
+      "grad_norm": 1.0895553828403004,
+      "learning_rate": 7.916036363997097e-05,
+      "loss": 2.9352,
+      "step": 5238
+    },
+    {
+      "epoch": 1.1150837988826816,
+      "grad_norm": 0.9717981104623846,
+      "learning_rate": 7.915030287032348e-05,
+      "loss": 2.9311,
+      "step": 5239
+    },
+    {
+      "epoch": 1.115296621441873,
+      "grad_norm": 0.894474534652692,
+      "learning_rate": 7.91402403124173e-05,
+      "loss": 2.9641,
+      "step": 5240
+    },
+    {
+      "epoch": 1.1155094440010642,
+      "grad_norm": 1.063037703117446,
+      "learning_rate": 7.913017596686968e-05,
+      "loss": 2.9306,
+      "step": 5241
+    },
+    {
+      "epoch": 1.1157222665602553,
+      "grad_norm": 0.9149442023577297,
+      "learning_rate": 7.91201098342981e-05,
+      "loss": 2.9085,
+      "step": 5242
+    },
+    {
+      "epoch": 1.1159350891194466,
+      "grad_norm": 0.9404171962196679,
+      "learning_rate": 7.911004191532e-05,
+      "loss": 2.9719,
+      "step": 5243
+    },
+    {
+      "epoch": 1.116147911678638,
+      "grad_norm": 1.0065142942909455,
+      "learning_rate": 7.909997221055308e-05,
+      "loss": 2.9114,
+      "step": 5244
+    },
+    {
+      "epoch": 1.1163607342378292,
+      "grad_norm": 1.0136263901043978,
+      "learning_rate": 7.908990072061504e-05,
+      "loss": 2.9013,
+      "step": 5245
+    },
+    {
+      "epoch": 1.1165735567970205,
+      "grad_norm": 1.1129676388362753,
+      "learning_rate": 7.907982744612372e-05,
+      "loss": 2.9613,
+      "step": 5246
+    },
+    {
+      "epoch": 1.1167863793562118,
+      "grad_norm": 0.9186764676830286,
+      "learning_rate": 7.906975238769709e-05,
+      "loss": 2.9506,
+      "step": 5247
+    },
+    {
+      "epoch": 1.116999201915403,
+      "grad_norm": 0.9108130727411384,
+      "learning_rate": 7.905967554595323e-05,
+      "loss": 2.9409,
+      "step": 5248
+    },
+    {
+      "epoch": 1.1172120244745942,
+      "grad_norm": 0.9155699340206455,
+      "learning_rate": 7.90495969215103e-05,
+      "loss": 2.8862,
+      "step": 5249
+    },
+    {
+      "epoch": 1.1174248470337855,
+      "grad_norm": 0.8842028652287625,
+      "learning_rate": 7.903951651498658e-05,
+      "loss": 2.9161,
+      "step": 5250
+    },
+    {
+      "epoch": 1.1176376695929768,
+      "grad_norm": 0.9509633154081567,
+      "learning_rate": 7.902943432700049e-05,
+      "loss": 2.9561,
+      "step": 5251
+    },
+    {
+      "epoch": 1.1178504921521681,
+      "grad_norm": 0.9350707602049457,
+      "learning_rate": 7.901935035817049e-05,
+      "loss": 2.8884,
+      "step": 5252
+    },
+    {
+      "epoch": 1.1180633147113594,
+      "grad_norm": 1.0800480832333585,
+      "learning_rate": 7.900926460911523e-05,
+      "loss": 2.9532,
+      "step": 5253
+    },
+    {
+      "epoch": 1.1182761372705508,
+      "grad_norm": 0.9551300724466183,
+      "learning_rate": 7.899917708045342e-05,
+      "loss": 2.9547,
+      "step": 5254
+    },
+    {
+      "epoch": 1.118488959829742,
+      "grad_norm": 0.867762137387985,
+      "learning_rate": 7.898908777280388e-05,
+      "loss": 2.9051,
+      "step": 5255
+    },
+    {
+      "epoch": 1.1187017823889331,
+      "grad_norm": 0.8516157330600572,
+      "learning_rate": 7.897899668678557e-05,
+      "loss": 2.9373,
+      "step": 5256
+    },
+    {
+      "epoch": 1.1189146049481244,
+      "grad_norm": 0.9101904968091641,
+      "learning_rate": 7.896890382301752e-05,
+      "loss": 2.9632,
+      "step": 5257
+    },
+    {
+      "epoch": 1.1191274275073158,
+      "grad_norm": 0.8722376687210492,
+      "learning_rate": 7.895880918211889e-05,
+      "loss": 2.9484,
+      "step": 5258
+    },
+    {
+      "epoch": 1.119340250066507,
+      "grad_norm": 0.8215182194480583,
+      "learning_rate": 7.894871276470897e-05,
+      "loss": 2.935,
+      "step": 5259
+    },
+    {
+      "epoch": 1.1195530726256984,
+      "grad_norm": 0.8628282151683435,
+      "learning_rate": 7.89386145714071e-05,
+      "loss": 2.9382,
+      "step": 5260
+    },
+    {
+      "epoch": 1.1197658951848897,
+      "grad_norm": 0.9316820984080129,
+      "learning_rate": 7.892851460283281e-05,
+      "loss": 2.9344,
+      "step": 5261
+    },
+    {
+      "epoch": 1.119978717744081,
+      "grad_norm": 0.9724347086957827,
+      "learning_rate": 7.891841285960567e-05,
+      "loss": 2.9361,
+      "step": 5262
+    },
+    {
+      "epoch": 1.120191540303272,
+      "grad_norm": 1.0377750946398772,
+      "learning_rate": 7.890830934234537e-05,
+      "loss": 2.9707,
+      "step": 5263
+    },
+    {
+      "epoch": 1.1204043628624634,
+      "grad_norm": 1.0639048410292185,
+      "learning_rate": 7.889820405167172e-05,
+      "loss": 2.9556,
+      "step": 5264
+    },
+    {
+      "epoch": 1.1206171854216547,
+      "grad_norm": 0.917819270837025,
+      "learning_rate": 7.888809698820468e-05,
+      "loss": 2.947,
+      "step": 5265
+    },
+    {
+      "epoch": 1.120830007980846,
+      "grad_norm": 0.8170085855304986,
+      "learning_rate": 7.887798815256425e-05,
+      "loss": 2.9408,
+      "step": 5266
+    },
+    {
+      "epoch": 1.1210428305400373,
+      "grad_norm": 0.8750386513185343,
+      "learning_rate": 7.886787754537057e-05,
+      "loss": 2.9657,
+      "step": 5267
+    },
+    {
+      "epoch": 1.1212556530992286,
+      "grad_norm": 1.0398276584201045,
+      "learning_rate": 7.885776516724388e-05,
+      "loss": 2.966,
+      "step": 5268
+    },
+    {
+      "epoch": 1.12146847565842,
+      "grad_norm": 1.195511157037487,
+      "learning_rate": 7.884765101880455e-05,
+      "loss": 2.9499,
+      "step": 5269
+    },
+    {
+      "epoch": 1.121681298217611,
+      "grad_norm": 0.8472851396738923,
+      "learning_rate": 7.883753510067306e-05,
+      "loss": 2.9523,
+      "step": 5270
+    },
+    {
+      "epoch": 1.1218941207768023,
+      "grad_norm": 0.9680745373142263,
+      "learning_rate": 7.882741741346994e-05,
+      "loss": 2.9468,
+      "step": 5271
+    },
+    {
+      "epoch": 1.1221069433359936,
+      "grad_norm": 1.049542292169334,
+      "learning_rate": 7.881729795781589e-05,
+      "loss": 2.9384,
+      "step": 5272
+    },
+    {
+      "epoch": 1.122319765895185,
+      "grad_norm": 0.8960861442831353,
+      "learning_rate": 7.880717673433172e-05,
+      "loss": 2.9606,
+      "step": 5273
+    },
+    {
+      "epoch": 1.1225325884543762,
+      "grad_norm": 0.9103148427581076,
+      "learning_rate": 7.879705374363831e-05,
+      "loss": 2.9684,
+      "step": 5274
+    },
+    {
+      "epoch": 1.1227454110135675,
+      "grad_norm": 0.9770079138420172,
+      "learning_rate": 7.878692898635665e-05,
+      "loss": 2.8978,
+      "step": 5275
+    },
+    {
+      "epoch": 1.1229582335727588,
+      "grad_norm": 1.0747231887640851,
+      "learning_rate": 7.877680246310788e-05,
+      "loss": 2.9662,
+      "step": 5276
+    },
+    {
+      "epoch": 1.12317105613195,
+      "grad_norm": 1.0297419443195643,
+      "learning_rate": 7.876667417451321e-05,
+      "loss": 2.9078,
+      "step": 5277
+    },
+    {
+      "epoch": 1.1233838786911412,
+      "grad_norm": 1.0820104866357718,
+      "learning_rate": 7.875654412119397e-05,
+      "loss": 2.9663,
+      "step": 5278
+    },
+    {
+      "epoch": 1.1235967012503325,
+      "grad_norm": 1.0027954177587477,
+      "learning_rate": 7.874641230377162e-05,
+      "loss": 2.9173,
+      "step": 5279
+    },
+    {
+      "epoch": 1.1238095238095238,
+      "grad_norm": 1.013307936991889,
+      "learning_rate": 7.87362787228677e-05,
+      "loss": 2.9043,
+      "step": 5280
+    },
+    {
+      "epoch": 1.1240223463687151,
+      "grad_norm": 0.9075333365856737,
+      "learning_rate": 7.872614337910384e-05,
+      "loss": 2.9479,
+      "step": 5281
+    },
+    {
+      "epoch": 1.1242351689279064,
+      "grad_norm": 0.948078885672334,
+      "learning_rate": 7.871600627310184e-05,
+      "loss": 2.9205,
+      "step": 5282
+    },
+    {
+      "epoch": 1.1244479914870977,
+      "grad_norm": 0.988584170833659,
+      "learning_rate": 7.870586740548355e-05,
+      "loss": 2.9582,
+      "step": 5283
+    },
+    {
+      "epoch": 1.1246608140462888,
+      "grad_norm": 1.006939867715615,
+      "learning_rate": 7.869572677687096e-05,
+      "loss": 2.9375,
+      "step": 5284
+    },
+    {
+      "epoch": 1.1248736366054801,
+      "grad_norm": 1.0522783309699342,
+      "learning_rate": 7.868558438788616e-05,
+      "loss": 2.9317,
+      "step": 5285
+    },
+    {
+      "epoch": 1.1250864591646714,
+      "grad_norm": 0.9086312349453509,
+      "learning_rate": 7.867544023915134e-05,
+      "loss": 2.908,
+      "step": 5286
+    },
+    {
+      "epoch": 1.1252992817238627,
+      "grad_norm": 0.9510208117496745,
+      "learning_rate": 7.86652943312888e-05,
+      "loss": 2.9621,
+      "step": 5287
+    },
+    {
+      "epoch": 1.125512104283054,
+      "grad_norm": 0.8939655118046865,
+      "learning_rate": 7.865514666492096e-05,
+      "loss": 2.9199,
+      "step": 5288
+    },
+    {
+      "epoch": 1.1257249268422453,
+      "grad_norm": 0.923421692293783,
+      "learning_rate": 7.864499724067033e-05,
+      "loss": 2.9652,
+      "step": 5289
+    },
+    {
+      "epoch": 1.1259377494014364,
+      "grad_norm": 0.9453657382234809,
+      "learning_rate": 7.863484605915956e-05,
+      "loss": 2.9381,
+      "step": 5290
+    },
+    {
+      "epoch": 1.1261505719606277,
+      "grad_norm": 0.9086529575943143,
+      "learning_rate": 7.862469312101138e-05,
+      "loss": 2.9581,
+      "step": 5291
+    },
+    {
+      "epoch": 1.126363394519819,
+      "grad_norm": 0.9711712420410823,
+      "learning_rate": 7.861453842684861e-05,
+      "loss": 2.9173,
+      "step": 5292
+    },
+    {
+      "epoch": 1.1265762170790103,
+      "grad_norm": 0.9962305588467758,
+      "learning_rate": 7.860438197729423e-05,
+      "loss": 2.9052,
+      "step": 5293
+    },
+    {
+      "epoch": 1.1267890396382017,
+      "grad_norm": 1.1038101183240425,
+      "learning_rate": 7.859422377297129e-05,
+      "loss": 2.9461,
+      "step": 5294
+    },
+    {
+      "epoch": 1.127001862197393,
+      "grad_norm": 0.9985218820171073,
+      "learning_rate": 7.858406381450294e-05,
+      "loss": 2.9291,
+      "step": 5295
+    },
+    {
+      "epoch": 1.1272146847565843,
+      "grad_norm": 1.0515658620565924,
+      "learning_rate": 7.857390210251248e-05,
+      "loss": 2.9185,
+      "step": 5296
+    },
+    {
+      "epoch": 1.1274275073157756,
+      "grad_norm": 0.988860118981504,
+      "learning_rate": 7.856373863762327e-05,
+      "loss": 2.9684,
+      "step": 5297
+    },
+    {
+      "epoch": 1.1276403298749667,
+      "grad_norm": 0.9601864993332495,
+      "learning_rate": 7.855357342045881e-05,
+      "loss": 2.9203,
+      "step": 5298
+    },
+    {
+      "epoch": 1.127853152434158,
+      "grad_norm": 0.9530860859031623,
+      "learning_rate": 7.85434064516427e-05,
+      "loss": 2.9257,
+      "step": 5299
+    },
+    {
+      "epoch": 1.1280659749933493,
+      "grad_norm": 1.0666774784491901,
+      "learning_rate": 7.853323773179863e-05,
+      "loss": 2.9577,
+      "step": 5300
+    },
+    {
+      "epoch": 1.1282787975525406,
+      "grad_norm": 1.0130299452581542,
+      "learning_rate": 7.852306726155045e-05,
+      "loss": 2.9614,
+      "step": 5301
+    },
+    {
+      "epoch": 1.1284916201117319,
+      "grad_norm": 1.1027330052871605,
+      "learning_rate": 7.851289504152201e-05,
+      "loss": 2.9771,
+      "step": 5302
+    },
+    {
+      "epoch": 1.1287044426709232,
+      "grad_norm": 0.8667412718794816,
+      "learning_rate": 7.850272107233742e-05,
+      "loss": 2.9405,
+      "step": 5303
+    },
+    {
+      "epoch": 1.1289172652301145,
+      "grad_norm": 0.8712101046853714,
+      "learning_rate": 7.849254535462075e-05,
+      "loss": 2.9076,
+      "step": 5304
+    },
+    {
+      "epoch": 1.1291300877893056,
+      "grad_norm": 0.9072922927076685,
+      "learning_rate": 7.848236788899626e-05,
+      "loss": 2.9683,
+      "step": 5305
+    },
+    {
+      "epoch": 1.1293429103484969,
+      "grad_norm": 0.9058960823688453,
+      "learning_rate": 7.84721886760883e-05,
+      "loss": 2.8951,
+      "step": 5306
+    },
+    {
+      "epoch": 1.1295557329076882,
+      "grad_norm": 0.9567679643638671,
+      "learning_rate": 7.846200771652135e-05,
+      "loss": 2.932,
+      "step": 5307
+    },
+    {
+      "epoch": 1.1297685554668795,
+      "grad_norm": 1.0133750956110854,
+      "learning_rate": 7.845182501091992e-05,
+      "loss": 2.8941,
+      "step": 5308
+    },
+    {
+      "epoch": 1.1299813780260708,
+      "grad_norm": 0.9295327349354797,
+      "learning_rate": 7.844164055990872e-05,
+      "loss": 2.9497,
+      "step": 5309
+    },
+    {
+      "epoch": 1.130194200585262,
+      "grad_norm": 1.031460730624097,
+      "learning_rate": 7.843145436411252e-05,
+      "loss": 2.9294,
+      "step": 5310
+    },
+    {
+      "epoch": 1.1304070231444534,
+      "grad_norm": 0.9872122057287973,
+      "learning_rate": 7.842126642415619e-05,
+      "loss": 2.9037,
+      "step": 5311
+    },
+    {
+      "epoch": 1.1306198457036447,
+      "grad_norm": 0.999314486930903,
+      "learning_rate": 7.841107674066473e-05,
+      "loss": 2.9197,
+      "step": 5312
+    },
+    {
+      "epoch": 1.1308326682628358,
+      "grad_norm": 0.9639067383561535,
+      "learning_rate": 7.840088531426325e-05,
+      "loss": 2.9467,
+      "step": 5313
+    },
+    {
+      "epoch": 1.131045490822027,
+      "grad_norm": 0.9381558233097387,
+      "learning_rate": 7.839069214557694e-05,
+      "loss": 2.9084,
+      "step": 5314
+    },
+    {
+      "epoch": 1.1312583133812184,
+      "grad_norm": 0.9630359991692603,
+      "learning_rate": 7.83804972352311e-05,
+      "loss": 2.9143,
+      "step": 5315
+    },
+    {
+      "epoch": 1.1314711359404097,
+      "grad_norm": 0.9245883466886297,
+      "learning_rate": 7.837030058385118e-05,
+      "loss": 2.9291,
+      "step": 5316
+    },
+    {
+      "epoch": 1.131683958499601,
+      "grad_norm": 0.9327038974771509,
+      "learning_rate": 7.836010219206269e-05,
+      "loss": 2.9798,
+      "step": 5317
+    },
+    {
+      "epoch": 1.1318967810587923,
+      "grad_norm": 1.0102838547123565,
+      "learning_rate": 7.834990206049124e-05,
+      "loss": 2.9004,
+      "step": 5318
+    },
+    {
+      "epoch": 1.1321096036179834,
+      "grad_norm": 0.8997939028238965,
+      "learning_rate": 7.83397001897626e-05,
+      "loss": 2.9541,
+      "step": 5319
+    },
+    {
+      "epoch": 1.1323224261771747,
+      "grad_norm": 0.7968441363268104,
+      "learning_rate": 7.832949658050262e-05,
+      "loss": 2.9259,
+      "step": 5320
+    },
+    {
+      "epoch": 1.132535248736366,
+      "grad_norm": 0.8999306435649913,
+      "learning_rate": 7.831929123333722e-05,
+      "loss": 2.9169,
+      "step": 5321
+    },
+    {
+      "epoch": 1.1327480712955573,
+      "grad_norm": 0.8881841024283725,
+      "learning_rate": 7.830908414889247e-05,
+      "loss": 2.9425,
+      "step": 5322
+    },
+    {
+      "epoch": 1.1329608938547486,
+      "grad_norm": 0.8412477806817138,
+      "learning_rate": 7.829887532779454e-05,
+      "loss": 2.9557,
+      "step": 5323
+    },
+    {
+      "epoch": 1.13317371641394,
+      "grad_norm": 0.9276955836538725,
+      "learning_rate": 7.828866477066971e-05,
+      "loss": 2.9317,
+      "step": 5324
+    },
+    {
+      "epoch": 1.1333865389731312,
+      "grad_norm": 0.9578316926976082,
+      "learning_rate": 7.827845247814433e-05,
+      "loss": 2.9476,
+      "step": 5325
+    },
+    {
+      "epoch": 1.1335993615323225,
+      "grad_norm": 0.8874904603412952,
+      "learning_rate": 7.826823845084493e-05,
+      "loss": 2.936,
+      "step": 5326
+    },
+    {
+      "epoch": 1.1338121840915136,
+      "grad_norm": 0.9400364045688964,
+      "learning_rate": 7.825802268939808e-05,
+      "loss": 2.882,
+      "step": 5327
+    },
+    {
+      "epoch": 1.134025006650705,
+      "grad_norm": 0.9580973409178237,
+      "learning_rate": 7.824780519443047e-05,
+      "loss": 2.9044,
+      "step": 5328
+    },
+    {
+      "epoch": 1.1342378292098962,
+      "grad_norm": 0.9647074625527345,
+      "learning_rate": 7.823758596656889e-05,
+      "loss": 2.9196,
+      "step": 5329
+    },
+    {
+      "epoch": 1.1344506517690875,
+      "grad_norm": 1.0340244182545333,
+      "learning_rate": 7.822736500644028e-05,
+      "loss": 2.9181,
+      "step": 5330
+    },
+    {
+      "epoch": 1.1346634743282789,
+      "grad_norm": 1.1605692001868657,
+      "learning_rate": 7.821714231467166e-05,
+      "loss": 2.9463,
+      "step": 5331
+    },
+    {
+      "epoch": 1.1348762968874702,
+      "grad_norm": 0.927126576861052,
+      "learning_rate": 7.820691789189012e-05,
+      "loss": 2.9439,
+      "step": 5332
+    },
+    {
+      "epoch": 1.1350891194466612,
+      "grad_norm": 0.9669521439041519,
+      "learning_rate": 7.819669173872291e-05,
+      "loss": 2.9248,
+      "step": 5333
+    },
+    {
+      "epoch": 1.1353019420058525,
+      "grad_norm": 0.9697361602015131,
+      "learning_rate": 7.818646385579736e-05,
+      "loss": 2.9476,
+      "step": 5334
+    },
+    {
+      "epoch": 1.1355147645650439,
+      "grad_norm": 1.0537967328568187,
+      "learning_rate": 7.817623424374089e-05,
+      "loss": 2.9799,
+      "step": 5335
+    },
+    {
+      "epoch": 1.1357275871242352,
+      "grad_norm": 0.9346828957311352,
+      "learning_rate": 7.81660029031811e-05,
+      "loss": 2.9007,
+      "step": 5336
+    },
+    {
+      "epoch": 1.1359404096834265,
+      "grad_norm": 0.8444354200589793,
+      "learning_rate": 7.815576983474562e-05,
+      "loss": 2.9082,
+      "step": 5337
+    },
+    {
+      "epoch": 1.1361532322426178,
+      "grad_norm": 0.8859746562148371,
+      "learning_rate": 7.814553503906217e-05,
+      "loss": 2.9489,
+      "step": 5338
+    },
+    {
+      "epoch": 1.136366054801809,
+      "grad_norm": 0.9262001436429197,
+      "learning_rate": 7.813529851675867e-05,
+      "loss": 2.9386,
+      "step": 5339
+    },
+    {
+      "epoch": 1.1365788773610004,
+      "grad_norm": 0.823328183848598,
+      "learning_rate": 7.812506026846307e-05,
+      "loss": 2.9428,
+      "step": 5340
+    },
+    {
+      "epoch": 1.1367916999201915,
+      "grad_norm": 0.8343643693565876,
+      "learning_rate": 7.811482029480344e-05,
+      "loss": 2.9381,
+      "step": 5341
+    },
+    {
+      "epoch": 1.1370045224793828,
+      "grad_norm": 0.9425841239872044,
+      "learning_rate": 7.810457859640798e-05,
+      "loss": 2.9598,
+      "step": 5342
+    },
+    {
+      "epoch": 1.137217345038574,
+      "grad_norm": 0.889760988002052,
+      "learning_rate": 7.809433517390495e-05,
+      "loss": 2.9028,
+      "step": 5343
+    },
+    {
+      "epoch": 1.1374301675977654,
+      "grad_norm": 0.9045900046572457,
+      "learning_rate": 7.808409002792277e-05,
+      "loss": 2.9188,
+      "step": 5344
+    },
+    {
+      "epoch": 1.1376429901569567,
+      "grad_norm": 1.0937575123135768,
+      "learning_rate": 7.807384315908993e-05,
+      "loss": 2.9143,
+      "step": 5345
+    },
+    {
+      "epoch": 1.137855812716148,
+      "grad_norm": 0.9701887081998151,
+      "learning_rate": 7.806359456803505e-05,
+      "loss": 2.9304,
+      "step": 5346
+    },
+    {
+      "epoch": 1.138068635275339,
+      "grad_norm": 1.0195365224609636,
+      "learning_rate": 7.805334425538682e-05,
+      "loss": 2.9202,
+      "step": 5347
+    },
+    {
+      "epoch": 1.1382814578345304,
+      "grad_norm": 0.937195093715146,
+      "learning_rate": 7.804309222177407e-05,
+      "loss": 2.9078,
+      "step": 5348
+    },
+    {
+      "epoch": 1.1384942803937217,
+      "grad_norm": 0.9370668047478216,
+      "learning_rate": 7.80328384678257e-05,
+      "loss": 2.9723,
+      "step": 5349
+    },
+    {
+      "epoch": 1.138707102952913,
+      "grad_norm": 0.9991536636078774,
+      "learning_rate": 7.802258299417078e-05,
+      "loss": 2.9771,
+      "step": 5350
+    },
+    {
+      "epoch": 1.1389199255121043,
+      "grad_norm": 1.1472121640024884,
+      "learning_rate": 7.801232580143842e-05,
+      "loss": 2.8911,
+      "step": 5351
+    },
+    {
+      "epoch": 1.1391327480712956,
+      "grad_norm": 0.8588464441958568,
+      "learning_rate": 7.800206689025785e-05,
+      "loss": 2.9664,
+      "step": 5352
+    },
+    {
+      "epoch": 1.139345570630487,
+      "grad_norm": 0.9257521443725658,
+      "learning_rate": 7.799180626125843e-05,
+      "loss": 2.9378,
+      "step": 5353
+    },
+    {
+      "epoch": 1.1395583931896782,
+      "grad_norm": 1.0727414804106477,
+      "learning_rate": 7.798154391506961e-05,
+      "loss": 2.9032,
+      "step": 5354
+    },
+    {
+      "epoch": 1.1397712157488693,
+      "grad_norm": 0.8901498829553984,
+      "learning_rate": 7.797127985232095e-05,
+      "loss": 2.9272,
+      "step": 5355
+    },
+    {
+      "epoch": 1.1399840383080606,
+      "grad_norm": 0.8835634689894396,
+      "learning_rate": 7.79610140736421e-05,
+      "loss": 2.9308,
+      "step": 5356
+    },
+    {
+      "epoch": 1.140196860867252,
+      "grad_norm": 0.8856773444706283,
+      "learning_rate": 7.79507465796628e-05,
+      "loss": 2.957,
+      "step": 5357
+    },
+    {
+      "epoch": 1.1404096834264432,
+      "grad_norm": 0.8675642723564407,
+      "learning_rate": 7.794047737101297e-05,
+      "loss": 2.9079,
+      "step": 5358
+    },
+    {
+      "epoch": 1.1406225059856345,
+      "grad_norm": 0.9150666476574398,
+      "learning_rate": 7.793020644832256e-05,
+      "loss": 2.9747,
+      "step": 5359
+    },
+    {
+      "epoch": 1.1408353285448258,
+      "grad_norm": 0.9135273756725082,
+      "learning_rate": 7.791993381222166e-05,
+      "loss": 2.9223,
+      "step": 5360
+    },
+    {
+      "epoch": 1.141048151104017,
+      "grad_norm": 0.8511874551343338,
+      "learning_rate": 7.790965946334045e-05,
+      "loss": 2.9324,
+      "step": 5361
+    },
+    {
+      "epoch": 1.1412609736632082,
+      "grad_norm": 0.8283247985942153,
+      "learning_rate": 7.789938340230923e-05,
+      "loss": 2.9419,
+      "step": 5362
+    },
+    {
+      "epoch": 1.1414737962223995,
+      "grad_norm": 0.7928390290749885,
+      "learning_rate": 7.78891056297584e-05,
+      "loss": 2.924,
+      "step": 5363
+    },
+    {
+      "epoch": 1.1416866187815908,
+      "grad_norm": 0.87169496812841,
+      "learning_rate": 7.787882614631843e-05,
+      "loss": 2.9518,
+      "step": 5364
+    },
+    {
+      "epoch": 1.1418994413407821,
+      "grad_norm": 0.9580022444028927,
+      "learning_rate": 7.786854495261997e-05,
+      "loss": 2.9384,
+      "step": 5365
+    },
+    {
+      "epoch": 1.1421122638999734,
+      "grad_norm": 0.9964879791282669,
+      "learning_rate": 7.78582620492937e-05,
+      "loss": 2.9466,
+      "step": 5366
+    },
+    {
+      "epoch": 1.1423250864591648,
+      "grad_norm": 1.0800889599063728,
+      "learning_rate": 7.784797743697045e-05,
+      "loss": 2.915,
+      "step": 5367
+    },
+    {
+      "epoch": 1.142537909018356,
+      "grad_norm": 0.9396886305153452,
+      "learning_rate": 7.783769111628114e-05,
+      "loss": 2.8922,
+      "step": 5368
+    },
+    {
+      "epoch": 1.1427507315775471,
+      "grad_norm": 0.9287353223290057,
+      "learning_rate": 7.78274030878568e-05,
+      "loss": 2.9261,
+      "step": 5369
+    },
+    {
+      "epoch": 1.1429635541367384,
+      "grad_norm": 0.9850919006175896,
+      "learning_rate": 7.781711335232856e-05,
+      "loss": 2.9826,
+      "step": 5370
+    },
+    {
+      "epoch": 1.1431763766959298,
+      "grad_norm": 1.0949930935490877,
+      "learning_rate": 7.780682191032765e-05,
+      "loss": 2.9281,
+      "step": 5371
+    },
+    {
+      "epoch": 1.143389199255121,
+      "grad_norm": 0.9374751249421087,
+      "learning_rate": 7.779652876248541e-05,
+      "loss": 2.8812,
+      "step": 5372
+    },
+    {
+      "epoch": 1.1436020218143124,
+      "grad_norm": 0.9827600168006331,
+      "learning_rate": 7.778623390943328e-05,
+      "loss": 2.8957,
+      "step": 5373
+    },
+    {
+      "epoch": 1.1438148443735037,
+      "grad_norm": 0.983527989427136,
+      "learning_rate": 7.777593735180282e-05,
+      "loss": 2.9469,
+      "step": 5374
+    },
+    {
+      "epoch": 1.1440276669326948,
+      "grad_norm": 0.9830020750357298,
+      "learning_rate": 7.77656390902257e-05,
+      "loss": 2.9617,
+      "step": 5375
+    },
+    {
+      "epoch": 1.144240489491886,
+      "grad_norm": 1.1296183264999715,
+      "learning_rate": 7.775533912533364e-05,
+      "loss": 2.9258,
+      "step": 5376
+    },
+    {
+      "epoch": 1.1444533120510774,
+      "grad_norm": 0.8358188379277425,
+      "learning_rate": 7.774503745775852e-05,
+      "loss": 2.9573,
+      "step": 5377
+    },
+    {
+      "epoch": 1.1446661346102687,
+      "grad_norm": 0.8497455911128802,
+      "learning_rate": 7.773473408813232e-05,
+      "loss": 2.9135,
+      "step": 5378
+    },
+    {
+      "epoch": 1.14487895716946,
+      "grad_norm": 0.8298391898393195,
+      "learning_rate": 7.772442901708709e-05,
+      "loss": 2.9213,
+      "step": 5379
+    },
+    {
+      "epoch": 1.1450917797286513,
+      "grad_norm": 0.8417855698360038,
+      "learning_rate": 7.771412224525502e-05,
+      "loss": 2.9278,
+      "step": 5380
+    },
+    {
+      "epoch": 1.1453046022878426,
+      "grad_norm": 0.954345546104913,
+      "learning_rate": 7.770381377326841e-05,
+      "loss": 2.9456,
+      "step": 5381
+    },
+    {
+      "epoch": 1.145517424847034,
+      "grad_norm": 1.0370303809421382,
+      "learning_rate": 7.769350360175962e-05,
+      "loss": 2.9258,
+      "step": 5382
+    },
+    {
+      "epoch": 1.145730247406225,
+      "grad_norm": 1.0231171658775111,
+      "learning_rate": 7.768319173136114e-05,
+      "loss": 2.9323,
+      "step": 5383
+    },
+    {
+      "epoch": 1.1459430699654163,
+      "grad_norm": 0.9783942486355288,
+      "learning_rate": 7.767287816270557e-05,
+      "loss": 2.9203,
+      "step": 5384
+    },
+    {
+      "epoch": 1.1461558925246076,
+      "grad_norm": 0.8672866664484602,
+      "learning_rate": 7.76625628964256e-05,
+      "loss": 2.9247,
+      "step": 5385
+    },
+    {
+      "epoch": 1.146368715083799,
+      "grad_norm": 0.880312752665827,
+      "learning_rate": 7.765224593315402e-05,
+      "loss": 2.8998,
+      "step": 5386
+    },
+    {
+      "epoch": 1.1465815376429902,
+      "grad_norm": 0.9022363600181423,
+      "learning_rate": 7.764192727352378e-05,
+      "loss": 2.9052,
+      "step": 5387
+    },
+    {
+      "epoch": 1.1467943602021815,
+      "grad_norm": 0.9582325893214825,
+      "learning_rate": 7.763160691816784e-05,
+      "loss": 2.8998,
+      "step": 5388
+    },
+    {
+      "epoch": 1.1470071827613726,
+      "grad_norm": 1.0032008033662703,
+      "learning_rate": 7.762128486771936e-05,
+      "loss": 2.9256,
+      "step": 5389
+    },
+    {
+      "epoch": 1.147220005320564,
+      "grad_norm": 0.9539388758682944,
+      "learning_rate": 7.76109611228115e-05,
+      "loss": 2.9525,
+      "step": 5390
+    },
+    {
+      "epoch": 1.1474328278797552,
+      "grad_norm": 0.9054837770747121,
+      "learning_rate": 7.760063568407764e-05,
+      "loss": 2.9331,
+      "step": 5391
+    },
+    {
+      "epoch": 1.1476456504389465,
+      "grad_norm": 0.9527186717490469,
+      "learning_rate": 7.759030855215119e-05,
+      "loss": 2.9174,
+      "step": 5392
+    },
+    {
+      "epoch": 1.1478584729981378,
+      "grad_norm": 0.9120294993701101,
+      "learning_rate": 7.757997972766566e-05,
+      "loss": 2.9044,
+      "step": 5393
+    },
+    {
+      "epoch": 1.1480712955573291,
+      "grad_norm": 0.8683491069952083,
+      "learning_rate": 7.75696492112547e-05,
+      "loss": 2.9747,
+      "step": 5394
+    },
+    {
+      "epoch": 1.1482841181165204,
+      "grad_norm": 0.9125642266940358,
+      "learning_rate": 7.755931700355205e-05,
+      "loss": 2.9132,
+      "step": 5395
+    },
+    {
+      "epoch": 1.1484969406757117,
+      "grad_norm": 0.9202716751149904,
+      "learning_rate": 7.754898310519154e-05,
+      "loss": 2.9015,
+      "step": 5396
+    },
+    {
+      "epoch": 1.1487097632349028,
+      "grad_norm": 0.8459144390599485,
+      "learning_rate": 7.75386475168071e-05,
+      "loss": 2.9202,
+      "step": 5397
+    },
+    {
+      "epoch": 1.1489225857940941,
+      "grad_norm": 0.7766231021559034,
+      "learning_rate": 7.752831023903282e-05,
+      "loss": 2.9276,
+      "step": 5398
+    },
+    {
+      "epoch": 1.1491354083532854,
+      "grad_norm": 0.8433433902428289,
+      "learning_rate": 7.751797127250284e-05,
+      "loss": 2.9184,
+      "step": 5399
+    },
+    {
+      "epoch": 1.1493482309124767,
+      "grad_norm": 0.9595342319734407,
+      "learning_rate": 7.750763061785138e-05,
+      "loss": 2.9475,
+      "step": 5400
+    },
+    {
+      "epoch": 1.149561053471668,
+      "grad_norm": 1.009959900072165,
+      "learning_rate": 7.749728827571284e-05,
+      "loss": 2.9492,
+      "step": 5401
+    },
+    {
+      "epoch": 1.1497738760308593,
+      "grad_norm": 0.9272759437997516,
+      "learning_rate": 7.748694424672166e-05,
+      "loss": 2.9509,
+      "step": 5402
+    },
+    {
+      "epoch": 1.1499866985900504,
+      "grad_norm": 0.995668672315105,
+      "learning_rate": 7.747659853151241e-05,
+      "loss": 2.9551,
+      "step": 5403
+    },
+    {
+      "epoch": 1.1501995211492417,
+      "grad_norm": 1.0441572464774431,
+      "learning_rate": 7.746625113071977e-05,
+      "loss": 2.9114,
+      "step": 5404
+    },
+    {
+      "epoch": 1.150412343708433,
+      "grad_norm": 0.8924670580382171,
+      "learning_rate": 7.74559020449785e-05,
+      "loss": 2.9491,
+      "step": 5405
+    },
+    {
+      "epoch": 1.1506251662676243,
+      "grad_norm": 0.8402175393405731,
+      "learning_rate": 7.74455512749235e-05,
+      "loss": 2.9312,
+      "step": 5406
+    },
+    {
+      "epoch": 1.1508379888268156,
+      "grad_norm": 0.8150320410977479,
+      "learning_rate": 7.743519882118973e-05,
+      "loss": 2.9389,
+      "step": 5407
+    },
+    {
+      "epoch": 1.151050811386007,
+      "grad_norm": 1.0124423581271622,
+      "learning_rate": 7.742484468441226e-05,
+      "loss": 2.9236,
+      "step": 5408
+    },
+    {
+      "epoch": 1.1512636339451983,
+      "grad_norm": 1.0229552305001193,
+      "learning_rate": 7.741448886522631e-05,
+      "loss": 2.8712,
+      "step": 5409
+    },
+    {
+      "epoch": 1.1514764565043896,
+      "grad_norm": 1.0386620746049202,
+      "learning_rate": 7.740413136426714e-05,
+      "loss": 2.9184,
+      "step": 5410
+    },
+    {
+      "epoch": 1.1516892790635807,
+      "grad_norm": 0.9392230173792743,
+      "learning_rate": 7.739377218217016e-05,
+      "loss": 2.9307,
+      "step": 5411
+    },
+    {
+      "epoch": 1.151902101622772,
+      "grad_norm": 0.9083462324304655,
+      "learning_rate": 7.738341131957085e-05,
+      "loss": 2.8914,
+      "step": 5412
+    },
+    {
+      "epoch": 1.1521149241819633,
+      "grad_norm": 0.8583845500595634,
+      "learning_rate": 7.737304877710482e-05,
+      "loss": 2.9289,
+      "step": 5413
+    },
+    {
+      "epoch": 1.1523277467411546,
+      "grad_norm": 0.9552207538655636,
+      "learning_rate": 7.736268455540781e-05,
+      "loss": 2.9043,
+      "step": 5414
+    },
+    {
+      "epoch": 1.1525405693003459,
+      "grad_norm": 1.019371589516151,
+      "learning_rate": 7.735231865511555e-05,
+      "loss": 2.9595,
+      "step": 5415
+    },
+    {
+      "epoch": 1.1527533918595372,
+      "grad_norm": 0.983198759611152,
+      "learning_rate": 7.7341951076864e-05,
+      "loss": 2.9301,
+      "step": 5416
+    },
+    {
+      "epoch": 1.1529662144187283,
+      "grad_norm": 0.8943107473054633,
+      "learning_rate": 7.733158182128914e-05,
+      "loss": 2.9163,
+      "step": 5417
+    },
+    {
+      "epoch": 1.1531790369779196,
+      "grad_norm": 0.8604780945972111,
+      "learning_rate": 7.732121088902711e-05,
+      "loss": 2.9559,
+      "step": 5418
+    },
+    {
+      "epoch": 1.1533918595371109,
+      "grad_norm": 0.9592643694889298,
+      "learning_rate": 7.731083828071412e-05,
+      "loss": 2.9186,
+      "step": 5419
+    },
+    {
+      "epoch": 1.1536046820963022,
+      "grad_norm": 1.1257954624628475,
+      "learning_rate": 7.730046399698648e-05,
+      "loss": 2.9043,
+      "step": 5420
+    },
+    {
+      "epoch": 1.1538175046554935,
+      "grad_norm": 0.9527521543538983,
+      "learning_rate": 7.729008803848063e-05,
+      "loss": 2.9701,
+      "step": 5421
+    },
+    {
+      "epoch": 1.1540303272146848,
+      "grad_norm": 0.9345654348739101,
+      "learning_rate": 7.727971040583307e-05,
+      "loss": 2.9,
+      "step": 5422
+    },
+    {
+      "epoch": 1.154243149773876,
+      "grad_norm": 0.9631406011614883,
+      "learning_rate": 7.726933109968044e-05,
+      "loss": 2.9147,
+      "step": 5423
+    },
+    {
+      "epoch": 1.1544559723330674,
+      "grad_norm": 0.985193745481683,
+      "learning_rate": 7.725895012065947e-05,
+      "loss": 2.9018,
+      "step": 5424
+    },
+    {
+      "epoch": 1.1546687948922585,
+      "grad_norm": 0.913352690110057,
+      "learning_rate": 7.724856746940701e-05,
+      "loss": 2.9352,
+      "step": 5425
+    },
+    {
+      "epoch": 1.1548816174514498,
+      "grad_norm": 0.880330338878286,
+      "learning_rate": 7.723818314655998e-05,
+      "loss": 2.88,
+      "step": 5426
+    },
+    {
+      "epoch": 1.155094440010641,
+      "grad_norm": 0.8631867689272269,
+      "learning_rate": 7.722779715275541e-05,
+      "loss": 2.9178,
+      "step": 5427
+    },
+    {
+      "epoch": 1.1553072625698324,
+      "grad_norm": 0.910513689588746,
+      "learning_rate": 7.721740948863043e-05,
+      "loss": 2.908,
+      "step": 5428
+    },
+    {
+      "epoch": 1.1555200851290237,
+      "grad_norm": 0.914073993371374,
+      "learning_rate": 7.720702015482233e-05,
+      "loss": 2.9552,
+      "step": 5429
+    },
+    {
+      "epoch": 1.155732907688215,
+      "grad_norm": 0.8637556706954062,
+      "learning_rate": 7.719662915196844e-05,
+      "loss": 2.9099,
+      "step": 5430
+    },
+    {
+      "epoch": 1.155945730247406,
+      "grad_norm": 0.8600550706356757,
+      "learning_rate": 7.718623648070615e-05,
+      "loss": 2.9103,
+      "step": 5431
+    },
+    {
+      "epoch": 1.1561585528065974,
+      "grad_norm": 0.8235947952133171,
+      "learning_rate": 7.717584214167308e-05,
+      "loss": 2.9404,
+      "step": 5432
+    },
+    {
+      "epoch": 1.1563713753657887,
+      "grad_norm": 0.8594470765973282,
+      "learning_rate": 7.716544613550687e-05,
+      "loss": 2.9655,
+      "step": 5433
+    },
+    {
+      "epoch": 1.15658419792498,
+      "grad_norm": 0.8096281186722398,
+      "learning_rate": 7.715504846284526e-05,
+      "loss": 2.8848,
+      "step": 5434
+    },
+    {
+      "epoch": 1.1567970204841713,
+      "grad_norm": 0.7090668021107963,
+      "learning_rate": 7.71446491243261e-05,
+      "loss": 2.9362,
+      "step": 5435
+    },
+    {
+      "epoch": 1.1570098430433626,
+      "grad_norm": 0.8704519433714029,
+      "learning_rate": 7.713424812058736e-05,
+      "loss": 2.9359,
+      "step": 5436
+    },
+    {
+      "epoch": 1.157222665602554,
+      "grad_norm": 0.8667956505630011,
+      "learning_rate": 7.712384545226711e-05,
+      "loss": 2.922,
+      "step": 5437
+    },
+    {
+      "epoch": 1.1574354881617452,
+      "grad_norm": 0.8936672408772566,
+      "learning_rate": 7.71134411200035e-05,
+      "loss": 2.9192,
+      "step": 5438
+    },
+    {
+      "epoch": 1.1576483107209363,
+      "grad_norm": 0.9288275329247936,
+      "learning_rate": 7.710303512443481e-05,
+      "loss": 2.9057,
+      "step": 5439
+    },
+    {
+      "epoch": 1.1578611332801276,
+      "grad_norm": 0.9139875417260322,
+      "learning_rate": 7.709262746619939e-05,
+      "loss": 2.8991,
+      "step": 5440
+    },
+    {
+      "epoch": 1.158073955839319,
+      "grad_norm": 0.8507343502503452,
+      "learning_rate": 7.708221814593573e-05,
+      "loss": 2.9204,
+      "step": 5441
+    },
+    {
+      "epoch": 1.1582867783985102,
+      "grad_norm": 0.8729221944690895,
+      "learning_rate": 7.707180716428237e-05,
+      "loss": 2.9459,
+      "step": 5442
+    },
+    {
+      "epoch": 1.1584996009577015,
+      "grad_norm": 0.9798255023617907,
+      "learning_rate": 7.706139452187802e-05,
+      "loss": 2.9331,
+      "step": 5443
+    },
+    {
+      "epoch": 1.1587124235168929,
+      "grad_norm": 1.0062442844604953,
+      "learning_rate": 7.705098021936142e-05,
+      "loss": 2.9511,
+      "step": 5444
+    },
+    {
+      "epoch": 1.1589252460760842,
+      "grad_norm": 0.9350357201949324,
+      "learning_rate": 7.704056425737148e-05,
+      "loss": 2.9325,
+      "step": 5445
+    },
+    {
+      "epoch": 1.1591380686352752,
+      "grad_norm": 0.9514484985592406,
+      "learning_rate": 7.703014663654717e-05,
+      "loss": 2.9062,
+      "step": 5446
+    },
+    {
+      "epoch": 1.1593508911944665,
+      "grad_norm": 0.9666369955347336,
+      "learning_rate": 7.701972735752756e-05,
+      "loss": 2.9514,
+      "step": 5447
+    },
+    {
+      "epoch": 1.1595637137536579,
+      "grad_norm": 0.9794125140701362,
+      "learning_rate": 7.700930642095184e-05,
+      "loss": 2.9108,
+      "step": 5448
+    },
+    {
+      "epoch": 1.1597765363128492,
+      "grad_norm": 0.9208104514843006,
+      "learning_rate": 7.69988838274593e-05,
+      "loss": 2.9312,
+      "step": 5449
+    },
+    {
+      "epoch": 1.1599893588720405,
+      "grad_norm": 0.9064346421176434,
+      "learning_rate": 7.698845957768932e-05,
+      "loss": 2.9098,
+      "step": 5450
+    },
+    {
+      "epoch": 1.1602021814312318,
+      "grad_norm": 0.9490022891579978,
+      "learning_rate": 7.697803367228136e-05,
+      "loss": 2.9081,
+      "step": 5451
+    },
+    {
+      "epoch": 1.160415003990423,
+      "grad_norm": 1.0388422645747903,
+      "learning_rate": 7.696760611187506e-05,
+      "loss": 2.8759,
+      "step": 5452
+    },
+    {
+      "epoch": 1.1606278265496144,
+      "grad_norm": 0.8172559573563178,
+      "learning_rate": 7.695717689711007e-05,
+      "loss": 2.9193,
+      "step": 5453
+    },
+    {
+      "epoch": 1.1608406491088055,
+      "grad_norm": 0.8025015758107642,
+      "learning_rate": 7.694674602862621e-05,
+      "loss": 2.9354,
+      "step": 5454
+    },
+    {
+      "epoch": 1.1610534716679968,
+      "grad_norm": 0.9553568424442996,
+      "learning_rate": 7.693631350706336e-05,
+      "loss": 2.8923,
+      "step": 5455
+    },
+    {
+      "epoch": 1.161266294227188,
+      "grad_norm": 1.108142751035006,
+      "learning_rate": 7.692587933306152e-05,
+      "loss": 2.9108,
+      "step": 5456
+    },
+    {
+      "epoch": 1.1614791167863794,
+      "grad_norm": 1.081908722277784,
+      "learning_rate": 7.69154435072608e-05,
+      "loss": 2.8848,
+      "step": 5457
+    },
+    {
+      "epoch": 1.1616919393455707,
+      "grad_norm": 0.8863284298105532,
+      "learning_rate": 7.690500603030135e-05,
+      "loss": 2.9048,
+      "step": 5458
+    },
+    {
+      "epoch": 1.161904761904762,
+      "grad_norm": 0.9076489199982916,
+      "learning_rate": 7.689456690282351e-05,
+      "loss": 2.892,
+      "step": 5459
+    },
+    {
+      "epoch": 1.162117584463953,
+      "grad_norm": 0.9656094223345267,
+      "learning_rate": 7.688412612546769e-05,
+      "loss": 2.9039,
+      "step": 5460
+    },
+    {
+      "epoch": 1.1623304070231444,
+      "grad_norm": 0.9482022060922478,
+      "learning_rate": 7.687368369887435e-05,
+      "loss": 2.925,
+      "step": 5461
+    },
+    {
+      "epoch": 1.1625432295823357,
+      "grad_norm": 0.8899098312131051,
+      "learning_rate": 7.686323962368411e-05,
+      "loss": 2.9265,
+      "step": 5462
+    },
+    {
+      "epoch": 1.162756052141527,
+      "grad_norm": 0.7952494847387725,
+      "learning_rate": 7.68527939005377e-05,
+      "loss": 2.8872,
+      "step": 5463
+    },
+    {
+      "epoch": 1.1629688747007183,
+      "grad_norm": 0.8236765584275473,
+      "learning_rate": 7.68423465300759e-05,
+      "loss": 2.891,
+      "step": 5464
+    },
+    {
+      "epoch": 1.1631816972599096,
+      "grad_norm": 0.7942885428942726,
+      "learning_rate": 7.68318975129396e-05,
+      "loss": 2.9146,
+      "step": 5465
+    },
+    {
+      "epoch": 1.163394519819101,
+      "grad_norm": 0.8730800861633203,
+      "learning_rate": 7.682144684976983e-05,
+      "loss": 2.8873,
+      "step": 5466
+    },
+    {
+      "epoch": 1.1636073423782922,
+      "grad_norm": 0.9604074189091254,
+      "learning_rate": 7.681099454120771e-05,
+      "loss": 2.9093,
+      "step": 5467
+    },
+    {
+      "epoch": 1.1638201649374833,
+      "grad_norm": 0.9878581496515197,
+      "learning_rate": 7.680054058789442e-05,
+      "loss": 2.9166,
+      "step": 5468
+    },
+    {
+      "epoch": 1.1640329874966746,
+      "grad_norm": 1.0059298039022166,
+      "learning_rate": 7.679008499047127e-05,
+      "loss": 2.942,
+      "step": 5469
+    },
+    {
+      "epoch": 1.164245810055866,
+      "grad_norm": 0.9854515910169157,
+      "learning_rate": 7.677962774957971e-05,
+      "loss": 2.9497,
+      "step": 5470
+    },
+    {
+      "epoch": 1.1644586326150572,
+      "grad_norm": 0.9830799153695737,
+      "learning_rate": 7.67691688658612e-05,
+      "loss": 2.902,
+      "step": 5471
+    },
+    {
+      "epoch": 1.1646714551742485,
+      "grad_norm": 1.0997462093786596,
+      "learning_rate": 7.67587083399574e-05,
+      "loss": 2.8928,
+      "step": 5472
+    },
+    {
+      "epoch": 1.1648842777334398,
+      "grad_norm": 1.0128123879064808,
+      "learning_rate": 7.674824617250998e-05,
+      "loss": 2.9742,
+      "step": 5473
+    },
+    {
+      "epoch": 1.165097100292631,
+      "grad_norm": 1.086705219164936,
+      "learning_rate": 7.67377823641608e-05,
+      "loss": 2.9323,
+      "step": 5474
+    },
+    {
+      "epoch": 1.1653099228518222,
+      "grad_norm": 0.8751690110131456,
+      "learning_rate": 7.672731691555173e-05,
+      "loss": 2.9344,
+      "step": 5475
+    },
+    {
+      "epoch": 1.1655227454110135,
+      "grad_norm": 0.8604770973144237,
+      "learning_rate": 7.671684982732479e-05,
+      "loss": 2.9381,
+      "step": 5476
+    },
+    {
+      "epoch": 1.1657355679702048,
+      "grad_norm": 0.9114095614217692,
+      "learning_rate": 7.670638110012212e-05,
+      "loss": 2.9515,
+      "step": 5477
+    },
+    {
+      "epoch": 1.1659483905293961,
+      "grad_norm": 0.945843839581014,
+      "learning_rate": 7.669591073458591e-05,
+      "loss": 2.9807,
+      "step": 5478
+    },
+    {
+      "epoch": 1.1661612130885874,
+      "grad_norm": 0.951537489993335,
+      "learning_rate": 7.668543873135852e-05,
+      "loss": 2.9324,
+      "step": 5479
+    },
+    {
+      "epoch": 1.1663740356477788,
+      "grad_norm": 0.8968770305890399,
+      "learning_rate": 7.667496509108232e-05,
+      "loss": 2.9224,
+      "step": 5480
+    },
+    {
+      "epoch": 1.16658685820697,
+      "grad_norm": 0.8949701760921108,
+      "learning_rate": 7.666448981439984e-05,
+      "loss": 2.914,
+      "step": 5481
+    },
+    {
+      "epoch": 1.1667996807661611,
+      "grad_norm": 1.0051638403039584,
+      "learning_rate": 7.66540129019537e-05,
+      "loss": 2.8936,
+      "step": 5482
+    },
+    {
+      "epoch": 1.1670125033253524,
+      "grad_norm": 1.0556491540996629,
+      "learning_rate": 7.664353435438663e-05,
+      "loss": 2.9285,
+      "step": 5483
+    },
+    {
+      "epoch": 1.1672253258845438,
+      "grad_norm": 0.9668473362717581,
+      "learning_rate": 7.663305417234145e-05,
+      "loss": 2.8869,
+      "step": 5484
+    },
+    {
+      "epoch": 1.167438148443735,
+      "grad_norm": 0.931447145346994,
+      "learning_rate": 7.662257235646104e-05,
+      "loss": 2.9413,
+      "step": 5485
+    },
+    {
+      "epoch": 1.1676509710029264,
+      "grad_norm": 0.898840965455417,
+      "learning_rate": 7.661208890738847e-05,
+      "loss": 2.8831,
+      "step": 5486
+    },
+    {
+      "epoch": 1.1678637935621177,
+      "grad_norm": 0.8744630268839054,
+      "learning_rate": 7.660160382576683e-05,
+      "loss": 2.9748,
+      "step": 5487
+    },
+    {
+      "epoch": 1.1680766161213088,
+      "grad_norm": 0.8246406362749444,
+      "learning_rate": 7.659111711223935e-05,
+      "loss": 2.9239,
+      "step": 5488
+    },
+    {
+      "epoch": 1.1682894386805,
+      "grad_norm": 0.9635398905885764,
+      "learning_rate": 7.658062876744934e-05,
+      "loss": 2.9356,
+      "step": 5489
+    },
+    {
+      "epoch": 1.1685022612396914,
+      "grad_norm": 0.9421974038455232,
+      "learning_rate": 7.657013879204022e-05,
+      "loss": 2.9485,
+      "step": 5490
+    },
+    {
+      "epoch": 1.1687150837988827,
+      "grad_norm": 0.870284179812366,
+      "learning_rate": 7.655964718665554e-05,
+      "loss": 2.951,
+      "step": 5491
+    },
+    {
+      "epoch": 1.168927906358074,
+      "grad_norm": 0.9246405511550115,
+      "learning_rate": 7.654915395193886e-05,
+      "loss": 2.9205,
+      "step": 5492
+    },
+    {
+      "epoch": 1.1691407289172653,
+      "grad_norm": 1.1052087880096977,
+      "learning_rate": 7.653865908853394e-05,
+      "loss": 2.9268,
+      "step": 5493
+    },
+    {
+      "epoch": 1.1693535514764566,
+      "grad_norm": 0.8514562982764782,
+      "learning_rate": 7.652816259708461e-05,
+      "loss": 2.9499,
+      "step": 5494
+    },
+    {
+      "epoch": 1.169566374035648,
+      "grad_norm": 0.8070442578776573,
+      "learning_rate": 7.651766447823476e-05,
+      "loss": 2.9446,
+      "step": 5495
+    },
+    {
+      "epoch": 1.169779196594839,
+      "grad_norm": 0.892471740137694,
+      "learning_rate": 7.650716473262842e-05,
+      "loss": 2.9045,
+      "step": 5496
+    },
+    {
+      "epoch": 1.1699920191540303,
+      "grad_norm": 1.0614601109215365,
+      "learning_rate": 7.649666336090971e-05,
+      "loss": 2.9164,
+      "step": 5497
+    },
+    {
+      "epoch": 1.1702048417132216,
+      "grad_norm": 1.0147397090122874,
+      "learning_rate": 7.648616036372287e-05,
+      "loss": 2.911,
+      "step": 5498
+    },
+    {
+      "epoch": 1.170417664272413,
+      "grad_norm": 0.9544525548354847,
+      "learning_rate": 7.647565574171217e-05,
+      "loss": 2.9411,
+      "step": 5499
+    },
+    {
+      "epoch": 1.1706304868316042,
+      "grad_norm": 0.9671451474356675,
+      "learning_rate": 7.646514949552206e-05,
+      "loss": 2.9276,
+      "step": 5500
+    },
+    {
+      "epoch": 1.1708433093907955,
+      "grad_norm": 0.9191705628931006,
+      "learning_rate": 7.645464162579705e-05,
+      "loss": 2.9217,
+      "step": 5501
+    },
+    {
+      "epoch": 1.1710561319499866,
+      "grad_norm": 0.9224244829626144,
+      "learning_rate": 7.644413213318177e-05,
+      "loss": 2.9868,
+      "step": 5502
+    },
+    {
+      "epoch": 1.171268954509178,
+      "grad_norm": 1.1047912382557057,
+      "learning_rate": 7.643362101832093e-05,
+      "loss": 2.8815,
+      "step": 5503
+    },
+    {
+      "epoch": 1.1714817770683692,
+      "grad_norm": 0.9224601273718405,
+      "learning_rate": 7.642310828185933e-05,
+      "loss": 2.9043,
+      "step": 5504
+    },
+    {
+      "epoch": 1.1716945996275605,
+      "grad_norm": 0.9095063515434599,
+      "learning_rate": 7.641259392444192e-05,
+      "loss": 2.9286,
+      "step": 5505
+    },
+    {
+      "epoch": 1.1719074221867518,
+      "grad_norm": 0.917600062868107,
+      "learning_rate": 7.640207794671368e-05,
+      "loss": 2.8996,
+      "step": 5506
+    },
+    {
+      "epoch": 1.1721202447459431,
+      "grad_norm": 0.8827746918454858,
+      "learning_rate": 7.639156034931974e-05,
+      "loss": 2.9049,
+      "step": 5507
+    },
+    {
+      "epoch": 1.1723330673051344,
+      "grad_norm": 0.9022829064530262,
+      "learning_rate": 7.638104113290532e-05,
+      "loss": 2.8949,
+      "step": 5508
+    },
+    {
+      "epoch": 1.1725458898643257,
+      "grad_norm": 0.9419863419262106,
+      "learning_rate": 7.637052029811571e-05,
+      "loss": 2.9322,
+      "step": 5509
+    },
+    {
+      "epoch": 1.1727587124235168,
+      "grad_norm": 1.0637061000933854,
+      "learning_rate": 7.635999784559636e-05,
+      "loss": 2.9334,
+      "step": 5510
+    },
+    {
+      "epoch": 1.1729715349827081,
+      "grad_norm": 0.9405307406022168,
+      "learning_rate": 7.634947377599273e-05,
+      "loss": 2.8982,
+      "step": 5511
+    },
+    {
+      "epoch": 1.1731843575418994,
+      "grad_norm": 0.7993395274619032,
+      "learning_rate": 7.63389480899505e-05,
+      "loss": 2.9263,
+      "step": 5512
+    },
+    {
+      "epoch": 1.1733971801010907,
+      "grad_norm": 0.9239083342967039,
+      "learning_rate": 7.632842078811532e-05,
+      "loss": 2.9468,
+      "step": 5513
+    },
+    {
+      "epoch": 1.173610002660282,
+      "grad_norm": 1.0681259477688005,
+      "learning_rate": 7.631789187113303e-05,
+      "loss": 2.9119,
+      "step": 5514
+    },
+    {
+      "epoch": 1.1738228252194733,
+      "grad_norm": 0.9090115786766532,
+      "learning_rate": 7.630736133964953e-05,
+      "loss": 2.9314,
+      "step": 5515
+    },
+    {
+      "epoch": 1.1740356477786644,
+      "grad_norm": 0.8727098846367082,
+      "learning_rate": 7.629682919431083e-05,
+      "loss": 2.8966,
+      "step": 5516
+    },
+    {
+      "epoch": 1.1742484703378557,
+      "grad_norm": 1.0642979017278547,
+      "learning_rate": 7.628629543576302e-05,
+      "loss": 2.9299,
+      "step": 5517
+    },
+    {
+      "epoch": 1.174461292897047,
+      "grad_norm": 1.1013416945762156,
+      "learning_rate": 7.627576006465234e-05,
+      "loss": 2.9285,
+      "step": 5518
+    },
+    {
+      "epoch": 1.1746741154562383,
+      "grad_norm": 0.9694632531197206,
+      "learning_rate": 7.626522308162507e-05,
+      "loss": 2.9691,
+      "step": 5519
+    },
+    {
+      "epoch": 1.1748869380154296,
+      "grad_norm": 0.9503204595028362,
+      "learning_rate": 7.62546844873276e-05,
+      "loss": 2.9516,
+      "step": 5520
+    },
+    {
+      "epoch": 1.175099760574621,
+      "grad_norm": 1.0008014707658706,
+      "learning_rate": 7.624414428240647e-05,
+      "loss": 2.8883,
+      "step": 5521
+    },
+    {
+      "epoch": 1.1753125831338123,
+      "grad_norm": 0.9632613551168897,
+      "learning_rate": 7.623360246750825e-05,
+      "loss": 2.9213,
+      "step": 5522
+    },
+    {
+      "epoch": 1.1755254056930036,
+      "grad_norm": 0.9258697095669499,
+      "learning_rate": 7.622305904327966e-05,
+      "loss": 2.9095,
+      "step": 5523
+    },
+    {
+      "epoch": 1.1757382282521947,
+      "grad_norm": 0.9604384178562116,
+      "learning_rate": 7.621251401036748e-05,
+      "loss": 2.9193,
+      "step": 5524
+    },
+    {
+      "epoch": 1.175951050811386,
+      "grad_norm": 0.9499804800297327,
+      "learning_rate": 7.620196736941863e-05,
+      "loss": 2.9483,
+      "step": 5525
+    },
+    {
+      "epoch": 1.1761638733705773,
+      "grad_norm": 1.0646906112568084,
+      "learning_rate": 7.619141912108008e-05,
+      "loss": 2.918,
+      "step": 5526
+    },
+    {
+      "epoch": 1.1763766959297686,
+      "grad_norm": 1.0940876382892886,
+      "learning_rate": 7.618086926599892e-05,
+      "loss": 2.9112,
+      "step": 5527
+    },
+    {
+      "epoch": 1.1765895184889599,
+      "grad_norm": 0.887168417611082,
+      "learning_rate": 7.617031780482237e-05,
+      "loss": 2.9474,
+      "step": 5528
+    },
+    {
+      "epoch": 1.1768023410481512,
+      "grad_norm": 0.9156283888178522,
+      "learning_rate": 7.615976473819772e-05,
+      "loss": 2.9621,
+      "step": 5529
+    },
+    {
+      "epoch": 1.1770151636073423,
+      "grad_norm": 1.0957663557673178,
+      "learning_rate": 7.614921006677236e-05,
+      "loss": 2.9387,
+      "step": 5530
+    },
+    {
+      "epoch": 1.1772279861665336,
+      "grad_norm": 0.9438715208306848,
+      "learning_rate": 7.613865379119374e-05,
+      "loss": 2.9445,
+      "step": 5531
+    },
+    {
+      "epoch": 1.1774408087257249,
+      "grad_norm": 0.9182081147590025,
+      "learning_rate": 7.61280959121095e-05,
+      "loss": 2.8941,
+      "step": 5532
+    },
+    {
+      "epoch": 1.1776536312849162,
+      "grad_norm": 0.9043995965707827,
+      "learning_rate": 7.61175364301673e-05,
+      "loss": 2.943,
+      "step": 5533
+    },
+    {
+      "epoch": 1.1778664538441075,
+      "grad_norm": 0.9759962647966182,
+      "learning_rate": 7.61069753460149e-05,
+      "loss": 2.9463,
+      "step": 5534
+    },
+    {
+      "epoch": 1.1780792764032988,
+      "grad_norm": 1.0739137826344212,
+      "learning_rate": 7.609641266030023e-05,
+      "loss": 2.932,
+      "step": 5535
+    },
+    {
+      "epoch": 1.17829209896249,
+      "grad_norm": 0.8107487767963347,
+      "learning_rate": 7.608584837367123e-05,
+      "loss": 2.8881,
+      "step": 5536
+    },
+    {
+      "epoch": 1.1785049215216814,
+      "grad_norm": 0.7288382511431009,
+      "learning_rate": 7.607528248677602e-05,
+      "loss": 2.9074,
+      "step": 5537
+    },
+    {
+      "epoch": 1.1787177440808725,
+      "grad_norm": 0.9291689851475028,
+      "learning_rate": 7.606471500026273e-05,
+      "loss": 2.8949,
+      "step": 5538
+    },
+    {
+      "epoch": 1.1789305666400638,
+      "grad_norm": 0.923890274355373,
+      "learning_rate": 7.605414591477966e-05,
+      "loss": 2.9057,
+      "step": 5539
+    },
+    {
+      "epoch": 1.179143389199255,
+      "grad_norm": 0.9599146035216409,
+      "learning_rate": 7.604357523097517e-05,
+      "loss": 2.9112,
+      "step": 5540
+    },
+    {
+      "epoch": 1.1793562117584464,
+      "grad_norm": 1.0678193225640378,
+      "learning_rate": 7.603300294949775e-05,
+      "loss": 2.8923,
+      "step": 5541
+    },
+    {
+      "epoch": 1.1795690343176377,
+      "grad_norm": 0.9640939317676461,
+      "learning_rate": 7.602242907099596e-05,
+      "loss": 2.9431,
+      "step": 5542
+    },
+    {
+      "epoch": 1.179781856876829,
+      "grad_norm": 0.9314590380677863,
+      "learning_rate": 7.601185359611845e-05,
+      "loss": 2.9289,
+      "step": 5543
+    },
+    {
+      "epoch": 1.17999467943602,
+      "grad_norm": 0.8804152658801928,
+      "learning_rate": 7.600127652551401e-05,
+      "loss": 2.9132,
+      "step": 5544
+    },
+    {
+      "epoch": 1.1802075019952114,
+      "grad_norm": 0.896047160082054,
+      "learning_rate": 7.599069785983148e-05,
+      "loss": 2.9243,
+      "step": 5545
+    },
+    {
+      "epoch": 1.1804203245544027,
+      "grad_norm": 0.9947682162059343,
+      "learning_rate": 7.598011759971985e-05,
+      "loss": 2.9109,
+      "step": 5546
+    },
+    {
+      "epoch": 1.180633147113594,
+      "grad_norm": 0.9491306216413035,
+      "learning_rate": 7.596953574582814e-05,
+      "loss": 2.9263,
+      "step": 5547
+    },
+    {
+      "epoch": 1.1808459696727853,
+      "grad_norm": 0.9097752020309523,
+      "learning_rate": 7.595895229880555e-05,
+      "loss": 2.9222,
+      "step": 5548
+    },
+    {
+      "epoch": 1.1810587922319766,
+      "grad_norm": 1.0027450278763683,
+      "learning_rate": 7.594836725930128e-05,
+      "loss": 2.9441,
+      "step": 5549
+    },
+    {
+      "epoch": 1.181271614791168,
+      "grad_norm": 1.0704068621792173,
+      "learning_rate": 7.593778062796472e-05,
+      "loss": 2.9487,
+      "step": 5550
+    },
+    {
+      "epoch": 1.1814844373503592,
+      "grad_norm": 0.9185790980334623,
+      "learning_rate": 7.592719240544533e-05,
+      "loss": 2.9492,
+      "step": 5551
+    },
+    {
+      "epoch": 1.1816972599095503,
+      "grad_norm": 0.9480966732225131,
+      "learning_rate": 7.591660259239262e-05,
+      "loss": 2.91,
+      "step": 5552
+    },
+    {
+      "epoch": 1.1819100824687416,
+      "grad_norm": 0.9471273225301093,
+      "learning_rate": 7.590601118945625e-05,
+      "loss": 2.9162,
+      "step": 5553
+    },
+    {
+      "epoch": 1.182122905027933,
+      "grad_norm": 1.0173386307499406,
+      "learning_rate": 7.589541819728595e-05,
+      "loss": 2.8814,
+      "step": 5554
+    },
+    {
+      "epoch": 1.1823357275871242,
+      "grad_norm": 0.9429837245085776,
+      "learning_rate": 7.588482361653159e-05,
+      "loss": 2.9053,
+      "step": 5555
+    },
+    {
+      "epoch": 1.1825485501463155,
+      "grad_norm": 1.0860209252377506,
+      "learning_rate": 7.587422744784311e-05,
+      "loss": 2.9478,
+      "step": 5556
+    },
+    {
+      "epoch": 1.1827613727055069,
+      "grad_norm": 0.9332856132828914,
+      "learning_rate": 7.586362969187048e-05,
+      "loss": 2.9361,
+      "step": 5557
+    },
+    {
+      "epoch": 1.182974195264698,
+      "grad_norm": 0.8437026627106397,
+      "learning_rate": 7.58530303492639e-05,
+      "loss": 2.9133,
+      "step": 5558
+    },
+    {
+      "epoch": 1.1831870178238892,
+      "grad_norm": 0.854756484025853,
+      "learning_rate": 7.584242942067356e-05,
+      "loss": 2.8741,
+      "step": 5559
+    },
+    {
+      "epoch": 1.1833998403830805,
+      "grad_norm": 1.062013213765455,
+      "learning_rate": 7.583182690674982e-05,
+      "loss": 2.9021,
+      "step": 5560
+    },
+    {
+      "epoch": 1.1836126629422719,
+      "grad_norm": 0.9620907954862195,
+      "learning_rate": 7.582122280814305e-05,
+      "loss": 2.9144,
+      "step": 5561
+    },
+    {
+      "epoch": 1.1838254855014632,
+      "grad_norm": 0.8504377392037714,
+      "learning_rate": 7.581061712550381e-05,
+      "loss": 2.9401,
+      "step": 5562
+    },
+    {
+      "epoch": 1.1840383080606545,
+      "grad_norm": 1.0095431814113631,
+      "learning_rate": 7.580000985948274e-05,
+      "loss": 2.9224,
+      "step": 5563
+    },
+    {
+      "epoch": 1.1842511306198458,
+      "grad_norm": 0.991368746918506,
+      "learning_rate": 7.578940101073048e-05,
+      "loss": 2.8929,
+      "step": 5564
+    },
+    {
+      "epoch": 1.184463953179037,
+      "grad_norm": 0.8819283813377704,
+      "learning_rate": 7.577879057989791e-05,
+      "loss": 2.928,
+      "step": 5565
+    },
+    {
+      "epoch": 1.1846767757382282,
+      "grad_norm": 0.9570654545086759,
+      "learning_rate": 7.576817856763594e-05,
+      "loss": 2.9577,
+      "step": 5566
+    },
+    {
+      "epoch": 1.1848895982974195,
+      "grad_norm": 0.9988528975850127,
+      "learning_rate": 7.575756497459552e-05,
+      "loss": 2.9299,
+      "step": 5567
+    },
+    {
+      "epoch": 1.1851024208566108,
+      "grad_norm": 1.0162119521392285,
+      "learning_rate": 7.574694980142779e-05,
+      "loss": 2.9012,
+      "step": 5568
+    },
+    {
+      "epoch": 1.185315243415802,
+      "grad_norm": 0.8416124751952584,
+      "learning_rate": 7.573633304878395e-05,
+      "loss": 2.8887,
+      "step": 5569
+    },
+    {
+      "epoch": 1.1855280659749934,
+      "grad_norm": 0.8270542791078749,
+      "learning_rate": 7.57257147173153e-05,
+      "loss": 2.9201,
+      "step": 5570
+    },
+    {
+      "epoch": 1.1857408885341847,
+      "grad_norm": 0.7969725290982311,
+      "learning_rate": 7.571509480767323e-05,
+      "loss": 2.9328,
+      "step": 5571
+    },
+    {
+      "epoch": 1.1859537110933758,
+      "grad_norm": 0.9133799972960975,
+      "learning_rate": 7.57044733205092e-05,
+      "loss": 2.9525,
+      "step": 5572
+    },
+    {
+      "epoch": 1.186166533652567,
+      "grad_norm": 0.856521818556254,
+      "learning_rate": 7.569385025647484e-05,
+      "loss": 2.884,
+      "step": 5573
+    },
+    {
+      "epoch": 1.1863793562117584,
+      "grad_norm": 0.7739096854600189,
+      "learning_rate": 7.568322561622183e-05,
+      "loss": 2.88,
+      "step": 5574
+    },
+    {
+      "epoch": 1.1865921787709497,
+      "grad_norm": 0.8744137410291543,
+      "learning_rate": 7.567259940040192e-05,
+      "loss": 2.9504,
+      "step": 5575
+    },
+    {
+      "epoch": 1.186805001330141,
+      "grad_norm": 0.8950710438951405,
+      "learning_rate": 7.566197160966701e-05,
+      "loss": 2.9303,
+      "step": 5576
+    },
+    {
+      "epoch": 1.1870178238893323,
+      "grad_norm": 0.9298861079648194,
+      "learning_rate": 7.565134224466907e-05,
+      "loss": 2.9209,
+      "step": 5577
+    },
+    {
+      "epoch": 1.1872306464485236,
+      "grad_norm": 1.0037294538556527,
+      "learning_rate": 7.564071130606017e-05,
+      "loss": 2.9085,
+      "step": 5578
+    },
+    {
+      "epoch": 1.187443469007715,
+      "grad_norm": 1.0971159926779164,
+      "learning_rate": 7.563007879449249e-05,
+      "loss": 2.9133,
+      "step": 5579
+    },
+    {
+      "epoch": 1.187656291566906,
+      "grad_norm": 0.9190082502071681,
+      "learning_rate": 7.561944471061826e-05,
+      "loss": 2.9062,
+      "step": 5580
+    },
+    {
+      "epoch": 1.1878691141260973,
+      "grad_norm": 0.9406949183577917,
+      "learning_rate": 7.560880905508988e-05,
+      "loss": 2.9191,
+      "step": 5581
+    },
+    {
+      "epoch": 1.1880819366852886,
+      "grad_norm": 1.0578088947727324,
+      "learning_rate": 7.559817182855977e-05,
+      "loss": 2.9085,
+      "step": 5582
+    },
+    {
+      "epoch": 1.18829475924448,
+      "grad_norm": 0.864079461316568,
+      "learning_rate": 7.558753303168053e-05,
+      "loss": 2.9064,
+      "step": 5583
+    },
+    {
+      "epoch": 1.1885075818036712,
+      "grad_norm": 0.9289952526198332,
+      "learning_rate": 7.557689266510475e-05,
+      "loss": 2.9322,
+      "step": 5584
+    },
+    {
+      "epoch": 1.1887204043628625,
+      "grad_norm": 0.8982913367903633,
+      "learning_rate": 7.556625072948522e-05,
+      "loss": 2.867,
+      "step": 5585
+    },
+    {
+      "epoch": 1.1889332269220538,
+      "grad_norm": 0.8630765870856223,
+      "learning_rate": 7.555560722547475e-05,
+      "loss": 2.9143,
+      "step": 5586
+    },
+    {
+      "epoch": 1.189146049481245,
+      "grad_norm": 0.9086151962229494,
+      "learning_rate": 7.55449621537263e-05,
+      "loss": 2.9323,
+      "step": 5587
+    },
+    {
+      "epoch": 1.1893588720404362,
+      "grad_norm": 1.01936931076255,
+      "learning_rate": 7.553431551489291e-05,
+      "loss": 2.916,
+      "step": 5588
+    },
+    {
+      "epoch": 1.1895716945996275,
+      "grad_norm": 0.8423804471829294,
+      "learning_rate": 7.55236673096277e-05,
+      "loss": 2.9568,
+      "step": 5589
+    },
+    {
+      "epoch": 1.1897845171588188,
+      "grad_norm": 0.8579597840539948,
+      "learning_rate": 7.55130175385839e-05,
+      "loss": 2.9365,
+      "step": 5590
+    },
+    {
+      "epoch": 1.1899973397180101,
+      "grad_norm": 0.9755903115260438,
+      "learning_rate": 7.55023662024148e-05,
+      "loss": 2.9551,
+      "step": 5591
+    },
+    {
+      "epoch": 1.1902101622772014,
+      "grad_norm": 1.0303176491698744,
+      "learning_rate": 7.549171330177387e-05,
+      "loss": 2.8944,
+      "step": 5592
+    },
+    {
+      "epoch": 1.1904229848363928,
+      "grad_norm": 1.016197012066764,
+      "learning_rate": 7.54810588373146e-05,
+      "loss": 2.9304,
+      "step": 5593
+    },
+    {
+      "epoch": 1.190635807395584,
+      "grad_norm": 1.032119042208343,
+      "learning_rate": 7.54704028096906e-05,
+      "loss": 2.8766,
+      "step": 5594
+    },
+    {
+      "epoch": 1.1908486299547751,
+      "grad_norm": 0.9753798784959324,
+      "learning_rate": 7.545974521955556e-05,
+      "loss": 2.9315,
+      "step": 5595
+    },
+    {
+      "epoch": 1.1910614525139664,
+      "grad_norm": 0.9260813335271301,
+      "learning_rate": 7.544908606756332e-05,
+      "loss": 2.9297,
+      "step": 5596
+    },
+    {
+      "epoch": 1.1912742750731578,
+      "grad_norm": 0.9196050511913489,
+      "learning_rate": 7.543842535436775e-05,
+      "loss": 2.9157,
+      "step": 5597
+    },
+    {
+      "epoch": 1.191487097632349,
+      "grad_norm": 1.0142049313640282,
+      "learning_rate": 7.542776308062285e-05,
+      "loss": 2.8997,
+      "step": 5598
+    },
+    {
+      "epoch": 1.1916999201915404,
+      "grad_norm": 1.0058806128909648,
+      "learning_rate": 7.541709924698272e-05,
+      "loss": 2.9023,
+      "step": 5599
+    },
+    {
+      "epoch": 1.1919127427507317,
+      "grad_norm": 0.8872473816112751,
+      "learning_rate": 7.540643385410152e-05,
+      "loss": 2.9163,
+      "step": 5600
+    },
+    {
+      "epoch": 1.1921255653099228,
+      "grad_norm": 0.8509610700238152,
+      "learning_rate": 7.539576690263355e-05,
+      "loss": 2.9087,
+      "step": 5601
+    },
+    {
+      "epoch": 1.192338387869114,
+      "grad_norm": 0.8928068468651637,
+      "learning_rate": 7.538509839323319e-05,
+      "loss": 2.9586,
+      "step": 5602
+    },
+    {
+      "epoch": 1.1925512104283054,
+      "grad_norm": 0.9882918986792175,
+      "learning_rate": 7.537442832655489e-05,
+      "loss": 2.8736,
+      "step": 5603
+    },
+    {
+      "epoch": 1.1927640329874967,
+      "grad_norm": 0.9669378670021538,
+      "learning_rate": 7.536375670325326e-05,
+      "loss": 2.9272,
+      "step": 5604
+    },
+    {
+      "epoch": 1.192976855546688,
+      "grad_norm": 1.011199822223607,
+      "learning_rate": 7.535308352398289e-05,
+      "loss": 2.9075,
+      "step": 5605
+    },
+    {
+      "epoch": 1.1931896781058793,
+      "grad_norm": 1.0336115777052175,
+      "learning_rate": 7.534240878939862e-05,
+      "loss": 2.9618,
+      "step": 5606
+    },
+    {
+      "epoch": 1.1934025006650706,
+      "grad_norm": 1.022702733152434,
+      "learning_rate": 7.533173250015526e-05,
+      "loss": 2.9314,
+      "step": 5607
+    },
+    {
+      "epoch": 1.193615323224262,
+      "grad_norm": 0.9974982536482392,
+      "learning_rate": 7.532105465690776e-05,
+      "loss": 2.9134,
+      "step": 5608
+    },
+    {
+      "epoch": 1.193828145783453,
+      "grad_norm": 0.9798846403703685,
+      "learning_rate": 7.531037526031117e-05,
+      "loss": 2.8967,
+      "step": 5609
+    },
+    {
+      "epoch": 1.1940409683426443,
+      "grad_norm": 1.0000584666559535,
+      "learning_rate": 7.529969431102064e-05,
+      "loss": 2.9141,
+      "step": 5610
+    },
+    {
+      "epoch": 1.1942537909018356,
+      "grad_norm": 0.9532939856927313,
+      "learning_rate": 7.528901180969138e-05,
+      "loss": 2.928,
+      "step": 5611
+    },
+    {
+      "epoch": 1.194466613461027,
+      "grad_norm": 0.9832744184753078,
+      "learning_rate": 7.527832775697874e-05,
+      "loss": 2.888,
+      "step": 5612
+    },
+    {
+      "epoch": 1.1946794360202182,
+      "grad_norm": 1.0300347798328768,
+      "learning_rate": 7.526764215353816e-05,
+      "loss": 2.8916,
+      "step": 5613
+    },
+    {
+      "epoch": 1.1948922585794095,
+      "grad_norm": 0.8876801704995875,
+      "learning_rate": 7.525695500002512e-05,
+      "loss": 2.8808,
+      "step": 5614
+    },
+    {
+      "epoch": 1.1951050811386006,
+      "grad_norm": 0.8609364091379923,
+      "learning_rate": 7.524626629709527e-05,
+      "loss": 2.9503,
+      "step": 5615
+    },
+    {
+      "epoch": 1.195317903697792,
+      "grad_norm": 0.9469362089866308,
+      "learning_rate": 7.52355760454043e-05,
+      "loss": 2.9098,
+      "step": 5616
+    },
+    {
+      "epoch": 1.1955307262569832,
+      "grad_norm": 1.050011441581453,
+      "learning_rate": 7.522488424560803e-05,
+      "loss": 2.9247,
+      "step": 5617
+    },
+    {
+      "epoch": 1.1957435488161745,
+      "grad_norm": 0.8989756762262168,
+      "learning_rate": 7.521419089836236e-05,
+      "loss": 2.8518,
+      "step": 5618
+    },
+    {
+      "epoch": 1.1959563713753658,
+      "grad_norm": 0.9132289637118183,
+      "learning_rate": 7.520349600432326e-05,
+      "loss": 2.9362,
+      "step": 5619
+    },
+    {
+      "epoch": 1.1961691939345571,
+      "grad_norm": 1.0020295554522902,
+      "learning_rate": 7.519279956414686e-05,
+      "loss": 2.8904,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1963820164937484,
+      "grad_norm": 1.036733851784299,
+      "learning_rate": 7.518210157848931e-05,
+      "loss": 2.9229,
+      "step": 5621
+    },
+    {
+      "epoch": 1.1965948390529397,
+      "grad_norm": 0.9188713121091339,
+      "learning_rate": 7.517140204800692e-05,
+      "loss": 2.923,
+      "step": 5622
+    },
+    {
+      "epoch": 1.1968076616121308,
+      "grad_norm": 0.8331360168186166,
+      "learning_rate": 7.516070097335605e-05,
+      "loss": 2.8904,
+      "step": 5623
+    },
+    {
+      "epoch": 1.1970204841713221,
+      "grad_norm": 1.003104357087714,
+      "learning_rate": 7.514999835519317e-05,
+      "loss": 2.9309,
+      "step": 5624
+    },
+    {
+      "epoch": 1.1972333067305134,
+      "grad_norm": 1.0958366071831769,
+      "learning_rate": 7.513929419417485e-05,
+      "loss": 2.8939,
+      "step": 5625
+    },
+    {
+      "epoch": 1.1974461292897047,
+      "grad_norm": 0.8922223282430124,
+      "learning_rate": 7.512858849095772e-05,
+      "loss": 2.8554,
+      "step": 5626
+    },
+    {
+      "epoch": 1.197658951848896,
+      "grad_norm": 0.9474382413964885,
+      "learning_rate": 7.511788124619858e-05,
+      "loss": 2.8781,
+      "step": 5627
+    },
+    {
+      "epoch": 1.1978717744080873,
+      "grad_norm": 0.8950284424723113,
+      "learning_rate": 7.510717246055426e-05,
+      "loss": 2.9522,
+      "step": 5628
+    },
+    {
+      "epoch": 1.1980845969672784,
+      "grad_norm": 0.9174862415745073,
+      "learning_rate": 7.509646213468169e-05,
+      "loss": 2.9155,
+      "step": 5629
+    },
+    {
+      "epoch": 1.1982974195264697,
+      "grad_norm": 0.9416283351669277,
+      "learning_rate": 7.508575026923792e-05,
+      "loss": 2.8989,
+      "step": 5630
+    },
+    {
+      "epoch": 1.198510242085661,
+      "grad_norm": 0.9217700094106879,
+      "learning_rate": 7.507503686488007e-05,
+      "loss": 2.915,
+      "step": 5631
+    },
+    {
+      "epoch": 1.1987230646448523,
+      "grad_norm": 0.9342811517629068,
+      "learning_rate": 7.506432192226537e-05,
+      "loss": 2.9218,
+      "step": 5632
+    },
+    {
+      "epoch": 1.1989358872040436,
+      "grad_norm": 0.9735497879024916,
+      "learning_rate": 7.505360544205115e-05,
+      "loss": 2.9183,
+      "step": 5633
+    },
+    {
+      "epoch": 1.199148709763235,
+      "grad_norm": 0.888378555132952,
+      "learning_rate": 7.504288742489483e-05,
+      "loss": 2.9456,
+      "step": 5634
+    },
+    {
+      "epoch": 1.1993615323224263,
+      "grad_norm": 0.898788137573992,
+      "learning_rate": 7.50321678714539e-05,
+      "loss": 2.8914,
+      "step": 5635
+    },
+    {
+      "epoch": 1.1995743548816176,
+      "grad_norm": 0.8752062031476254,
+      "learning_rate": 7.502144678238598e-05,
+      "loss": 2.871,
+      "step": 5636
+    },
+    {
+      "epoch": 1.1997871774408086,
+      "grad_norm": 0.8512567875524272,
+      "learning_rate": 7.501072415834875e-05,
+      "loss": 2.9005,
+      "step": 5637
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9410257902667597,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 2.9123,
+      "step": 5638
+    },
+    {
+      "epoch": 1.2002128225591913,
+      "grad_norm": 1.0077783725756053,
+      "learning_rate": 7.498927430799765e-05,
+      "loss": 2.8686,
+      "step": 5639
+    },
+    {
+      "epoch": 1.2004256451183826,
+      "grad_norm": 1.0245667869988337,
+      "learning_rate": 7.497854708299963e-05,
+      "loss": 2.9442,
+      "step": 5640
+    },
+    {
+      "epoch": 1.2006384676775739,
+      "grad_norm": 0.9987693058122277,
+      "learning_rate": 7.496781832566407e-05,
+      "loss": 2.9032,
+      "step": 5641
+    },
+    {
+      "epoch": 1.2008512902367652,
+      "grad_norm": 1.0289176463748375,
+      "learning_rate": 7.495708803664906e-05,
+      "loss": 2.9698,
+      "step": 5642
+    },
+    {
+      "epoch": 1.2010641127959563,
+      "grad_norm": 0.9742196698355712,
+      "learning_rate": 7.494635621661295e-05,
+      "loss": 2.9152,
+      "step": 5643
+    },
+    {
+      "epoch": 1.2012769353551476,
+      "grad_norm": 0.9858223975858353,
+      "learning_rate": 7.493562286621405e-05,
+      "loss": 2.8975,
+      "step": 5644
+    },
+    {
+      "epoch": 1.2014897579143389,
+      "grad_norm": 0.9532089107636169,
+      "learning_rate": 7.492488798611082e-05,
+      "loss": 2.8752,
+      "step": 5645
+    },
+    {
+      "epoch": 1.2017025804735302,
+      "grad_norm": 0.9900284882132869,
+      "learning_rate": 7.491415157696179e-05,
+      "loss": 2.9284,
+      "step": 5646
+    },
+    {
+      "epoch": 1.2019154030327215,
+      "grad_norm": 0.8925597506409777,
+      "learning_rate": 7.490341363942559e-05,
+      "loss": 2.9021,
+      "step": 5647
+    },
+    {
+      "epoch": 1.2021282255919128,
+      "grad_norm": 0.8359310662010928,
+      "learning_rate": 7.4892674174161e-05,
+      "loss": 2.9511,
+      "step": 5648
+    },
+    {
+      "epoch": 1.202341048151104,
+      "grad_norm": 0.928290192725012,
+      "learning_rate": 7.488193318182679e-05,
+      "loss": 2.9364,
+      "step": 5649
+    },
+    {
+      "epoch": 1.2025538707102954,
+      "grad_norm": 0.9414873875362358,
+      "learning_rate": 7.48711906630819e-05,
+      "loss": 2.9174,
+      "step": 5650
+    },
+    {
+      "epoch": 1.2027666932694865,
+      "grad_norm": 1.0239453793917097,
+      "learning_rate": 7.486044661858533e-05,
+      "loss": 2.8775,
+      "step": 5651
+    },
+    {
+      "epoch": 1.2029795158286778,
+      "grad_norm": 0.9403140331438457,
+      "learning_rate": 7.484970104899624e-05,
+      "loss": 2.8703,
+      "step": 5652
+    },
+    {
+      "epoch": 1.203192338387869,
+      "grad_norm": 0.8981177581483087,
+      "learning_rate": 7.483895395497376e-05,
+      "loss": 2.9213,
+      "step": 5653
+    },
+    {
+      "epoch": 1.2034051609470604,
+      "grad_norm": 0.9069218471655623,
+      "learning_rate": 7.48282053371772e-05,
+      "loss": 2.8962,
+      "step": 5654
+    },
+    {
+      "epoch": 1.2036179835062517,
+      "grad_norm": 0.9129047862048317,
+      "learning_rate": 7.481745519626599e-05,
+      "loss": 2.8942,
+      "step": 5655
+    },
+    {
+      "epoch": 1.203830806065443,
+      "grad_norm": 0.8800898011992594,
+      "learning_rate": 7.480670353289954e-05,
+      "loss": 2.9091,
+      "step": 5656
+    },
+    {
+      "epoch": 1.204043628624634,
+      "grad_norm": 0.7373828680984423,
+      "learning_rate": 7.479595034773748e-05,
+      "loss": 2.9458,
+      "step": 5657
+    },
+    {
+      "epoch": 1.2042564511838254,
+      "grad_norm": 0.8045752414231245,
+      "learning_rate": 7.478519564143946e-05,
+      "loss": 2.8836,
+      "step": 5658
+    },
+    {
+      "epoch": 1.2044692737430167,
+      "grad_norm": 0.8872756006724122,
+      "learning_rate": 7.477443941466522e-05,
+      "loss": 2.9148,
+      "step": 5659
+    },
+    {
+      "epoch": 1.204682096302208,
+      "grad_norm": 0.7891996707128265,
+      "learning_rate": 7.476368166807462e-05,
+      "loss": 2.911,
+      "step": 5660
+    },
+    {
+      "epoch": 1.2048949188613993,
+      "grad_norm": 0.7368455914373861,
+      "learning_rate": 7.475292240232763e-05,
+      "loss": 2.8997,
+      "step": 5661
+    },
+    {
+      "epoch": 1.2051077414205906,
+      "grad_norm": 0.8445521278726591,
+      "learning_rate": 7.47421616180843e-05,
+      "loss": 2.892,
+      "step": 5662
+    },
+    {
+      "epoch": 1.205320563979782,
+      "grad_norm": 0.7484002879135395,
+      "learning_rate": 7.473139931600469e-05,
+      "loss": 2.8969,
+      "step": 5663
+    },
+    {
+      "epoch": 1.2055333865389732,
+      "grad_norm": 0.7349246757230375,
+      "learning_rate": 7.47206354967491e-05,
+      "loss": 2.9387,
+      "step": 5664
+    },
+    {
+      "epoch": 1.2057462090981643,
+      "grad_norm": 0.8368358860489874,
+      "learning_rate": 7.470987016097781e-05,
+      "loss": 2.8922,
+      "step": 5665
+    },
+    {
+      "epoch": 1.2059590316573556,
+      "grad_norm": 0.8779702390686259,
+      "learning_rate": 7.469910330935126e-05,
+      "loss": 2.9159,
+      "step": 5666
+    },
+    {
+      "epoch": 1.206171854216547,
+      "grad_norm": 0.7582326662267823,
+      "learning_rate": 7.468833494252992e-05,
+      "loss": 2.9184,
+      "step": 5667
+    },
+    {
+      "epoch": 1.2063846767757382,
+      "grad_norm": 0.861114280646347,
+      "learning_rate": 7.467756506117441e-05,
+      "loss": 2.9112,
+      "step": 5668
+    },
+    {
+      "epoch": 1.2065974993349295,
+      "grad_norm": 0.8712197197815084,
+      "learning_rate": 7.466679366594542e-05,
+      "loss": 2.9083,
+      "step": 5669
+    },
+    {
+      "epoch": 1.2068103218941209,
+      "grad_norm": 0.8067551953549976,
+      "learning_rate": 7.465602075750373e-05,
+      "loss": 2.8613,
+      "step": 5670
+    },
+    {
+      "epoch": 1.207023144453312,
+      "grad_norm": 0.8552979492980403,
+      "learning_rate": 7.464524633651022e-05,
+      "loss": 2.9088,
+      "step": 5671
+    },
+    {
+      "epoch": 1.2072359670125032,
+      "grad_norm": 0.851656807767561,
+      "learning_rate": 7.463447040362586e-05,
+      "loss": 2.9196,
+      "step": 5672
+    },
+    {
+      "epoch": 1.2074487895716945,
+      "grad_norm": 0.8926746151386497,
+      "learning_rate": 7.462369295951169e-05,
+      "loss": 2.9375,
+      "step": 5673
+    },
+    {
+      "epoch": 1.2076616121308859,
+      "grad_norm": 0.9317053722932661,
+      "learning_rate": 7.461291400482888e-05,
+      "loss": 2.9104,
+      "step": 5674
+    },
+    {
+      "epoch": 1.2078744346900772,
+      "grad_norm": 0.9519435503046547,
+      "learning_rate": 7.460213354023871e-05,
+      "loss": 2.9282,
+      "step": 5675
+    },
+    {
+      "epoch": 1.2080872572492685,
+      "grad_norm": 0.9564334585433355,
+      "learning_rate": 7.459135156640247e-05,
+      "loss": 2.947,
+      "step": 5676
+    },
+    {
+      "epoch": 1.2083000798084598,
+      "grad_norm": 0.9490890566597154,
+      "learning_rate": 7.458056808398161e-05,
+      "loss": 2.9498,
+      "step": 5677
+    },
+    {
+      "epoch": 1.208512902367651,
+      "grad_norm": 0.9375058025144619,
+      "learning_rate": 7.456978309363766e-05,
+      "loss": 2.9364,
+      "step": 5678
+    },
+    {
+      "epoch": 1.2087257249268422,
+      "grad_norm": 0.8830077262692924,
+      "learning_rate": 7.455899659603223e-05,
+      "loss": 2.898,
+      "step": 5679
+    },
+    {
+      "epoch": 1.2089385474860335,
+      "grad_norm": 0.8173464232847227,
+      "learning_rate": 7.454820859182706e-05,
+      "loss": 2.9566,
+      "step": 5680
+    },
+    {
+      "epoch": 1.2091513700452248,
+      "grad_norm": 0.8423673547767226,
+      "learning_rate": 7.453741908168389e-05,
+      "loss": 2.9444,
+      "step": 5681
+    },
+    {
+      "epoch": 1.209364192604416,
+      "grad_norm": 0.9359133041334328,
+      "learning_rate": 7.452662806626468e-05,
+      "loss": 2.8897,
+      "step": 5682
+    },
+    {
+      "epoch": 1.2095770151636074,
+      "grad_norm": 0.8985151843536208,
+      "learning_rate": 7.451583554623137e-05,
+      "loss": 2.922,
+      "step": 5683
+    },
+    {
+      "epoch": 1.2097898377227987,
+      "grad_norm": 0.8680959470869185,
+      "learning_rate": 7.450504152224608e-05,
+      "loss": 2.9246,
+      "step": 5684
+    },
+    {
+      "epoch": 1.2100026602819898,
+      "grad_norm": 0.8925998624155842,
+      "learning_rate": 7.449424599497094e-05,
+      "loss": 2.9228,
+      "step": 5685
+    },
+    {
+      "epoch": 1.210215482841181,
+      "grad_norm": 0.95525991971377,
+      "learning_rate": 7.448344896506825e-05,
+      "loss": 2.9173,
+      "step": 5686
+    },
+    {
+      "epoch": 1.2104283054003724,
+      "grad_norm": 1.018738806454174,
+      "learning_rate": 7.447265043320036e-05,
+      "loss": 2.9525,
+      "step": 5687
+    },
+    {
+      "epoch": 1.2106411279595637,
+      "grad_norm": 1.0792186475985734,
+      "learning_rate": 7.446185040002968e-05,
+      "loss": 2.9496,
+      "step": 5688
+    },
+    {
+      "epoch": 1.210853950518755,
+      "grad_norm": 0.9002675548956026,
+      "learning_rate": 7.445104886621879e-05,
+      "loss": 2.8793,
+      "step": 5689
+    },
+    {
+      "epoch": 1.2110667730779463,
+      "grad_norm": 0.8511521432919628,
+      "learning_rate": 7.444024583243033e-05,
+      "loss": 2.9047,
+      "step": 5690
+    },
+    {
+      "epoch": 1.2112795956371376,
+      "grad_norm": 0.9551392464432824,
+      "learning_rate": 7.442944129932698e-05,
+      "loss": 2.8991,
+      "step": 5691
+    },
+    {
+      "epoch": 1.211492418196329,
+      "grad_norm": 0.9299816893174164,
+      "learning_rate": 7.44186352675716e-05,
+      "loss": 2.9121,
+      "step": 5692
+    },
+    {
+      "epoch": 1.21170524075552,
+      "grad_norm": 0.8919794738141987,
+      "learning_rate": 7.440782773782709e-05,
+      "loss": 2.9274,
+      "step": 5693
+    },
+    {
+      "epoch": 1.2119180633147113,
+      "grad_norm": 0.8684082377916814,
+      "learning_rate": 7.439701871075641e-05,
+      "loss": 2.9099,
+      "step": 5694
+    },
+    {
+      "epoch": 1.2121308858739026,
+      "grad_norm": 0.8029989516277681,
+      "learning_rate": 7.43862081870227e-05,
+      "loss": 2.8772,
+      "step": 5695
+    },
+    {
+      "epoch": 1.212343708433094,
+      "grad_norm": 0.8279002804975517,
+      "learning_rate": 7.437539616728915e-05,
+      "loss": 2.8949,
+      "step": 5696
+    },
+    {
+      "epoch": 1.2125565309922852,
+      "grad_norm": 0.8517393581843367,
+      "learning_rate": 7.436458265221899e-05,
+      "loss": 2.8977,
+      "step": 5697
+    },
+    {
+      "epoch": 1.2127693535514765,
+      "grad_norm": 0.9232326496846478,
+      "learning_rate": 7.43537676424756e-05,
+      "loss": 2.9178,
+      "step": 5698
+    },
+    {
+      "epoch": 1.2129821761106676,
+      "grad_norm": 0.9748009576076049,
+      "learning_rate": 7.434295113872247e-05,
+      "loss": 2.9182,
+      "step": 5699
+    },
+    {
+      "epoch": 1.213194998669859,
+      "grad_norm": 0.9476959471162809,
+      "learning_rate": 7.433213314162313e-05,
+      "loss": 2.8974,
+      "step": 5700
+    },
+    {
+      "epoch": 1.2134078212290502,
+      "grad_norm": 0.8782337381106722,
+      "learning_rate": 7.432131365184122e-05,
+      "loss": 2.8865,
+      "step": 5701
+    },
+    {
+      "epoch": 1.2136206437882415,
+      "grad_norm": 0.9222305527380797,
+      "learning_rate": 7.431049267004046e-05,
+      "loss": 2.8944,
+      "step": 5702
+    },
+    {
+      "epoch": 1.2138334663474328,
+      "grad_norm": 0.8892324404157792,
+      "learning_rate": 7.429967019688473e-05,
+      "loss": 2.8584,
+      "step": 5703
+    },
+    {
+      "epoch": 1.2140462889066241,
+      "grad_norm": 0.8359203444771827,
+      "learning_rate": 7.428884623303788e-05,
+      "loss": 2.9133,
+      "step": 5704
+    },
+    {
+      "epoch": 1.2142591114658154,
+      "grad_norm": 0.8348971010711247,
+      "learning_rate": 7.427802077916396e-05,
+      "loss": 2.9064,
+      "step": 5705
+    },
+    {
+      "epoch": 1.2144719340250067,
+      "grad_norm": 0.8147386272131023,
+      "learning_rate": 7.426719383592706e-05,
+      "loss": 2.8923,
+      "step": 5706
+    },
+    {
+      "epoch": 1.2146847565841978,
+      "grad_norm": 0.7306609272862328,
+      "learning_rate": 7.425636540399135e-05,
+      "loss": 2.882,
+      "step": 5707
+    },
+    {
+      "epoch": 1.2148975791433891,
+      "grad_norm": 0.7503854309445073,
+      "learning_rate": 7.424553548402116e-05,
+      "loss": 2.8975,
+      "step": 5708
+    },
+    {
+      "epoch": 1.2151104017025804,
+      "grad_norm": 0.7827480231102844,
+      "learning_rate": 7.423470407668081e-05,
+      "loss": 2.9263,
+      "step": 5709
+    },
+    {
+      "epoch": 1.2153232242617718,
+      "grad_norm": 0.8767089661310228,
+      "learning_rate": 7.422387118263482e-05,
+      "loss": 2.8646,
+      "step": 5710
+    },
+    {
+      "epoch": 1.215536046820963,
+      "grad_norm": 0.9492729219362679,
+      "learning_rate": 7.421303680254769e-05,
+      "loss": 2.9289,
+      "step": 5711
+    },
+    {
+      "epoch": 1.2157488693801544,
+      "grad_norm": 1.0026069879600172,
+      "learning_rate": 7.42022009370841e-05,
+      "loss": 2.9105,
+      "step": 5712
+    },
+    {
+      "epoch": 1.2159616919393454,
+      "grad_norm": 0.8791725300021614,
+      "learning_rate": 7.419136358690878e-05,
+      "loss": 2.8609,
+      "step": 5713
+    },
+    {
+      "epoch": 1.2161745144985368,
+      "grad_norm": 0.9300466270978262,
+      "learning_rate": 7.418052475268659e-05,
+      "loss": 2.9462,
+      "step": 5714
+    },
+    {
+      "epoch": 1.216387337057728,
+      "grad_norm": 0.8841952637504495,
+      "learning_rate": 7.41696844350824e-05,
+      "loss": 2.8717,
+      "step": 5715
+    },
+    {
+      "epoch": 1.2166001596169194,
+      "grad_norm": 0.8328967141696776,
+      "learning_rate": 7.415884263476125e-05,
+      "loss": 2.867,
+      "step": 5716
+    },
+    {
+      "epoch": 1.2168129821761107,
+      "grad_norm": 0.7839687155998951,
+      "learning_rate": 7.414799935238824e-05,
+      "loss": 2.9099,
+      "step": 5717
+    },
+    {
+      "epoch": 1.217025804735302,
+      "grad_norm": 0.8397430258802595,
+      "learning_rate": 7.413715458862856e-05,
+      "loss": 2.8797,
+      "step": 5718
+    },
+    {
+      "epoch": 1.2172386272944933,
+      "grad_norm": 0.8220652515073787,
+      "learning_rate": 7.412630834414748e-05,
+      "loss": 2.9122,
+      "step": 5719
+    },
+    {
+      "epoch": 1.2174514498536846,
+      "grad_norm": 0.7367580662524659,
+      "learning_rate": 7.411546061961041e-05,
+      "loss": 2.9004,
+      "step": 5720
+    },
+    {
+      "epoch": 1.2176642724128757,
+      "grad_norm": 0.8011419684274295,
+      "learning_rate": 7.41046114156828e-05,
+      "loss": 2.9533,
+      "step": 5721
+    },
+    {
+      "epoch": 1.217877094972067,
+      "grad_norm": 0.8438773676876783,
+      "learning_rate": 7.40937607330302e-05,
+      "loss": 2.8462,
+      "step": 5722
+    },
+    {
+      "epoch": 1.2180899175312583,
+      "grad_norm": 0.9009176375780146,
+      "learning_rate": 7.408290857231825e-05,
+      "loss": 2.9032,
+      "step": 5723
+    },
+    {
+      "epoch": 1.2183027400904496,
+      "grad_norm": 0.9730406754054,
+      "learning_rate": 7.407205493421271e-05,
+      "loss": 2.935,
+      "step": 5724
+    },
+    {
+      "epoch": 1.218515562649641,
+      "grad_norm": 0.9615790515782806,
+      "learning_rate": 7.406119981937941e-05,
+      "loss": 2.873,
+      "step": 5725
+    },
+    {
+      "epoch": 1.2187283852088322,
+      "grad_norm": 0.9385314238167638,
+      "learning_rate": 7.405034322848423e-05,
+      "loss": 2.9137,
+      "step": 5726
+    },
+    {
+      "epoch": 1.2189412077680235,
+      "grad_norm": 0.8990419223591429,
+      "learning_rate": 7.403948516219322e-05,
+      "loss": 2.8704,
+      "step": 5727
+    },
+    {
+      "epoch": 1.2191540303272146,
+      "grad_norm": 0.7900190997893566,
+      "learning_rate": 7.402862562117248e-05,
+      "loss": 2.906,
+      "step": 5728
+    },
+    {
+      "epoch": 1.219366852886406,
+      "grad_norm": 0.7403585228627363,
+      "learning_rate": 7.401776460608818e-05,
+      "loss": 2.8944,
+      "step": 5729
+    },
+    {
+      "epoch": 1.2195796754455972,
+      "grad_norm": 0.7936054493454968,
+      "learning_rate": 7.400690211760662e-05,
+      "loss": 2.8882,
+      "step": 5730
+    },
+    {
+      "epoch": 1.2197924980047885,
+      "grad_norm": 0.8004584295293679,
+      "learning_rate": 7.399603815639416e-05,
+      "loss": 2.9052,
+      "step": 5731
+    },
+    {
+      "epoch": 1.2200053205639798,
+      "grad_norm": 0.8783166919970413,
+      "learning_rate": 7.398517272311725e-05,
+      "loss": 2.8994,
+      "step": 5732
+    },
+    {
+      "epoch": 1.2202181431231711,
+      "grad_norm": 0.9371625207299059,
+      "learning_rate": 7.397430581844246e-05,
+      "loss": 2.9083,
+      "step": 5733
+    },
+    {
+      "epoch": 1.2204309656823624,
+      "grad_norm": 0.8763278507558409,
+      "learning_rate": 7.396343744303645e-05,
+      "loss": 2.881,
+      "step": 5734
+    },
+    {
+      "epoch": 1.2206437882415537,
+      "grad_norm": 0.8091785520976571,
+      "learning_rate": 7.395256759756593e-05,
+      "loss": 2.8856,
+      "step": 5735
+    },
+    {
+      "epoch": 1.2208566108007448,
+      "grad_norm": 0.870507190382192,
+      "learning_rate": 7.394169628269772e-05,
+      "loss": 2.8982,
+      "step": 5736
+    },
+    {
+      "epoch": 1.2210694333599361,
+      "grad_norm": 0.970124816275725,
+      "learning_rate": 7.393082349909872e-05,
+      "loss": 2.9012,
+      "step": 5737
+    },
+    {
+      "epoch": 1.2212822559191274,
+      "grad_norm": 0.9426911630239412,
+      "learning_rate": 7.391994924743597e-05,
+      "loss": 2.8932,
+      "step": 5738
+    },
+    {
+      "epoch": 1.2214950784783187,
+      "grad_norm": 0.8197958515789061,
+      "learning_rate": 7.390907352837654e-05,
+      "loss": 2.9017,
+      "step": 5739
+    },
+    {
+      "epoch": 1.22170790103751,
+      "grad_norm": 0.7358228497187347,
+      "learning_rate": 7.389819634258764e-05,
+      "loss": 2.9186,
+      "step": 5740
+    },
+    {
+      "epoch": 1.2219207235967013,
+      "grad_norm": 0.7691851648424858,
+      "learning_rate": 7.38873176907365e-05,
+      "loss": 2.9451,
+      "step": 5741
+    },
+    {
+      "epoch": 1.2221335461558924,
+      "grad_norm": 0.7888806710195416,
+      "learning_rate": 7.387643757349053e-05,
+      "loss": 2.9266,
+      "step": 5742
+    },
+    {
+      "epoch": 1.2223463687150837,
+      "grad_norm": 0.800769940121225,
+      "learning_rate": 7.386555599151712e-05,
+      "loss": 2.8768,
+      "step": 5743
+    },
+    {
+      "epoch": 1.222559191274275,
+      "grad_norm": 0.9265528755412326,
+      "learning_rate": 7.38546729454839e-05,
+      "loss": 2.8881,
+      "step": 5744
+    },
+    {
+      "epoch": 1.2227720138334663,
+      "grad_norm": 1.0837695669648724,
+      "learning_rate": 7.384378843605843e-05,
+      "loss": 2.9278,
+      "step": 5745
+    },
+    {
+      "epoch": 1.2229848363926576,
+      "grad_norm": 0.9667251585471287,
+      "learning_rate": 7.383290246390845e-05,
+      "loss": 2.8814,
+      "step": 5746
+    },
+    {
+      "epoch": 1.223197658951849,
+      "grad_norm": 0.9578089543942648,
+      "learning_rate": 7.38220150297018e-05,
+      "loss": 2.923,
+      "step": 5747
+    },
+    {
+      "epoch": 1.2234104815110403,
+      "grad_norm": 1.0516112047811086,
+      "learning_rate": 7.381112613410636e-05,
+      "loss": 2.9325,
+      "step": 5748
+    },
+    {
+      "epoch": 1.2236233040702316,
+      "grad_norm": 0.8363169112983063,
+      "learning_rate": 7.380023577779012e-05,
+      "loss": 2.9096,
+      "step": 5749
+    },
+    {
+      "epoch": 1.2238361266294226,
+      "grad_norm": 0.9497389650087045,
+      "learning_rate": 7.378934396142115e-05,
+      "loss": 2.904,
+      "step": 5750
+    },
+    {
+      "epoch": 1.224048949188614,
+      "grad_norm": 1.0275865083993418,
+      "learning_rate": 7.377845068566767e-05,
+      "loss": 2.8831,
+      "step": 5751
+    },
+    {
+      "epoch": 1.2242617717478053,
+      "grad_norm": 1.059437614456302,
+      "learning_rate": 7.376755595119789e-05,
+      "loss": 2.9204,
+      "step": 5752
+    },
+    {
+      "epoch": 1.2244745943069966,
+      "grad_norm": 0.9289648617801914,
+      "learning_rate": 7.375665975868017e-05,
+      "loss": 2.9144,
+      "step": 5753
+    },
+    {
+      "epoch": 1.2246874168661879,
+      "grad_norm": 0.9450717868310298,
+      "learning_rate": 7.374576210878297e-05,
+      "loss": 2.8827,
+      "step": 5754
+    },
+    {
+      "epoch": 1.2249002394253792,
+      "grad_norm": 0.9143776382964741,
+      "learning_rate": 7.37348630021748e-05,
+      "loss": 2.8981,
+      "step": 5755
+    },
+    {
+      "epoch": 1.2251130619845703,
+      "grad_norm": 0.8589849486424641,
+      "learning_rate": 7.372396243952429e-05,
+      "loss": 2.8932,
+      "step": 5756
+    },
+    {
+      "epoch": 1.2253258845437616,
+      "grad_norm": 0.8691305679629203,
+      "learning_rate": 7.371306042150012e-05,
+      "loss": 2.8812,
+      "step": 5757
+    },
+    {
+      "epoch": 1.2255387071029529,
+      "grad_norm": 0.851388747353653,
+      "learning_rate": 7.370215694877114e-05,
+      "loss": 2.9394,
+      "step": 5758
+    },
+    {
+      "epoch": 1.2257515296621442,
+      "grad_norm": 0.8344553596336082,
+      "learning_rate": 7.369125202200618e-05,
+      "loss": 2.9318,
+      "step": 5759
+    },
+    {
+      "epoch": 1.2259643522213355,
+      "grad_norm": 0.8576433307098433,
+      "learning_rate": 7.368034564187425e-05,
+      "loss": 2.8849,
+      "step": 5760
+    },
+    {
+      "epoch": 1.2261771747805268,
+      "grad_norm": 0.886124280486067,
+      "learning_rate": 7.36694378090444e-05,
+      "loss": 2.9362,
+      "step": 5761
+    },
+    {
+      "epoch": 1.226389997339718,
+      "grad_norm": 0.9600530663155059,
+      "learning_rate": 7.36585285241858e-05,
+      "loss": 2.9097,
+      "step": 5762
+    },
+    {
+      "epoch": 1.2266028198989094,
+      "grad_norm": 1.0895216352905752,
+      "learning_rate": 7.364761778796766e-05,
+      "loss": 2.9202,
+      "step": 5763
+    },
+    {
+      "epoch": 1.2268156424581005,
+      "grad_norm": 1.0082923357411024,
+      "learning_rate": 7.363670560105934e-05,
+      "loss": 2.9054,
+      "step": 5764
+    },
+    {
+      "epoch": 1.2270284650172918,
+      "grad_norm": 0.9242692372680255,
+      "learning_rate": 7.362579196413025e-05,
+      "loss": 2.8909,
+      "step": 5765
+    },
+    {
+      "epoch": 1.227241287576483,
+      "grad_norm": 0.853465376391155,
+      "learning_rate": 7.361487687784989e-05,
+      "loss": 2.9379,
+      "step": 5766
+    },
+    {
+      "epoch": 1.2274541101356744,
+      "grad_norm": 0.8655133311675459,
+      "learning_rate": 7.360396034288789e-05,
+      "loss": 2.8831,
+      "step": 5767
+    },
+    {
+      "epoch": 1.2276669326948657,
+      "grad_norm": 0.8258831951405713,
+      "learning_rate": 7.359304235991391e-05,
+      "loss": 2.8949,
+      "step": 5768
+    },
+    {
+      "epoch": 1.227879755254057,
+      "grad_norm": 0.7639269649824568,
+      "learning_rate": 7.358212292959774e-05,
+      "loss": 2.9049,
+      "step": 5769
+    },
+    {
+      "epoch": 1.228092577813248,
+      "grad_norm": 0.8210256043720014,
+      "learning_rate": 7.357120205260924e-05,
+      "loss": 2.9149,
+      "step": 5770
+    },
+    {
+      "epoch": 1.2283054003724394,
+      "grad_norm": 0.8372637916257544,
+      "learning_rate": 7.356027972961836e-05,
+      "loss": 2.8757,
+      "step": 5771
+    },
+    {
+      "epoch": 1.2285182229316307,
+      "grad_norm": 0.7429089463319632,
+      "learning_rate": 7.354935596129513e-05,
+      "loss": 2.9054,
+      "step": 5772
+    },
+    {
+      "epoch": 1.228731045490822,
+      "grad_norm": 0.7801285538228605,
+      "learning_rate": 7.35384307483097e-05,
+      "loss": 2.8795,
+      "step": 5773
+    },
+    {
+      "epoch": 1.2289438680500133,
+      "grad_norm": 0.8074251240043256,
+      "learning_rate": 7.35275040913323e-05,
+      "loss": 2.9109,
+      "step": 5774
+    },
+    {
+      "epoch": 1.2291566906092046,
+      "grad_norm": 0.8474471526557289,
+      "learning_rate": 7.351657599103321e-05,
+      "loss": 2.8749,
+      "step": 5775
+    },
+    {
+      "epoch": 1.229369513168396,
+      "grad_norm": 0.7880222709816622,
+      "learning_rate": 7.350564644808284e-05,
+      "loss": 2.9048,
+      "step": 5776
+    },
+    {
+      "epoch": 1.2295823357275872,
+      "grad_norm": 0.7577532410495811,
+      "learning_rate": 7.349471546315168e-05,
+      "loss": 2.9448,
+      "step": 5777
+    },
+    {
+      "epoch": 1.2297951582867783,
+      "grad_norm": 0.8325333650181729,
+      "learning_rate": 7.34837830369103e-05,
+      "loss": 2.8681,
+      "step": 5778
+    },
+    {
+      "epoch": 1.2300079808459696,
+      "grad_norm": 0.8864433262556378,
+      "learning_rate": 7.347284917002937e-05,
+      "loss": 2.9105,
+      "step": 5779
+    },
+    {
+      "epoch": 1.230220803405161,
+      "grad_norm": 0.9659416545603916,
+      "learning_rate": 7.346191386317961e-05,
+      "loss": 2.9138,
+      "step": 5780
+    },
+    {
+      "epoch": 1.2304336259643522,
+      "grad_norm": 1.003992533759489,
+      "learning_rate": 7.345097711703191e-05,
+      "loss": 2.907,
+      "step": 5781
+    },
+    {
+      "epoch": 1.2306464485235435,
+      "grad_norm": 0.854684999260696,
+      "learning_rate": 7.344003893225715e-05,
+      "loss": 2.8961,
+      "step": 5782
+    },
+    {
+      "epoch": 1.2308592710827349,
+      "grad_norm": 0.9278243018929903,
+      "learning_rate": 7.342909930952636e-05,
+      "loss": 2.8917,
+      "step": 5783
+    },
+    {
+      "epoch": 1.231072093641926,
+      "grad_norm": 0.9070405817492188,
+      "learning_rate": 7.341815824951066e-05,
+      "loss": 2.907,
+      "step": 5784
+    },
+    {
+      "epoch": 1.2312849162011172,
+      "grad_norm": 0.9524001916454856,
+      "learning_rate": 7.340721575288123e-05,
+      "loss": 2.8934,
+      "step": 5785
+    },
+    {
+      "epoch": 1.2314977387603085,
+      "grad_norm": 0.9527973639885471,
+      "learning_rate": 7.339627182030934e-05,
+      "loss": 2.8827,
+      "step": 5786
+    },
+    {
+      "epoch": 1.2317105613194999,
+      "grad_norm": 0.9695086204774449,
+      "learning_rate": 7.338532645246637e-05,
+      "loss": 2.9083,
+      "step": 5787
+    },
+    {
+      "epoch": 1.2319233838786912,
+      "grad_norm": 0.9621161977181427,
+      "learning_rate": 7.337437965002379e-05,
+      "loss": 2.9105,
+      "step": 5788
+    },
+    {
+      "epoch": 1.2321362064378825,
+      "grad_norm": 1.053183307511502,
+      "learning_rate": 7.33634314136531e-05,
+      "loss": 2.8894,
+      "step": 5789
+    },
+    {
+      "epoch": 1.2323490289970738,
+      "grad_norm": 0.8619167893677281,
+      "learning_rate": 7.335248174402598e-05,
+      "loss": 2.9237,
+      "step": 5790
+    },
+    {
+      "epoch": 1.232561851556265,
+      "grad_norm": 0.7717845742374541,
+      "learning_rate": 7.334153064181413e-05,
+      "loss": 2.9018,
+      "step": 5791
+    },
+    {
+      "epoch": 1.2327746741154562,
+      "grad_norm": 0.9932375937562833,
+      "learning_rate": 7.333057810768934e-05,
+      "loss": 2.8645,
+      "step": 5792
+    },
+    {
+      "epoch": 1.2329874966746475,
+      "grad_norm": 1.0052474271407095,
+      "learning_rate": 7.331962414232354e-05,
+      "loss": 2.9144,
+      "step": 5793
+    },
+    {
+      "epoch": 1.2332003192338388,
+      "grad_norm": 1.0076824582872017,
+      "learning_rate": 7.330866874638869e-05,
+      "loss": 2.9177,
+      "step": 5794
+    },
+    {
+      "epoch": 1.23341314179303,
+      "grad_norm": 0.9532960735337511,
+      "learning_rate": 7.329771192055688e-05,
+      "loss": 2.8834,
+      "step": 5795
+    },
+    {
+      "epoch": 1.2336259643522214,
+      "grad_norm": 0.9308694264812548,
+      "learning_rate": 7.328675366550024e-05,
+      "loss": 2.8712,
+      "step": 5796
+    },
+    {
+      "epoch": 1.2338387869114127,
+      "grad_norm": 0.82204247794824,
+      "learning_rate": 7.327579398189104e-05,
+      "loss": 2.8665,
+      "step": 5797
+    },
+    {
+      "epoch": 1.2340516094706038,
+      "grad_norm": 1.0020727793779443,
+      "learning_rate": 7.326483287040162e-05,
+      "loss": 2.8934,
+      "step": 5798
+    },
+    {
+      "epoch": 1.234264432029795,
+      "grad_norm": 1.2062460994275936,
+      "learning_rate": 7.325387033170438e-05,
+      "loss": 2.9213,
+      "step": 5799
+    },
+    {
+      "epoch": 1.2344772545889864,
+      "grad_norm": 0.7777427315215187,
+      "learning_rate": 7.324290636647184e-05,
+      "loss": 2.9396,
+      "step": 5800
+    },
+    {
+      "epoch": 1.2346900771481777,
+      "grad_norm": 0.8221683824140842,
+      "learning_rate": 7.323194097537659e-05,
+      "loss": 2.9412,
+      "step": 5801
+    },
+    {
+      "epoch": 1.234902899707369,
+      "grad_norm": 0.7769330517233224,
+      "learning_rate": 7.322097415909133e-05,
+      "loss": 2.9175,
+      "step": 5802
+    },
+    {
+      "epoch": 1.2351157222665603,
+      "grad_norm": 0.7754583966916861,
+      "learning_rate": 7.321000591828882e-05,
+      "loss": 2.8942,
+      "step": 5803
+    },
+    {
+      "epoch": 1.2353285448257516,
+      "grad_norm": 0.8587671780152419,
+      "learning_rate": 7.319903625364192e-05,
+      "loss": 2.9009,
+      "step": 5804
+    },
+    {
+      "epoch": 1.235541367384943,
+      "grad_norm": 0.9162798503462298,
+      "learning_rate": 7.318806516582359e-05,
+      "loss": 2.8736,
+      "step": 5805
+    },
+    {
+      "epoch": 1.235754189944134,
+      "grad_norm": 0.9234398412806117,
+      "learning_rate": 7.317709265550685e-05,
+      "loss": 2.8796,
+      "step": 5806
+    },
+    {
+      "epoch": 1.2359670125033253,
+      "grad_norm": 0.8526454987707259,
+      "learning_rate": 7.316611872336484e-05,
+      "loss": 2.8934,
+      "step": 5807
+    },
+    {
+      "epoch": 1.2361798350625166,
+      "grad_norm": 0.8275245083327185,
+      "learning_rate": 7.315514337007071e-05,
+      "loss": 2.9311,
+      "step": 5808
+    },
+    {
+      "epoch": 1.236392657621708,
+      "grad_norm": 0.8669516159603449,
+      "learning_rate": 7.314416659629784e-05,
+      "loss": 2.9225,
+      "step": 5809
+    },
+    {
+      "epoch": 1.2366054801808992,
+      "grad_norm": 0.900890435563352,
+      "learning_rate": 7.313318840271954e-05,
+      "loss": 2.9165,
+      "step": 5810
+    },
+    {
+      "epoch": 1.2368183027400905,
+      "grad_norm": 0.9525961095477032,
+      "learning_rate": 7.312220879000933e-05,
+      "loss": 2.9126,
+      "step": 5811
+    },
+    {
+      "epoch": 1.2370311252992816,
+      "grad_norm": 0.9527355077191314,
+      "learning_rate": 7.311122775884074e-05,
+      "loss": 2.8812,
+      "step": 5812
+    },
+    {
+      "epoch": 1.237243947858473,
+      "grad_norm": 0.9137915268768042,
+      "learning_rate": 7.310024530988744e-05,
+      "loss": 2.9399,
+      "step": 5813
+    },
+    {
+      "epoch": 1.2374567704176642,
+      "grad_norm": 0.8868020755591409,
+      "learning_rate": 7.308926144382312e-05,
+      "loss": 2.8911,
+      "step": 5814
+    },
+    {
+      "epoch": 1.2376695929768555,
+      "grad_norm": 0.9822670946060003,
+      "learning_rate": 7.307827616132163e-05,
+      "loss": 2.8774,
+      "step": 5815
+    },
+    {
+      "epoch": 1.2378824155360468,
+      "grad_norm": 1.036169731912602,
+      "learning_rate": 7.306728946305687e-05,
+      "loss": 2.913,
+      "step": 5816
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 0.9679995021503693,
+      "learning_rate": 7.305630134970281e-05,
+      "loss": 2.8755,
+      "step": 5817
+    },
+    {
+      "epoch": 1.2383080606544294,
+      "grad_norm": 0.930384011158455,
+      "learning_rate": 7.304531182193357e-05,
+      "loss": 2.9113,
+      "step": 5818
+    },
+    {
+      "epoch": 1.2385208832136207,
+      "grad_norm": 1.038467197822583,
+      "learning_rate": 7.303432088042327e-05,
+      "loss": 2.9043,
+      "step": 5819
+    },
+    {
+      "epoch": 1.2387337057728118,
+      "grad_norm": 0.8310209013036196,
+      "learning_rate": 7.302332852584618e-05,
+      "loss": 2.8753,
+      "step": 5820
+    },
+    {
+      "epoch": 1.2389465283320031,
+      "grad_norm": 0.8543759542937989,
+      "learning_rate": 7.301233475887666e-05,
+      "loss": 2.9189,
+      "step": 5821
+    },
+    {
+      "epoch": 1.2391593508911944,
+      "grad_norm": 1.027109562574985,
+      "learning_rate": 7.300133958018911e-05,
+      "loss": 2.9068,
+      "step": 5822
+    },
+    {
+      "epoch": 1.2393721734503858,
+      "grad_norm": 0.9012927393493684,
+      "learning_rate": 7.299034299045805e-05,
+      "loss": 2.9049,
+      "step": 5823
+    },
+    {
+      "epoch": 1.239584996009577,
+      "grad_norm": 0.9086609054282317,
+      "learning_rate": 7.297934499035805e-05,
+      "loss": 2.923,
+      "step": 5824
+    },
+    {
+      "epoch": 1.2397978185687684,
+      "grad_norm": 0.998141652390602,
+      "learning_rate": 7.296834558056386e-05,
+      "loss": 2.9419,
+      "step": 5825
+    },
+    {
+      "epoch": 1.2400106411279594,
+      "grad_norm": 0.90259900413139,
+      "learning_rate": 7.29573447617502e-05,
+      "loss": 2.8904,
+      "step": 5826
+    },
+    {
+      "epoch": 1.2402234636871508,
+      "grad_norm": 0.943875034970832,
+      "learning_rate": 7.294634253459192e-05,
+      "loss": 2.8643,
+      "step": 5827
+    },
+    {
+      "epoch": 1.240436286246342,
+      "grad_norm": 0.8807728607382121,
+      "learning_rate": 7.293533889976401e-05,
+      "loss": 2.9126,
+      "step": 5828
+    },
+    {
+      "epoch": 1.2406491088055334,
+      "grad_norm": 0.7653774953088552,
+      "learning_rate": 7.292433385794146e-05,
+      "loss": 2.9087,
+      "step": 5829
+    },
+    {
+      "epoch": 1.2408619313647247,
+      "grad_norm": 0.8160052248686912,
+      "learning_rate": 7.291332740979941e-05,
+      "loss": 2.919,
+      "step": 5830
+    },
+    {
+      "epoch": 1.241074753923916,
+      "grad_norm": 0.973975993780239,
+      "learning_rate": 7.290231955601305e-05,
+      "loss": 2.8995,
+      "step": 5831
+    },
+    {
+      "epoch": 1.2412875764831073,
+      "grad_norm": 0.9932641887735802,
+      "learning_rate": 7.289131029725769e-05,
+      "loss": 2.8737,
+      "step": 5832
+    },
+    {
+      "epoch": 1.2415003990422986,
+      "grad_norm": 0.983192018877059,
+      "learning_rate": 7.288029963420867e-05,
+      "loss": 2.916,
+      "step": 5833
+    },
+    {
+      "epoch": 1.2417132216014897,
+      "grad_norm": 0.9966489272868515,
+      "learning_rate": 7.286928756754147e-05,
+      "loss": 2.8837,
+      "step": 5834
+    },
+    {
+      "epoch": 1.241926044160681,
+      "grad_norm": 1.1082855937074734,
+      "learning_rate": 7.285827409793166e-05,
+      "loss": 2.9027,
+      "step": 5835
+    },
+    {
+      "epoch": 1.2421388667198723,
+      "grad_norm": 0.8166881993138766,
+      "learning_rate": 7.284725922605484e-05,
+      "loss": 2.868,
+      "step": 5836
+    },
+    {
+      "epoch": 1.2423516892790636,
+      "grad_norm": 0.7991332187874967,
+      "learning_rate": 7.283624295258675e-05,
+      "loss": 2.9177,
+      "step": 5837
+    },
+    {
+      "epoch": 1.242564511838255,
+      "grad_norm": 0.9832831697103097,
+      "learning_rate": 7.282522527820319e-05,
+      "loss": 2.9355,
+      "step": 5838
+    },
+    {
+      "epoch": 1.2427773343974462,
+      "grad_norm": 1.0304165415854676,
+      "learning_rate": 7.281420620358006e-05,
+      "loss": 2.8934,
+      "step": 5839
+    },
+    {
+      "epoch": 1.2429901569566373,
+      "grad_norm": 0.8830680687816133,
+      "learning_rate": 7.28031857293933e-05,
+      "loss": 2.9385,
+      "step": 5840
+    },
+    {
+      "epoch": 1.2432029795158286,
+      "grad_norm": 0.9238446923396759,
+      "learning_rate": 7.279216385631903e-05,
+      "loss": 2.8857,
+      "step": 5841
+    },
+    {
+      "epoch": 1.24341580207502,
+      "grad_norm": 1.0307100290458122,
+      "learning_rate": 7.278114058503337e-05,
+      "loss": 2.94,
+      "step": 5842
+    },
+    {
+      "epoch": 1.2436286246342112,
+      "grad_norm": 0.9868061454042524,
+      "learning_rate": 7.277011591621254e-05,
+      "loss": 2.9031,
+      "step": 5843
+    },
+    {
+      "epoch": 1.2438414471934025,
+      "grad_norm": 1.0818035554783574,
+      "learning_rate": 7.27590898505329e-05,
+      "loss": 2.8959,
+      "step": 5844
+    },
+    {
+      "epoch": 1.2440542697525938,
+      "grad_norm": 0.8348236860673863,
+      "learning_rate": 7.274806238867082e-05,
+      "loss": 2.8944,
+      "step": 5845
+    },
+    {
+      "epoch": 1.2442670923117851,
+      "grad_norm": 0.7623380484660086,
+      "learning_rate": 7.273703353130282e-05,
+      "loss": 2.9131,
+      "step": 5846
+    },
+    {
+      "epoch": 1.2444799148709764,
+      "grad_norm": 0.9663300673611739,
+      "learning_rate": 7.272600327910544e-05,
+      "loss": 2.9143,
+      "step": 5847
+    },
+    {
+      "epoch": 1.2446927374301675,
+      "grad_norm": 0.9543363656099371,
+      "learning_rate": 7.27149716327554e-05,
+      "loss": 2.9025,
+      "step": 5848
+    },
+    {
+      "epoch": 1.2449055599893588,
+      "grad_norm": 0.8746510929175788,
+      "learning_rate": 7.270393859292938e-05,
+      "loss": 2.907,
+      "step": 5849
+    },
+    {
+      "epoch": 1.2451183825485501,
+      "grad_norm": 0.8444147228217593,
+      "learning_rate": 7.26929041603043e-05,
+      "loss": 2.9429,
+      "step": 5850
+    },
+    {
+      "epoch": 1.2453312051077414,
+      "grad_norm": 1.0071531033539936,
+      "learning_rate": 7.268186833555701e-05,
+      "loss": 2.8611,
+      "step": 5851
+    },
+    {
+      "epoch": 1.2455440276669327,
+      "grad_norm": 1.0763217554067714,
+      "learning_rate": 7.267083111936453e-05,
+      "loss": 2.8892,
+      "step": 5852
+    },
+    {
+      "epoch": 1.245756850226124,
+      "grad_norm": 0.7843330065468113,
+      "learning_rate": 7.265979251240397e-05,
+      "loss": 2.9238,
+      "step": 5853
+    },
+    {
+      "epoch": 1.2459696727853151,
+      "grad_norm": 0.8470371220744883,
+      "learning_rate": 7.264875251535248e-05,
+      "loss": 2.9134,
+      "step": 5854
+    },
+    {
+      "epoch": 1.2461824953445064,
+      "grad_norm": 0.9465019841283392,
+      "learning_rate": 7.263771112888735e-05,
+      "loss": 2.9181,
+      "step": 5855
+    },
+    {
+      "epoch": 1.2463953179036977,
+      "grad_norm": 0.8468509360169725,
+      "learning_rate": 7.26266683536859e-05,
+      "loss": 2.9262,
+      "step": 5856
+    },
+    {
+      "epoch": 1.246608140462889,
+      "grad_norm": 0.8323331740414041,
+      "learning_rate": 7.261562419042558e-05,
+      "loss": 2.8823,
+      "step": 5857
+    },
+    {
+      "epoch": 1.2468209630220803,
+      "grad_norm": 0.8193404158817255,
+      "learning_rate": 7.260457863978389e-05,
+      "loss": 2.888,
+      "step": 5858
+    },
+    {
+      "epoch": 1.2470337855812716,
+      "grad_norm": 0.8833293577986313,
+      "learning_rate": 7.259353170243844e-05,
+      "loss": 2.8901,
+      "step": 5859
+    },
+    {
+      "epoch": 1.247246608140463,
+      "grad_norm": 1.002352916576254,
+      "learning_rate": 7.258248337906695e-05,
+      "loss": 2.9434,
+      "step": 5860
+    },
+    {
+      "epoch": 1.2474594306996543,
+      "grad_norm": 0.9345342730733198,
+      "learning_rate": 7.257143367034712e-05,
+      "loss": 2.9338,
+      "step": 5861
+    },
+    {
+      "epoch": 1.2476722532588453,
+      "grad_norm": 0.9270516842376634,
+      "learning_rate": 7.256038257695687e-05,
+      "loss": 2.9237,
+      "step": 5862
+    },
+    {
+      "epoch": 1.2478850758180366,
+      "grad_norm": 1.0115709057019895,
+      "learning_rate": 7.254933009957413e-05,
+      "loss": 2.8825,
+      "step": 5863
+    },
+    {
+      "epoch": 1.248097898377228,
+      "grad_norm": 0.9376995752275987,
+      "learning_rate": 7.25382762388769e-05,
+      "loss": 2.9149,
+      "step": 5864
+    },
+    {
+      "epoch": 1.2483107209364193,
+      "grad_norm": 0.8961684734694363,
+      "learning_rate": 7.252722099554331e-05,
+      "loss": 2.9002,
+      "step": 5865
+    },
+    {
+      "epoch": 1.2485235434956106,
+      "grad_norm": 0.8838540878751335,
+      "learning_rate": 7.251616437025156e-05,
+      "loss": 2.9206,
+      "step": 5866
+    },
+    {
+      "epoch": 1.2487363660548019,
+      "grad_norm": 0.8478125453590863,
+      "learning_rate": 7.250510636367993e-05,
+      "loss": 2.8816,
+      "step": 5867
+    },
+    {
+      "epoch": 1.2489491886139932,
+      "grad_norm": 0.852742863168413,
+      "learning_rate": 7.249404697650678e-05,
+      "loss": 2.9031,
+      "step": 5868
+    },
+    {
+      "epoch": 1.2491620111731843,
+      "grad_norm": 0.8242407319470206,
+      "learning_rate": 7.248298620941057e-05,
+      "loss": 2.9023,
+      "step": 5869
+    },
+    {
+      "epoch": 1.2493748337323756,
+      "grad_norm": 0.8942539010068692,
+      "learning_rate": 7.247192406306981e-05,
+      "loss": 2.8757,
+      "step": 5870
+    },
+    {
+      "epoch": 1.2495876562915669,
+      "grad_norm": 0.9346320422085475,
+      "learning_rate": 7.246086053816316e-05,
+      "loss": 2.9068,
+      "step": 5871
+    },
+    {
+      "epoch": 1.2498004788507582,
+      "grad_norm": 1.025932791244546,
+      "learning_rate": 7.244979563536929e-05,
+      "loss": 2.9395,
+      "step": 5872
+    },
+    {
+      "epoch": 1.2500133014099495,
+      "grad_norm": 0.935706767611859,
+      "learning_rate": 7.243872935536702e-05,
+      "loss": 2.8732,
+      "step": 5873
+    },
+    {
+      "epoch": 1.2502261239691408,
+      "grad_norm": 0.8835605985395366,
+      "learning_rate": 7.242766169883519e-05,
+      "loss": 2.9409,
+      "step": 5874
+    },
+    {
+      "epoch": 1.250438946528332,
+      "grad_norm": 0.864152766319211,
+      "learning_rate": 7.241659266645277e-05,
+      "loss": 2.8855,
+      "step": 5875
+    },
+    {
+      "epoch": 1.2506517690875234,
+      "grad_norm": 0.9467918222299329,
+      "learning_rate": 7.240552225889882e-05,
+      "loss": 2.9255,
+      "step": 5876
+    },
+    {
+      "epoch": 1.2508645916467145,
+      "grad_norm": 0.9212558471242682,
+      "learning_rate": 7.239445047685245e-05,
+      "loss": 2.9171,
+      "step": 5877
+    },
+    {
+      "epoch": 1.2510774142059058,
+      "grad_norm": 0.8146325637949544,
+      "learning_rate": 7.238337732099288e-05,
+      "loss": 2.8959,
+      "step": 5878
+    },
+    {
+      "epoch": 1.251290236765097,
+      "grad_norm": 0.7532985230981948,
+      "learning_rate": 7.237230279199938e-05,
+      "loss": 2.9117,
+      "step": 5879
+    },
+    {
+      "epoch": 1.2515030593242884,
+      "grad_norm": 0.7218814927060226,
+      "learning_rate": 7.236122689055138e-05,
+      "loss": 2.8637,
+      "step": 5880
+    },
+    {
+      "epoch": 1.2517158818834797,
+      "grad_norm": 0.7840309102620475,
+      "learning_rate": 7.235014961732829e-05,
+      "loss": 2.8951,
+      "step": 5881
+    },
+    {
+      "epoch": 1.2519287044426708,
+      "grad_norm": 0.8502531910617268,
+      "learning_rate": 7.23390709730097e-05,
+      "loss": 2.8788,
+      "step": 5882
+    },
+    {
+      "epoch": 1.252141527001862,
+      "grad_norm": 0.8630055382240117,
+      "learning_rate": 7.232799095827521e-05,
+      "loss": 2.8947,
+      "step": 5883
+    },
+    {
+      "epoch": 1.2523543495610534,
+      "grad_norm": 0.892477643753136,
+      "learning_rate": 7.231690957380454e-05,
+      "loss": 2.8769,
+      "step": 5884
+    },
+    {
+      "epoch": 1.2525671721202447,
+      "grad_norm": 0.8500349903723373,
+      "learning_rate": 7.230582682027753e-05,
+      "loss": 2.8912,
+      "step": 5885
+    },
+    {
+      "epoch": 1.252779994679436,
+      "grad_norm": 0.7588758626440041,
+      "learning_rate": 7.2294742698374e-05,
+      "loss": 2.9079,
+      "step": 5886
+    },
+    {
+      "epoch": 1.2529928172386273,
+      "grad_norm": 0.8667857917374637,
+      "learning_rate": 7.228365720877399e-05,
+      "loss": 2.8559,
+      "step": 5887
+    },
+    {
+      "epoch": 1.2532056397978186,
+      "grad_norm": 0.9317031452319816,
+      "learning_rate": 7.227257035215749e-05,
+      "loss": 2.9162,
+      "step": 5888
+    },
+    {
+      "epoch": 1.25341846235701,
+      "grad_norm": 0.9178044377740133,
+      "learning_rate": 7.226148212920467e-05,
+      "loss": 2.9133,
+      "step": 5889
+    },
+    {
+      "epoch": 1.2536312849162012,
+      "grad_norm": 0.9707028527485106,
+      "learning_rate": 7.225039254059574e-05,
+      "loss": 2.8968,
+      "step": 5890
+    },
+    {
+      "epoch": 1.2538441074753923,
+      "grad_norm": 1.0204973887071191,
+      "learning_rate": 7.2239301587011e-05,
+      "loss": 2.9136,
+      "step": 5891
+    },
+    {
+      "epoch": 1.2540569300345836,
+      "grad_norm": 0.9248915586236025,
+      "learning_rate": 7.222820926913085e-05,
+      "loss": 2.8968,
+      "step": 5892
+    },
+    {
+      "epoch": 1.254269752593775,
+      "grad_norm": 0.9951269551298589,
+      "learning_rate": 7.221711558763573e-05,
+      "loss": 2.9163,
+      "step": 5893
+    },
+    {
+      "epoch": 1.2544825751529662,
+      "grad_norm": 0.9270292772249541,
+      "learning_rate": 7.220602054320625e-05,
+      "loss": 2.8812,
+      "step": 5894
+    },
+    {
+      "epoch": 1.2546953977121575,
+      "grad_norm": 0.7279712168195077,
+      "learning_rate": 7.2194924136523e-05,
+      "loss": 2.9115,
+      "step": 5895
+    },
+    {
+      "epoch": 1.2549082202713486,
+      "grad_norm": 0.8384884977954169,
+      "learning_rate": 7.21838263682667e-05,
+      "loss": 2.8891,
+      "step": 5896
+    },
+    {
+      "epoch": 1.25512104283054,
+      "grad_norm": 0.9402310992556464,
+      "learning_rate": 7.217272723911818e-05,
+      "loss": 2.8811,
+      "step": 5897
+    },
+    {
+      "epoch": 1.2553338653897312,
+      "grad_norm": 0.8576082969290322,
+      "learning_rate": 7.216162674975834e-05,
+      "loss": 2.9249,
+      "step": 5898
+    },
+    {
+      "epoch": 1.2555466879489225,
+      "grad_norm": 0.8667932176070541,
+      "learning_rate": 7.215052490086812e-05,
+      "loss": 2.9529,
+      "step": 5899
+    },
+    {
+      "epoch": 1.2557595105081139,
+      "grad_norm": 1.0189046813167046,
+      "learning_rate": 7.213942169312859e-05,
+      "loss": 2.89,
+      "step": 5900
+    },
+    {
+      "epoch": 1.2559723330673052,
+      "grad_norm": 0.9597548215700917,
+      "learning_rate": 7.212831712722088e-05,
+      "loss": 2.887,
+      "step": 5901
+    },
+    {
+      "epoch": 1.2561851556264965,
+      "grad_norm": 0.9070879748194965,
+      "learning_rate": 7.211721120382623e-05,
+      "loss": 2.9136,
+      "step": 5902
+    },
+    {
+      "epoch": 1.2563979781856878,
+      "grad_norm": 0.8119650111388,
+      "learning_rate": 7.210610392362593e-05,
+      "loss": 2.8685,
+      "step": 5903
+    },
+    {
+      "epoch": 1.256610800744879,
+      "grad_norm": 0.6989780631484794,
+      "learning_rate": 7.209499528730138e-05,
+      "loss": 2.8903,
+      "step": 5904
+    },
+    {
+      "epoch": 1.2568236233040702,
+      "grad_norm": 0.8166561073425321,
+      "learning_rate": 7.208388529553405e-05,
+      "loss": 2.898,
+      "step": 5905
+    },
+    {
+      "epoch": 1.2570364458632615,
+      "grad_norm": 0.9911121296559893,
+      "learning_rate": 7.207277394900548e-05,
+      "loss": 2.9214,
+      "step": 5906
+    },
+    {
+      "epoch": 1.2572492684224528,
+      "grad_norm": 1.017500315042128,
+      "learning_rate": 7.206166124839732e-05,
+      "loss": 2.8986,
+      "step": 5907
+    },
+    {
+      "epoch": 1.257462090981644,
+      "grad_norm": 0.9479275514349785,
+      "learning_rate": 7.205054719439131e-05,
+      "loss": 2.9184,
+      "step": 5908
+    },
+    {
+      "epoch": 1.2576749135408354,
+      "grad_norm": 0.9170727105426545,
+      "learning_rate": 7.20394317876692e-05,
+      "loss": 2.903,
+      "step": 5909
+    },
+    {
+      "epoch": 1.2578877361000265,
+      "grad_norm": 0.8920048018333276,
+      "learning_rate": 7.202831502891295e-05,
+      "loss": 2.891,
+      "step": 5910
+    },
+    {
+      "epoch": 1.2581005586592178,
+      "grad_norm": 0.89209254413333,
+      "learning_rate": 7.201719691880446e-05,
+      "loss": 2.9125,
+      "step": 5911
+    },
+    {
+      "epoch": 1.258313381218409,
+      "grad_norm": 0.8806522187898083,
+      "learning_rate": 7.200607745802583e-05,
+      "loss": 2.9231,
+      "step": 5912
+    },
+    {
+      "epoch": 1.2585262037776004,
+      "grad_norm": 0.9403399747066937,
+      "learning_rate": 7.199495664725919e-05,
+      "loss": 2.9185,
+      "step": 5913
+    },
+    {
+      "epoch": 1.2587390263367917,
+      "grad_norm": 0.9883257822665718,
+      "learning_rate": 7.198383448718675e-05,
+      "loss": 2.8791,
+      "step": 5914
+    },
+    {
+      "epoch": 1.258951848895983,
+      "grad_norm": 0.960644780150262,
+      "learning_rate": 7.197271097849081e-05,
+      "loss": 2.9308,
+      "step": 5915
+    },
+    {
+      "epoch": 1.2591646714551743,
+      "grad_norm": 1.1192721025916688,
+      "learning_rate": 7.196158612185375e-05,
+      "loss": 2.946,
+      "step": 5916
+    },
+    {
+      "epoch": 1.2593774940143656,
+      "grad_norm": 0.9823227762288766,
+      "learning_rate": 7.195045991795806e-05,
+      "loss": 2.8769,
+      "step": 5917
+    },
+    {
+      "epoch": 1.259590316573557,
+      "grad_norm": 0.9898012809937805,
+      "learning_rate": 7.193933236748627e-05,
+      "loss": 2.9065,
+      "step": 5918
+    },
+    {
+      "epoch": 1.259803139132748,
+      "grad_norm": 0.9511204886414062,
+      "learning_rate": 7.192820347112101e-05,
+      "loss": 2.918,
+      "step": 5919
+    },
+    {
+      "epoch": 1.2600159616919393,
+      "grad_norm": 1.1239877480421363,
+      "learning_rate": 7.1917073229545e-05,
+      "loss": 2.8795,
+      "step": 5920
+    },
+    {
+      "epoch": 1.2602287842511306,
+      "grad_norm": 0.8379363883293247,
+      "learning_rate": 7.190594164344105e-05,
+      "loss": 2.898,
+      "step": 5921
+    },
+    {
+      "epoch": 1.260441606810322,
+      "grad_norm": 0.8357938615321354,
+      "learning_rate": 7.1894808713492e-05,
+      "loss": 2.9013,
+      "step": 5922
+    },
+    {
+      "epoch": 1.2606544293695132,
+      "grad_norm": 0.9324014560691078,
+      "learning_rate": 7.188367444038087e-05,
+      "loss": 2.9216,
+      "step": 5923
+    },
+    {
+      "epoch": 1.2608672519287045,
+      "grad_norm": 0.931363458766825,
+      "learning_rate": 7.187253882479067e-05,
+      "loss": 2.8729,
+      "step": 5924
+    },
+    {
+      "epoch": 1.2610800744878956,
+      "grad_norm": 0.8494956067329418,
+      "learning_rate": 7.186140186740453e-05,
+      "loss": 2.9005,
+      "step": 5925
+    },
+    {
+      "epoch": 1.261292897047087,
+      "grad_norm": 0.9721630655729293,
+      "learning_rate": 7.185026356890566e-05,
+      "loss": 2.9084,
+      "step": 5926
+    },
+    {
+      "epoch": 1.2615057196062782,
+      "grad_norm": 0.9093926053901633,
+      "learning_rate": 7.183912392997737e-05,
+      "loss": 2.9309,
+      "step": 5927
+    },
+    {
+      "epoch": 1.2617185421654695,
+      "grad_norm": 0.7746555918942335,
+      "learning_rate": 7.1827982951303e-05,
+      "loss": 2.9118,
+      "step": 5928
+    },
+    {
+      "epoch": 1.2619313647246608,
+      "grad_norm": 0.8369053404478378,
+      "learning_rate": 7.181684063356603e-05,
+      "loss": 2.9165,
+      "step": 5929
+    },
+    {
+      "epoch": 1.2621441872838521,
+      "grad_norm": 0.8367536511411944,
+      "learning_rate": 7.180569697744998e-05,
+      "loss": 2.8985,
+      "step": 5930
+    },
+    {
+      "epoch": 1.2623570098430434,
+      "grad_norm": 0.8529649451624185,
+      "learning_rate": 7.179455198363851e-05,
+      "loss": 2.8881,
+      "step": 5931
+    },
+    {
+      "epoch": 1.2625698324022347,
+      "grad_norm": 0.9303146490937054,
+      "learning_rate": 7.178340565281527e-05,
+      "loss": 2.969,
+      "step": 5932
+    },
+    {
+      "epoch": 1.2627826549614258,
+      "grad_norm": 0.9426167539070308,
+      "learning_rate": 7.177225798566407e-05,
+      "loss": 2.8884,
+      "step": 5933
+    },
+    {
+      "epoch": 1.2629954775206171,
+      "grad_norm": 0.9068508480130961,
+      "learning_rate": 7.176110898286878e-05,
+      "loss": 2.8895,
+      "step": 5934
+    },
+    {
+      "epoch": 1.2632083000798084,
+      "grad_norm": 0.9213340768289172,
+      "learning_rate": 7.174995864511336e-05,
+      "loss": 2.8769,
+      "step": 5935
+    },
+    {
+      "epoch": 1.2634211226389997,
+      "grad_norm": 0.9468215111856432,
+      "learning_rate": 7.173880697308179e-05,
+      "loss": 2.9338,
+      "step": 5936
+    },
+    {
+      "epoch": 1.263633945198191,
+      "grad_norm": 0.9205187554451131,
+      "learning_rate": 7.172765396745824e-05,
+      "loss": 2.908,
+      "step": 5937
+    },
+    {
+      "epoch": 1.2638467677573824,
+      "grad_norm": 0.8391400564100278,
+      "learning_rate": 7.17164996289269e-05,
+      "loss": 2.8994,
+      "step": 5938
+    },
+    {
+      "epoch": 1.2640595903165734,
+      "grad_norm": 0.7559200424038143,
+      "learning_rate": 7.1705343958172e-05,
+      "loss": 2.8948,
+      "step": 5939
+    },
+    {
+      "epoch": 1.2642724128757648,
+      "grad_norm": 0.7775618937375436,
+      "learning_rate": 7.169418695587791e-05,
+      "loss": 2.8871,
+      "step": 5940
+    },
+    {
+      "epoch": 1.264485235434956,
+      "grad_norm": 0.8047387269525517,
+      "learning_rate": 7.16830286227291e-05,
+      "loss": 2.8473,
+      "step": 5941
+    },
+    {
+      "epoch": 1.2646980579941474,
+      "grad_norm": 0.8718975951674267,
+      "learning_rate": 7.167186895941007e-05,
+      "loss": 2.8868,
+      "step": 5942
+    },
+    {
+      "epoch": 1.2649108805533387,
+      "grad_norm": 0.8260544968057987,
+      "learning_rate": 7.166070796660544e-05,
+      "loss": 2.9486,
+      "step": 5943
+    },
+    {
+      "epoch": 1.26512370311253,
+      "grad_norm": 0.8529088855582337,
+      "learning_rate": 7.164954564499985e-05,
+      "loss": 2.9052,
+      "step": 5944
+    },
+    {
+      "epoch": 1.2653365256717213,
+      "grad_norm": 0.9001647884477209,
+      "learning_rate": 7.163838199527812e-05,
+      "loss": 2.8892,
+      "step": 5945
+    },
+    {
+      "epoch": 1.2655493482309126,
+      "grad_norm": 0.8374582413722083,
+      "learning_rate": 7.162721701812505e-05,
+      "loss": 2.888,
+      "step": 5946
+    },
+    {
+      "epoch": 1.2657621707901037,
+      "grad_norm": 0.7978441816260619,
+      "learning_rate": 7.161605071422561e-05,
+      "loss": 2.8877,
+      "step": 5947
+    },
+    {
+      "epoch": 1.265974993349295,
+      "grad_norm": 0.7993526272011483,
+      "learning_rate": 7.160488308426477e-05,
+      "loss": 2.8979,
+      "step": 5948
+    },
+    {
+      "epoch": 1.2661878159084863,
+      "grad_norm": 0.840373151351831,
+      "learning_rate": 7.159371412892767e-05,
+      "loss": 2.8995,
+      "step": 5949
+    },
+    {
+      "epoch": 1.2664006384676776,
+      "grad_norm": 0.8397411934774639,
+      "learning_rate": 7.158254384889945e-05,
+      "loss": 2.9043,
+      "step": 5950
+    },
+    {
+      "epoch": 1.266613461026869,
+      "grad_norm": 0.7636177789864327,
+      "learning_rate": 7.157137224486533e-05,
+      "loss": 2.9319,
+      "step": 5951
+    },
+    {
+      "epoch": 1.2668262835860602,
+      "grad_norm": 0.8015519735994763,
+      "learning_rate": 7.156019931751072e-05,
+      "loss": 2.8985,
+      "step": 5952
+    },
+    {
+      "epoch": 1.2670391061452513,
+      "grad_norm": 0.8546897953006629,
+      "learning_rate": 7.154902506752096e-05,
+      "loss": 2.8422,
+      "step": 5953
+    },
+    {
+      "epoch": 1.2672519287044426,
+      "grad_norm": 0.839801582967653,
+      "learning_rate": 7.153784949558162e-05,
+      "loss": 2.9171,
+      "step": 5954
+    },
+    {
+      "epoch": 1.267464751263634,
+      "grad_norm": 0.7360276185875477,
+      "learning_rate": 7.152667260237823e-05,
+      "loss": 2.8523,
+      "step": 5955
+    },
+    {
+      "epoch": 1.2676775738228252,
+      "grad_norm": 0.6996594461658556,
+      "learning_rate": 7.151549438859646e-05,
+      "loss": 2.9347,
+      "step": 5956
+    },
+    {
+      "epoch": 1.2678903963820165,
+      "grad_norm": 0.7801391983215019,
+      "learning_rate": 7.150431485492207e-05,
+      "loss": 2.8495,
+      "step": 5957
+    },
+    {
+      "epoch": 1.2681032189412078,
+      "grad_norm": 0.8823152741973547,
+      "learning_rate": 7.149313400204083e-05,
+      "loss": 2.894,
+      "step": 5958
+    },
+    {
+      "epoch": 1.2683160415003991,
+      "grad_norm": 0.943790675840979,
+      "learning_rate": 7.148195183063871e-05,
+      "loss": 2.8613,
+      "step": 5959
+    },
+    {
+      "epoch": 1.2685288640595904,
+      "grad_norm": 1.0364174563077881,
+      "learning_rate": 7.147076834140163e-05,
+      "loss": 2.9081,
+      "step": 5960
+    },
+    {
+      "epoch": 1.2687416866187817,
+      "grad_norm": 1.061473397856767,
+      "learning_rate": 7.14595835350157e-05,
+      "loss": 2.9263,
+      "step": 5961
+    },
+    {
+      "epoch": 1.2689545091779728,
+      "grad_norm": 1.0422356327971813,
+      "learning_rate": 7.144839741216703e-05,
+      "loss": 2.9191,
+      "step": 5962
+    },
+    {
+      "epoch": 1.2691673317371641,
+      "grad_norm": 0.9610956176617148,
+      "learning_rate": 7.143720997354189e-05,
+      "loss": 2.9032,
+      "step": 5963
+    },
+    {
+      "epoch": 1.2693801542963554,
+      "grad_norm": 0.9303575561700671,
+      "learning_rate": 7.142602121982653e-05,
+      "loss": 2.8846,
+      "step": 5964
+    },
+    {
+      "epoch": 1.2695929768555467,
+      "grad_norm": 0.7630527096166332,
+      "learning_rate": 7.141483115170739e-05,
+      "loss": 2.9031,
+      "step": 5965
+    },
+    {
+      "epoch": 1.269805799414738,
+      "grad_norm": 0.8867867285559936,
+      "learning_rate": 7.140363976987092e-05,
+      "loss": 2.8996,
+      "step": 5966
+    },
+    {
+      "epoch": 1.2700186219739291,
+      "grad_norm": 1.1052267900283619,
+      "learning_rate": 7.139244707500363e-05,
+      "loss": 2.904,
+      "step": 5967
+    },
+    {
+      "epoch": 1.2702314445331204,
+      "grad_norm": 0.8959850210725274,
+      "learning_rate": 7.138125306779221e-05,
+      "loss": 2.8837,
+      "step": 5968
+    },
+    {
+      "epoch": 1.2704442670923117,
+      "grad_norm": 0.8305895160718476,
+      "learning_rate": 7.137005774892334e-05,
+      "loss": 2.9046,
+      "step": 5969
+    },
+    {
+      "epoch": 1.270657089651503,
+      "grad_norm": 0.8692135058760887,
+      "learning_rate": 7.135886111908378e-05,
+      "loss": 2.8854,
+      "step": 5970
+    },
+    {
+      "epoch": 1.2708699122106943,
+      "grad_norm": 0.85705837322571,
+      "learning_rate": 7.134766317896046e-05,
+      "loss": 2.9145,
+      "step": 5971
+    },
+    {
+      "epoch": 1.2710827347698856,
+      "grad_norm": 0.9397652606439232,
+      "learning_rate": 7.133646392924029e-05,
+      "loss": 2.9097,
+      "step": 5972
+    },
+    {
+      "epoch": 1.271295557329077,
+      "grad_norm": 0.8511011045428145,
+      "learning_rate": 7.132526337061031e-05,
+      "loss": 2.8748,
+      "step": 5973
+    },
+    {
+      "epoch": 1.2715083798882683,
+      "grad_norm": 0.6867861239112908,
+      "learning_rate": 7.131406150375763e-05,
+      "loss": 2.854,
+      "step": 5974
+    },
+    {
+      "epoch": 1.2717212024474596,
+      "grad_norm": 0.9285028532723174,
+      "learning_rate": 7.130285832936944e-05,
+      "loss": 2.905,
+      "step": 5975
+    },
+    {
+      "epoch": 1.2719340250066506,
+      "grad_norm": 1.0814245162692238,
+      "learning_rate": 7.129165384813303e-05,
+      "loss": 2.897,
+      "step": 5976
+    },
+    {
+      "epoch": 1.272146847565842,
+      "grad_norm": 1.039479638021501,
+      "learning_rate": 7.128044806073573e-05,
+      "loss": 2.8833,
+      "step": 5977
+    },
+    {
+      "epoch": 1.2723596701250333,
+      "grad_norm": 0.9200405036291311,
+      "learning_rate": 7.126924096786496e-05,
+      "loss": 2.9054,
+      "step": 5978
+    },
+    {
+      "epoch": 1.2725724926842246,
+      "grad_norm": 0.8292824424209747,
+      "learning_rate": 7.125803257020829e-05,
+      "loss": 2.8603,
+      "step": 5979
+    },
+    {
+      "epoch": 1.2727853152434159,
+      "grad_norm": 0.9605341950687789,
+      "learning_rate": 7.124682286845323e-05,
+      "loss": 2.9222,
+      "step": 5980
+    },
+    {
+      "epoch": 1.272998137802607,
+      "grad_norm": 1.0218064920404804,
+      "learning_rate": 7.123561186328753e-05,
+      "loss": 2.8928,
+      "step": 5981
+    },
+    {
+      "epoch": 1.2732109603617983,
+      "grad_norm": 0.9289281384300934,
+      "learning_rate": 7.122439955539889e-05,
+      "loss": 2.8745,
+      "step": 5982
+    },
+    {
+      "epoch": 1.2734237829209896,
+      "grad_norm": 0.8855768503118451,
+      "learning_rate": 7.121318594547516e-05,
+      "loss": 2.9178,
+      "step": 5983
+    },
+    {
+      "epoch": 1.2736366054801809,
+      "grad_norm": 0.8705027417321456,
+      "learning_rate": 7.120197103420426e-05,
+      "loss": 2.9175,
+      "step": 5984
+    },
+    {
+      "epoch": 1.2738494280393722,
+      "grad_norm": 1.032212657038542,
+      "learning_rate": 7.119075482227414e-05,
+      "loss": 2.8938,
+      "step": 5985
+    },
+    {
+      "epoch": 1.2740622505985635,
+      "grad_norm": 0.9799509321283991,
+      "learning_rate": 7.117953731037293e-05,
+      "loss": 2.9098,
+      "step": 5986
+    },
+    {
+      "epoch": 1.2742750731577548,
+      "grad_norm": 0.8698751367613154,
+      "learning_rate": 7.116831849918874e-05,
+      "loss": 2.8723,
+      "step": 5987
+    },
+    {
+      "epoch": 1.274487895716946,
+      "grad_norm": 0.8186656950651546,
+      "learning_rate": 7.115709838940983e-05,
+      "loss": 2.9116,
+      "step": 5988
+    },
+    {
+      "epoch": 1.2747007182761374,
+      "grad_norm": 0.8373407932585464,
+      "learning_rate": 7.114587698172449e-05,
+      "loss": 2.8986,
+      "step": 5989
+    },
+    {
+      "epoch": 1.2749135408353285,
+      "grad_norm": 0.8083367249118475,
+      "learning_rate": 7.113465427682112e-05,
+      "loss": 2.8755,
+      "step": 5990
+    },
+    {
+      "epoch": 1.2751263633945198,
+      "grad_norm": 0.8570352827759461,
+      "learning_rate": 7.112343027538818e-05,
+      "loss": 2.894,
+      "step": 5991
+    },
+    {
+      "epoch": 1.275339185953711,
+      "grad_norm": 0.8176094391513459,
+      "learning_rate": 7.111220497811421e-05,
+      "loss": 2.8981,
+      "step": 5992
+    },
+    {
+      "epoch": 1.2755520085129024,
+      "grad_norm": 0.832764463649728,
+      "learning_rate": 7.110097838568788e-05,
+      "loss": 2.901,
+      "step": 5993
+    },
+    {
+      "epoch": 1.2757648310720937,
+      "grad_norm": 0.810857332861436,
+      "learning_rate": 7.108975049879784e-05,
+      "loss": 2.8996,
+      "step": 5994
+    },
+    {
+      "epoch": 1.2759776536312848,
+      "grad_norm": 0.688170795468895,
+      "learning_rate": 7.107852131813294e-05,
+      "loss": 2.9013,
+      "step": 5995
+    },
+    {
+      "epoch": 1.276190476190476,
+      "grad_norm": 0.8563428849127295,
+      "learning_rate": 7.1067290844382e-05,
+      "loss": 2.8665,
+      "step": 5996
+    },
+    {
+      "epoch": 1.2764032987496674,
+      "grad_norm": 0.8027721165464734,
+      "learning_rate": 7.1056059078234e-05,
+      "loss": 2.8634,
+      "step": 5997
+    },
+    {
+      "epoch": 1.2766161213088587,
+      "grad_norm": 0.8766223972132371,
+      "learning_rate": 7.104482602037793e-05,
+      "loss": 2.8588,
+      "step": 5998
+    },
+    {
+      "epoch": 1.27682894386805,
+      "grad_norm": 0.9118929164968365,
+      "learning_rate": 7.103359167150291e-05,
+      "loss": 2.9071,
+      "step": 5999
+    },
+    {
+      "epoch": 1.2770417664272413,
+      "grad_norm": 0.8484683038776094,
+      "learning_rate": 7.102235603229814e-05,
+      "loss": 2.9305,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 14094,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 3000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1449491064422400.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}