diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,91532 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 13070,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001530221882172915,
+      "grad_norm": 13.516160750382397,
+      "learning_rate": 5.0890585241730285e-08,
+      "loss": 1.0816,
+      "step": 1
+    },
+    {
+      "epoch": 0.000306044376434583,
+      "grad_norm": 13.656487119130613,
+      "learning_rate": 1.0178117048346057e-07,
+      "loss": 1.0528,
+      "step": 2
+    },
+    {
+      "epoch": 0.0004590665646518745,
+      "grad_norm": 16.846354425859488,
+      "learning_rate": 1.5267175572519085e-07,
+      "loss": 1.1708,
+      "step": 3
+    },
+    {
+      "epoch": 0.000612088752869166,
+      "grad_norm": 13.046572036423532,
+      "learning_rate": 2.0356234096692114e-07,
+      "loss": 1.0517,
+      "step": 4
+    },
+    {
+      "epoch": 0.0007651109410864575,
+      "grad_norm": 12.49233347364335,
+      "learning_rate": 2.544529262086514e-07,
+      "loss": 1.0145,
+      "step": 5
+    },
+    {
+      "epoch": 0.000918133129303749,
+      "grad_norm": 14.751771565656806,
+      "learning_rate": 3.053435114503817e-07,
+      "loss": 1.0758,
+      "step": 6
+    },
+    {
+      "epoch": 0.0010711553175210406,
+      "grad_norm": 11.653404038618469,
+      "learning_rate": 3.56234096692112e-07,
+      "loss": 1.0636,
+      "step": 7
+    },
+    {
+      "epoch": 0.001224177505738332,
+      "grad_norm": 13.668951288685138,
+      "learning_rate": 4.071246819338423e-07,
+      "loss": 1.1952,
+      "step": 8
+    },
+    {
+      "epoch": 0.0013771996939556236,
+      "grad_norm": 12.046531599063504,
+      "learning_rate": 4.5801526717557257e-07,
+      "loss": 1.0257,
+      "step": 9
+    },
+    {
+      "epoch": 0.001530221882172915,
+      "grad_norm": 13.723654296994994,
+      "learning_rate": 5.089058524173028e-07,
+      "loss": 1.059,
+      "step": 10
+    },
+    {
+      "epoch": 0.0016832440703902067,
+      "grad_norm": 13.902265225322944,
+      "learning_rate": 5.597964376590332e-07,
+      "loss": 1.0858,
+      "step": 11
+    },
+    {
+      "epoch": 0.001836266258607498,
+      "grad_norm": 12.528149564606679,
+      "learning_rate": 6.106870229007634e-07,
+      "loss": 1.0277,
+      "step": 12
+    },
+    {
+      "epoch": 0.0019892884468247895,
+      "grad_norm": 13.1123093332078,
+      "learning_rate": 6.615776081424936e-07,
+      "loss": 0.9474,
+      "step": 13
+    },
+    {
+      "epoch": 0.002142310635042081,
+      "grad_norm": 11.039557385133689,
+      "learning_rate": 7.12468193384224e-07,
+      "loss": 1.0024,
+      "step": 14
+    },
+    {
+      "epoch": 0.0022953328232593728,
+      "grad_norm": 8.906284572299981,
+      "learning_rate": 7.633587786259543e-07,
+      "loss": 0.9381,
+      "step": 15
+    },
+    {
+      "epoch": 0.002448355011476664,
+      "grad_norm": 10.05716027554367,
+      "learning_rate": 8.142493638676846e-07,
+      "loss": 0.9707,
+      "step": 16
+    },
+    {
+      "epoch": 0.0026013771996939556,
+      "grad_norm": 9.006177898036631,
+      "learning_rate": 8.651399491094148e-07,
+      "loss": 0.9394,
+      "step": 17
+    },
+    {
+      "epoch": 0.0027543993879112472,
+      "grad_norm": 21.183255574001187,
+      "learning_rate": 9.160305343511451e-07,
+      "loss": 1.0931,
+      "step": 18
+    },
+    {
+      "epoch": 0.002907421576128539,
+      "grad_norm": 10.995986966246456,
+      "learning_rate": 9.669211195928755e-07,
+      "loss": 0.984,
+      "step": 19
+    },
+    {
+      "epoch": 0.00306044376434583,
+      "grad_norm": 11.307646984532525,
+      "learning_rate": 1.0178117048346056e-06,
+      "loss": 0.9666,
+      "step": 20
+    },
+    {
+      "epoch": 0.0032134659525631217,
+      "grad_norm": 14.964095843531805,
+      "learning_rate": 1.068702290076336e-06,
+      "loss": 1.0085,
+      "step": 21
+    },
+    {
+      "epoch": 0.0033664881407804133,
+      "grad_norm": 9.3514966763225,
+      "learning_rate": 1.1195928753180663e-06,
+      "loss": 0.9647,
+      "step": 22
+    },
+    {
+      "epoch": 0.0035195103289977045,
+      "grad_norm": 7.7748996347946,
+      "learning_rate": 1.1704834605597967e-06,
+      "loss": 1.0044,
+      "step": 23
+    },
+    {
+      "epoch": 0.003672532517214996,
+      "grad_norm": 9.478805120179091,
+      "learning_rate": 1.2213740458015268e-06,
+      "loss": 0.9446,
+      "step": 24
+    },
+    {
+      "epoch": 0.0038255547054322878,
+      "grad_norm": 6.758033563189933,
+      "learning_rate": 1.2722646310432571e-06,
+      "loss": 0.9163,
+      "step": 25
+    },
+    {
+      "epoch": 0.003978576893649579,
+      "grad_norm": 5.984168999533934,
+      "learning_rate": 1.3231552162849873e-06,
+      "loss": 0.8452,
+      "step": 26
+    },
+    {
+      "epoch": 0.004131599081866871,
+      "grad_norm": 6.141293673107684,
+      "learning_rate": 1.3740458015267178e-06,
+      "loss": 0.9215,
+      "step": 27
+    },
+    {
+      "epoch": 0.004284621270084162,
+      "grad_norm": 8.443582458602195,
+      "learning_rate": 1.424936386768448e-06,
+      "loss": 0.9543,
+      "step": 28
+    },
+    {
+      "epoch": 0.004437643458301454,
+      "grad_norm": 6.219529039996243,
+      "learning_rate": 1.475826972010178e-06,
+      "loss": 1.1055,
+      "step": 29
+    },
+    {
+      "epoch": 0.0045906656465187455,
+      "grad_norm": 6.723218647037651,
+      "learning_rate": 1.5267175572519086e-06,
+      "loss": 0.8657,
+      "step": 30
+    },
+    {
+      "epoch": 0.004743687834736037,
+      "grad_norm": 6.925216418396402,
+      "learning_rate": 1.5776081424936388e-06,
+      "loss": 1.0145,
+      "step": 31
+    },
+    {
+      "epoch": 0.004896710022953328,
+      "grad_norm": 5.817861415648484,
+      "learning_rate": 1.6284987277353691e-06,
+      "loss": 0.9494,
+      "step": 32
+    },
+    {
+      "epoch": 0.0050497322111706195,
+      "grad_norm": 6.517661419074156,
+      "learning_rate": 1.6793893129770995e-06,
+      "loss": 1.0573,
+      "step": 33
+    },
+    {
+      "epoch": 0.005202754399387911,
+      "grad_norm": 7.507479750968244,
+      "learning_rate": 1.7302798982188296e-06,
+      "loss": 0.8793,
+      "step": 34
+    },
+    {
+      "epoch": 0.005355776587605203,
+      "grad_norm": 6.961283303110031,
+      "learning_rate": 1.78117048346056e-06,
+      "loss": 0.9334,
+      "step": 35
+    },
+    {
+      "epoch": 0.0055087987758224944,
+      "grad_norm": 7.154082162667194,
+      "learning_rate": 1.8320610687022903e-06,
+      "loss": 0.8061,
+      "step": 36
+    },
+    {
+      "epoch": 0.005661820964039786,
+      "grad_norm": 6.110703133559988,
+      "learning_rate": 1.8829516539440204e-06,
+      "loss": 0.8045,
+      "step": 37
+    },
+    {
+      "epoch": 0.005814843152257078,
+      "grad_norm": 6.144648736565948,
+      "learning_rate": 1.933842239185751e-06,
+      "loss": 0.7255,
+      "step": 38
+    },
+    {
+      "epoch": 0.0059678653404743685,
+      "grad_norm": 6.104408552237334,
+      "learning_rate": 1.984732824427481e-06,
+      "loss": 0.8799,
+      "step": 39
+    },
+    {
+      "epoch": 0.00612088752869166,
+      "grad_norm": 6.224216203164156,
+      "learning_rate": 2.0356234096692112e-06,
+      "loss": 0.8629,
+      "step": 40
+    },
+    {
+      "epoch": 0.006273909716908952,
+      "grad_norm": 5.354540996861285,
+      "learning_rate": 2.0865139949109416e-06,
+      "loss": 0.9332,
+      "step": 41
+    },
+    {
+      "epoch": 0.006426931905126243,
+      "grad_norm": 7.248232082302233,
+      "learning_rate": 2.137404580152672e-06,
+      "loss": 0.879,
+      "step": 42
+    },
+    {
+      "epoch": 0.006579954093343535,
+      "grad_norm": 6.526634157834574,
+      "learning_rate": 2.1882951653944023e-06,
+      "loss": 0.8931,
+      "step": 43
+    },
+    {
+      "epoch": 0.006732976281560827,
+      "grad_norm": 5.558985637856043,
+      "learning_rate": 2.2391857506361326e-06,
+      "loss": 0.9254,
+      "step": 44
+    },
+    {
+      "epoch": 0.006885998469778117,
+      "grad_norm": 5.143083416175365,
+      "learning_rate": 2.2900763358778625e-06,
+      "loss": 0.7931,
+      "step": 45
+    },
+    {
+      "epoch": 0.007039020657995409,
+      "grad_norm": 6.230004131099689,
+      "learning_rate": 2.3409669211195933e-06,
+      "loss": 0.8783,
+      "step": 46
+    },
+    {
+      "epoch": 0.007192042846212701,
+      "grad_norm": 6.96270555642169,
+      "learning_rate": 2.3918575063613232e-06,
+      "loss": 0.8966,
+      "step": 47
+    },
+    {
+      "epoch": 0.007345065034429992,
+      "grad_norm": 4.725287444763657,
+      "learning_rate": 2.4427480916030536e-06,
+      "loss": 0.792,
+      "step": 48
+    },
+    {
+      "epoch": 0.007498087222647284,
+      "grad_norm": 4.516169085236729,
+      "learning_rate": 2.493638676844784e-06,
+      "loss": 0.8188,
+      "step": 49
+    },
+    {
+      "epoch": 0.0076511094108645756,
+      "grad_norm": 6.169099136800104,
+      "learning_rate": 2.5445292620865143e-06,
+      "loss": 0.9132,
+      "step": 50
+    },
+    {
+      "epoch": 0.007804131599081867,
+      "grad_norm": 5.030562592462142,
+      "learning_rate": 2.595419847328244e-06,
+      "loss": 0.9125,
+      "step": 51
+    },
+    {
+      "epoch": 0.007957153787299158,
+      "grad_norm": 5.741056522914908,
+      "learning_rate": 2.6463104325699745e-06,
+      "loss": 0.8294,
+      "step": 52
+    },
+    {
+      "epoch": 0.00811017597551645,
+      "grad_norm": 5.3502817029143515,
+      "learning_rate": 2.6972010178117053e-06,
+      "loss": 0.8005,
+      "step": 53
+    },
+    {
+      "epoch": 0.008263198163733741,
+      "grad_norm": 5.153661369992958,
+      "learning_rate": 2.7480916030534356e-06,
+      "loss": 0.8175,
+      "step": 54
+    },
+    {
+      "epoch": 0.008416220351951033,
+      "grad_norm": 6.046120010370828,
+      "learning_rate": 2.7989821882951656e-06,
+      "loss": 0.8508,
+      "step": 55
+    },
+    {
+      "epoch": 0.008569242540168324,
+      "grad_norm": 6.292867560774823,
+      "learning_rate": 2.849872773536896e-06,
+      "loss": 0.9253,
+      "step": 56
+    },
+    {
+      "epoch": 0.008722264728385616,
+      "grad_norm": 6.105008240723325,
+      "learning_rate": 2.900763358778626e-06,
+      "loss": 0.9678,
+      "step": 57
+    },
+    {
+      "epoch": 0.008875286916602908,
+      "grad_norm": 4.459519530250151,
+      "learning_rate": 2.951653944020356e-06,
+      "loss": 0.857,
+      "step": 58
+    },
+    {
+      "epoch": 0.0090283091048202,
+      "grad_norm": 4.966961705007526,
+      "learning_rate": 3.002544529262087e-06,
+      "loss": 0.8162,
+      "step": 59
+    },
+    {
+      "epoch": 0.009181331293037491,
+      "grad_norm": 5.179614147649331,
+      "learning_rate": 3.0534351145038173e-06,
+      "loss": 0.7892,
+      "step": 60
+    },
+    {
+      "epoch": 0.009334353481254783,
+      "grad_norm": 6.585863543513201,
+      "learning_rate": 3.104325699745547e-06,
+      "loss": 0.81,
+      "step": 61
+    },
+    {
+      "epoch": 0.009487375669472074,
+      "grad_norm": 5.684014842731852,
+      "learning_rate": 3.1552162849872775e-06,
+      "loss": 0.9146,
+      "step": 62
+    },
+    {
+      "epoch": 0.009640397857689364,
+      "grad_norm": 4.941334064996768,
+      "learning_rate": 3.206106870229008e-06,
+      "loss": 0.8218,
+      "step": 63
+    },
+    {
+      "epoch": 0.009793420045906656,
+      "grad_norm": 5.436505901693147,
+      "learning_rate": 3.2569974554707382e-06,
+      "loss": 0.7455,
+      "step": 64
+    },
+    {
+      "epoch": 0.009946442234123947,
+      "grad_norm": 4.827550970928569,
+      "learning_rate": 3.3078880407124686e-06,
+      "loss": 0.7799,
+      "step": 65
+    },
+    {
+      "epoch": 0.010099464422341239,
+      "grad_norm": 5.1126560035818125,
+      "learning_rate": 3.358778625954199e-06,
+      "loss": 0.8829,
+      "step": 66
+    },
+    {
+      "epoch": 0.01025248661055853,
+      "grad_norm": 4.734944949714936,
+      "learning_rate": 3.409669211195929e-06,
+      "loss": 0.8362,
+      "step": 67
+    },
+    {
+      "epoch": 0.010405508798775822,
+      "grad_norm": 5.1152605564609726,
+      "learning_rate": 3.460559796437659e-06,
+      "loss": 0.806,
+      "step": 68
+    },
+    {
+      "epoch": 0.010558530986993114,
+      "grad_norm": 5.501335018085708,
+      "learning_rate": 3.5114503816793895e-06,
+      "loss": 0.8836,
+      "step": 69
+    },
+    {
+      "epoch": 0.010711553175210406,
+      "grad_norm": 5.6907042929238,
+      "learning_rate": 3.56234096692112e-06,
+      "loss": 0.7729,
+      "step": 70
+    },
+    {
+      "epoch": 0.010864575363427697,
+      "grad_norm": 4.7874920288700755,
+      "learning_rate": 3.6132315521628502e-06,
+      "loss": 0.7401,
+      "step": 71
+    },
+    {
+      "epoch": 0.011017597551644989,
+      "grad_norm": 5.34946991015153,
+      "learning_rate": 3.6641221374045806e-06,
+      "loss": 0.804,
+      "step": 72
+    },
+    {
+      "epoch": 0.01117061973986228,
+      "grad_norm": 6.0799857986381,
+      "learning_rate": 3.7150127226463105e-06,
+      "loss": 0.845,
+      "step": 73
+    },
+    {
+      "epoch": 0.011323641928079572,
+      "grad_norm": 6.075178425573325,
+      "learning_rate": 3.765903307888041e-06,
+      "loss": 0.7568,
+      "step": 74
+    },
+    {
+      "epoch": 0.011476664116296864,
+      "grad_norm": 4.759854629433448,
+      "learning_rate": 3.816793893129772e-06,
+      "loss": 0.7735,
+      "step": 75
+    },
+    {
+      "epoch": 0.011629686304514155,
+      "grad_norm": 5.486398129731232,
+      "learning_rate": 3.867684478371502e-06,
+      "loss": 0.788,
+      "step": 76
+    },
+    {
+      "epoch": 0.011782708492731445,
+      "grad_norm": 5.097255872801085,
+      "learning_rate": 3.918575063613232e-06,
+      "loss": 0.7627,
+      "step": 77
+    },
+    {
+      "epoch": 0.011935730680948737,
+      "grad_norm": 4.242028635114688,
+      "learning_rate": 3.969465648854962e-06,
+      "loss": 0.7489,
+      "step": 78
+    },
+    {
+      "epoch": 0.012088752869166029,
+      "grad_norm": 5.948491901588452,
+      "learning_rate": 4.020356234096692e-06,
+      "loss": 0.8135,
+      "step": 79
+    },
+    {
+      "epoch": 0.01224177505738332,
+      "grad_norm": 6.895223832667884,
+      "learning_rate": 4.0712468193384225e-06,
+      "loss": 0.84,
+      "step": 80
+    },
+    {
+      "epoch": 0.012394797245600612,
+      "grad_norm": 4.672739171576903,
+      "learning_rate": 4.122137404580153e-06,
+      "loss": 0.7619,
+      "step": 81
+    },
+    {
+      "epoch": 0.012547819433817903,
+      "grad_norm": 4.6442577318217815,
+      "learning_rate": 4.173027989821883e-06,
+      "loss": 0.8648,
+      "step": 82
+    },
+    {
+      "epoch": 0.012700841622035195,
+      "grad_norm": 5.053558792090741,
+      "learning_rate": 4.2239185750636135e-06,
+      "loss": 0.8448,
+      "step": 83
+    },
+    {
+      "epoch": 0.012853863810252487,
+      "grad_norm": 4.530911050956989,
+      "learning_rate": 4.274809160305344e-06,
+      "loss": 0.8076,
+      "step": 84
+    },
+    {
+      "epoch": 0.013006885998469778,
+      "grad_norm": 6.262686478254499,
+      "learning_rate": 4.325699745547074e-06,
+      "loss": 0.9201,
+      "step": 85
+    },
+    {
+      "epoch": 0.01315990818668707,
+      "grad_norm": 4.640895988619081,
+      "learning_rate": 4.3765903307888045e-06,
+      "loss": 0.7195,
+      "step": 86
+    },
+    {
+      "epoch": 0.013312930374904362,
+      "grad_norm": 5.0565684924592755,
+      "learning_rate": 4.427480916030535e-06,
+      "loss": 0.8627,
+      "step": 87
+    },
+    {
+      "epoch": 0.013465952563121653,
+      "grad_norm": 5.283088934487553,
+      "learning_rate": 4.478371501272265e-06,
+      "loss": 0.7922,
+      "step": 88
+    },
+    {
+      "epoch": 0.013618974751338945,
+      "grad_norm": 5.026171701957064,
+      "learning_rate": 4.5292620865139956e-06,
+      "loss": 0.8261,
+      "step": 89
+    },
+    {
+      "epoch": 0.013771996939556235,
+      "grad_norm": 5.884463834573161,
+      "learning_rate": 4.580152671755725e-06,
+      "loss": 0.8093,
+      "step": 90
+    },
+    {
+      "epoch": 0.013925019127773526,
+      "grad_norm": 4.438561387505246,
+      "learning_rate": 4.631043256997455e-06,
+      "loss": 0.7371,
+      "step": 91
+    },
+    {
+      "epoch": 0.014078041315990818,
+      "grad_norm": 5.305169117726312,
+      "learning_rate": 4.681933842239187e-06,
+      "loss": 0.8009,
+      "step": 92
+    },
+    {
+      "epoch": 0.01423106350420811,
+      "grad_norm": 5.419418332275794,
+      "learning_rate": 4.732824427480917e-06,
+      "loss": 0.8181,
+      "step": 93
+    },
+    {
+      "epoch": 0.014384085692425401,
+      "grad_norm": 5.040887136301689,
+      "learning_rate": 4.7837150127226464e-06,
+      "loss": 0.7963,
+      "step": 94
+    },
+    {
+      "epoch": 0.014537107880642693,
+      "grad_norm": 6.45075012480531,
+      "learning_rate": 4.834605597964377e-06,
+      "loss": 0.6717,
+      "step": 95
+    },
+    {
+      "epoch": 0.014690130068859985,
+      "grad_norm": 5.015628494483557,
+      "learning_rate": 4.885496183206107e-06,
+      "loss": 0.7777,
+      "step": 96
+    },
+    {
+      "epoch": 0.014843152257077276,
+      "grad_norm": 4.517147075856258,
+      "learning_rate": 4.9363867684478375e-06,
+      "loss": 0.7682,
+      "step": 97
+    },
+    {
+      "epoch": 0.014996174445294568,
+      "grad_norm": 5.015648995454229,
+      "learning_rate": 4.987277353689568e-06,
+      "loss": 0.793,
+      "step": 98
+    },
+    {
+      "epoch": 0.01514919663351186,
+      "grad_norm": 4.91359741535642,
+      "learning_rate": 5.038167938931297e-06,
+      "loss": 0.7853,
+      "step": 99
+    },
+    {
+      "epoch": 0.015302218821729151,
+      "grad_norm": 4.88280014065132,
+      "learning_rate": 5.0890585241730285e-06,
+      "loss": 0.7867,
+      "step": 100
+    },
+    {
+      "epoch": 0.015455241009946443,
+      "grad_norm": 5.310203953123644,
+      "learning_rate": 5.139949109414759e-06,
+      "loss": 0.8291,
+      "step": 101
+    },
+    {
+      "epoch": 0.015608263198163734,
+      "grad_norm": 5.063524860125988,
+      "learning_rate": 5.190839694656488e-06,
+      "loss": 0.8207,
+      "step": 102
+    },
+    {
+      "epoch": 0.015761285386381024,
+      "grad_norm": 5.476946259775138,
+      "learning_rate": 5.2417302798982195e-06,
+      "loss": 0.7438,
+      "step": 103
+    },
+    {
+      "epoch": 0.015914307574598316,
+      "grad_norm": 4.563268298277021,
+      "learning_rate": 5.292620865139949e-06,
+      "loss": 0.8014,
+      "step": 104
+    },
+    {
+      "epoch": 0.016067329762815608,
+      "grad_norm": 4.766384946361356,
+      "learning_rate": 5.34351145038168e-06,
+      "loss": 0.7385,
+      "step": 105
+    },
+    {
+      "epoch": 0.0162203519510329,
+      "grad_norm": 4.959052123725921,
+      "learning_rate": 5.394402035623411e-06,
+      "loss": 0.6631,
+      "step": 106
+    },
+    {
+      "epoch": 0.01637337413925019,
+      "grad_norm": 4.660517289576387,
+      "learning_rate": 5.44529262086514e-06,
+      "loss": 0.7539,
+      "step": 107
+    },
+    {
+      "epoch": 0.016526396327467482,
+      "grad_norm": 4.760332379137938,
+      "learning_rate": 5.496183206106871e-06,
+      "loss": 0.8044,
+      "step": 108
+    },
+    {
+      "epoch": 0.016679418515684774,
+      "grad_norm": 5.213877297259624,
+      "learning_rate": 5.547073791348601e-06,
+      "loss": 0.8032,
+      "step": 109
+    },
+    {
+      "epoch": 0.016832440703902066,
+      "grad_norm": 4.500414357173698,
+      "learning_rate": 5.597964376590331e-06,
+      "loss": 0.851,
+      "step": 110
+    },
+    {
+      "epoch": 0.016985462892119357,
+      "grad_norm": 5.20758083198332,
+      "learning_rate": 5.648854961832062e-06,
+      "loss": 0.8376,
+      "step": 111
+    },
+    {
+      "epoch": 0.01713848508033665,
+      "grad_norm": 4.313432158821755,
+      "learning_rate": 5.699745547073792e-06,
+      "loss": 0.7425,
+      "step": 112
+    },
+    {
+      "epoch": 0.01729150726855394,
+      "grad_norm": 4.384745758633966,
+      "learning_rate": 5.750636132315522e-06,
+      "loss": 0.9117,
+      "step": 113
+    },
+    {
+      "epoch": 0.017444529456771232,
+      "grad_norm": 5.289188278919928,
+      "learning_rate": 5.801526717557252e-06,
+      "loss": 0.7709,
+      "step": 114
+    },
+    {
+      "epoch": 0.017597551644988524,
+      "grad_norm": 4.718040596929263,
+      "learning_rate": 5.852417302798983e-06,
+      "loss": 0.825,
+      "step": 115
+    },
+    {
+      "epoch": 0.017750573833205815,
+      "grad_norm": 4.4740329916884045,
+      "learning_rate": 5.903307888040712e-06,
+      "loss": 0.8031,
+      "step": 116
+    },
+    {
+      "epoch": 0.017903596021423107,
+      "grad_norm": 3.8294489939326657,
+      "learning_rate": 5.9541984732824435e-06,
+      "loss": 0.7537,
+      "step": 117
+    },
+    {
+      "epoch": 0.0180566182096404,
+      "grad_norm": 5.3921477281799906,
+      "learning_rate": 6.005089058524174e-06,
+      "loss": 0.8343,
+      "step": 118
+    },
+    {
+      "epoch": 0.01820964039785769,
+      "grad_norm": 4.585186066396906,
+      "learning_rate": 6.055979643765903e-06,
+      "loss": 0.8059,
+      "step": 119
+    },
+    {
+      "epoch": 0.018362662586074982,
+      "grad_norm": 4.164993298962277,
+      "learning_rate": 6.1068702290076346e-06,
+      "loss": 0.7707,
+      "step": 120
+    },
+    {
+      "epoch": 0.018515684774292274,
+      "grad_norm": 4.7470966220472715,
+      "learning_rate": 6.157760814249364e-06,
+      "loss": 0.7635,
+      "step": 121
+    },
+    {
+      "epoch": 0.018668706962509565,
+      "grad_norm": 4.57053122827436,
+      "learning_rate": 6.208651399491094e-06,
+      "loss": 0.7802,
+      "step": 122
+    },
+    {
+      "epoch": 0.018821729150726857,
+      "grad_norm": 4.764792340656931,
+      "learning_rate": 6.259541984732826e-06,
+      "loss": 0.7754,
+      "step": 123
+    },
+    {
+      "epoch": 0.01897475133894415,
+      "grad_norm": 4.673906229696043,
+      "learning_rate": 6.310432569974555e-06,
+      "loss": 0.7227,
+      "step": 124
+    },
+    {
+      "epoch": 0.019127773527161437,
+      "grad_norm": 4.254889278087384,
+      "learning_rate": 6.3613231552162854e-06,
+      "loss": 0.7967,
+      "step": 125
+    },
+    {
+      "epoch": 0.01928079571537873,
+      "grad_norm": 3.9781988510098314,
+      "learning_rate": 6.412213740458016e-06,
+      "loss": 0.6657,
+      "step": 126
+    },
+    {
+      "epoch": 0.01943381790359602,
+      "grad_norm": 4.294430988412905,
+      "learning_rate": 6.463104325699746e-06,
+      "loss": 0.8075,
+      "step": 127
+    },
+    {
+      "epoch": 0.01958684009181331,
+      "grad_norm": 4.562471061393905,
+      "learning_rate": 6.5139949109414765e-06,
+      "loss": 0.8503,
+      "step": 128
+    },
+    {
+      "epoch": 0.019739862280030603,
+      "grad_norm": 4.549436052521284,
+      "learning_rate": 6.564885496183207e-06,
+      "loss": 0.8009,
+      "step": 129
+    },
+    {
+      "epoch": 0.019892884468247895,
+      "grad_norm": 3.835829653833588,
+      "learning_rate": 6.615776081424937e-06,
+      "loss": 0.7992,
+      "step": 130
+    },
+    {
+      "epoch": 0.020045906656465187,
+      "grad_norm": 4.267030813968025,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7283,
+      "step": 131
+    },
+    {
+      "epoch": 0.020198928844682478,
+      "grad_norm": 3.839774979418662,
+      "learning_rate": 6.717557251908398e-06,
+      "loss": 0.8088,
+      "step": 132
+    },
+    {
+      "epoch": 0.02035195103289977,
+      "grad_norm": 4.05482208608787,
+      "learning_rate": 6.768447837150128e-06,
+      "loss": 0.691,
+      "step": 133
+    },
+    {
+      "epoch": 0.02050497322111706,
+      "grad_norm": 4.39540221039343,
+      "learning_rate": 6.819338422391858e-06,
+      "loss": 0.769,
+      "step": 134
+    },
+    {
+      "epoch": 0.020657995409334353,
+      "grad_norm": 4.889279260558998,
+      "learning_rate": 6.870229007633589e-06,
+      "loss": 0.7327,
+      "step": 135
+    },
+    {
+      "epoch": 0.020811017597551645,
+      "grad_norm": 4.256993693098611,
+      "learning_rate": 6.921119592875318e-06,
+      "loss": 0.8079,
+      "step": 136
+    },
+    {
+      "epoch": 0.020964039785768936,
+      "grad_norm": 3.907777183976378,
+      "learning_rate": 6.972010178117049e-06,
+      "loss": 0.7274,
+      "step": 137
+    },
+    {
+      "epoch": 0.021117061973986228,
+      "grad_norm": 5.169921684229104,
+      "learning_rate": 7.022900763358779e-06,
+      "loss": 0.7762,
+      "step": 138
+    },
+    {
+      "epoch": 0.02127008416220352,
+      "grad_norm": 3.705954278589657,
+      "learning_rate": 7.073791348600509e-06,
+      "loss": 0.7769,
+      "step": 139
+    },
+    {
+      "epoch": 0.02142310635042081,
+      "grad_norm": 4.415407291310754,
+      "learning_rate": 7.12468193384224e-06,
+      "loss": 0.7743,
+      "step": 140
+    },
+    {
+      "epoch": 0.021576128538638103,
+      "grad_norm": 3.9001765524312577,
+      "learning_rate": 7.17557251908397e-06,
+      "loss": 0.7702,
+      "step": 141
+    },
+    {
+      "epoch": 0.021729150726855394,
+      "grad_norm": 4.010966551248577,
+      "learning_rate": 7.2264631043257004e-06,
+      "loss": 0.794,
+      "step": 142
+    },
+    {
+      "epoch": 0.021882172915072686,
+      "grad_norm": 4.547219140713333,
+      "learning_rate": 7.27735368956743e-06,
+      "loss": 0.8035,
+      "step": 143
+    },
+    {
+      "epoch": 0.022035195103289978,
+      "grad_norm": 3.895437930156407,
+      "learning_rate": 7.328244274809161e-06,
+      "loss": 0.7342,
+      "step": 144
+    },
+    {
+      "epoch": 0.02218821729150727,
+      "grad_norm": 3.7831765310187784,
+      "learning_rate": 7.3791348600508915e-06,
+      "loss": 0.7207,
+      "step": 145
+    },
+    {
+      "epoch": 0.02234123947972456,
+      "grad_norm": 4.02944403589167,
+      "learning_rate": 7.430025445292621e-06,
+      "loss": 0.6893,
+      "step": 146
+    },
+    {
+      "epoch": 0.022494261667941853,
+      "grad_norm": 3.955758005563262,
+      "learning_rate": 7.480916030534352e-06,
+      "loss": 0.744,
+      "step": 147
+    },
+    {
+      "epoch": 0.022647283856159144,
+      "grad_norm": 4.067883709052432,
+      "learning_rate": 7.531806615776082e-06,
+      "loss": 0.755,
+      "step": 148
+    },
+    {
+      "epoch": 0.022800306044376436,
+      "grad_norm": 4.2402099809259,
+      "learning_rate": 7.582697201017812e-06,
+      "loss": 0.7422,
+      "step": 149
+    },
+    {
+      "epoch": 0.022953328232593728,
+      "grad_norm": 4.262612871583228,
+      "learning_rate": 7.633587786259543e-06,
+      "loss": 0.8447,
+      "step": 150
+    },
+    {
+      "epoch": 0.02310635042081102,
+      "grad_norm": 3.7947232149122043,
+      "learning_rate": 7.684478371501274e-06,
+      "loss": 0.745,
+      "step": 151
+    },
+    {
+      "epoch": 0.02325937260902831,
+      "grad_norm": 4.143133247827625,
+      "learning_rate": 7.735368956743004e-06,
+      "loss": 0.8246,
+      "step": 152
+    },
+    {
+      "epoch": 0.0234123947972456,
+      "grad_norm": 3.9840463660552916,
+      "learning_rate": 7.786259541984733e-06,
+      "loss": 0.817,
+      "step": 153
+    },
+    {
+      "epoch": 0.02356541698546289,
+      "grad_norm": 4.101939658807749,
+      "learning_rate": 7.837150127226465e-06,
+      "loss": 0.8084,
+      "step": 154
+    },
+    {
+      "epoch": 0.023718439173680182,
+      "grad_norm": 4.061501850729827,
+      "learning_rate": 7.888040712468195e-06,
+      "loss": 0.8093,
+      "step": 155
+    },
+    {
+      "epoch": 0.023871461361897474,
+      "grad_norm": 4.822404469329433,
+      "learning_rate": 7.938931297709924e-06,
+      "loss": 0.8627,
+      "step": 156
+    },
+    {
+      "epoch": 0.024024483550114765,
+      "grad_norm": 4.64280647954761,
+      "learning_rate": 7.989821882951656e-06,
+      "loss": 0.7448,
+      "step": 157
+    },
+    {
+      "epoch": 0.024177505738332057,
+      "grad_norm": 4.03108430157329,
+      "learning_rate": 8.040712468193384e-06,
+      "loss": 0.7671,
+      "step": 158
+    },
+    {
+      "epoch": 0.02433052792654935,
+      "grad_norm": 4.4845780168203575,
+      "learning_rate": 8.091603053435115e-06,
+      "loss": 0.813,
+      "step": 159
+    },
+    {
+      "epoch": 0.02448355011476664,
+      "grad_norm": 4.345621286114522,
+      "learning_rate": 8.142493638676845e-06,
+      "loss": 0.7408,
+      "step": 160
+    },
+    {
+      "epoch": 0.024636572302983932,
+      "grad_norm": 4.6919797838381605,
+      "learning_rate": 8.193384223918575e-06,
+      "loss": 0.8106,
+      "step": 161
+    },
+    {
+      "epoch": 0.024789594491201224,
+      "grad_norm": 4.333359656214333,
+      "learning_rate": 8.244274809160306e-06,
+      "loss": 0.8115,
+      "step": 162
+    },
+    {
+      "epoch": 0.024942616679418515,
+      "grad_norm": 3.81217702704086,
+      "learning_rate": 8.295165394402036e-06,
+      "loss": 0.7482,
+      "step": 163
+    },
+    {
+      "epoch": 0.025095638867635807,
+      "grad_norm": 4.44804919658539,
+      "learning_rate": 8.346055979643766e-06,
+      "loss": 0.8544,
+      "step": 164
+    },
+    {
+      "epoch": 0.0252486610558531,
+      "grad_norm": 4.131700271663872,
+      "learning_rate": 8.396946564885497e-06,
+      "loss": 0.7262,
+      "step": 165
+    },
+    {
+      "epoch": 0.02540168324407039,
+      "grad_norm": 4.951792235136878,
+      "learning_rate": 8.447837150127227e-06,
+      "loss": 0.853,
+      "step": 166
+    },
+    {
+      "epoch": 0.025554705432287682,
+      "grad_norm": 4.326303123701613,
+      "learning_rate": 8.498727735368957e-06,
+      "loss": 0.7414,
+      "step": 167
+    },
+    {
+      "epoch": 0.025707727620504973,
+      "grad_norm": 3.955742466785765,
+      "learning_rate": 8.549618320610688e-06,
+      "loss": 0.6963,
+      "step": 168
+    },
+    {
+      "epoch": 0.025860749808722265,
+      "grad_norm": 3.9376734300269995,
+      "learning_rate": 8.600508905852418e-06,
+      "loss": 0.655,
+      "step": 169
+    },
+    {
+      "epoch": 0.026013771996939557,
+      "grad_norm": 5.053922872384057,
+      "learning_rate": 8.651399491094148e-06,
+      "loss": 0.8598,
+      "step": 170
+    },
+    {
+      "epoch": 0.02616679418515685,
+      "grad_norm": 3.940496090207759,
+      "learning_rate": 8.702290076335879e-06,
+      "loss": 0.7512,
+      "step": 171
+    },
+    {
+      "epoch": 0.02631981637337414,
+      "grad_norm": 4.22917739285267,
+      "learning_rate": 8.753180661577609e-06,
+      "loss": 0.9123,
+      "step": 172
+    },
+    {
+      "epoch": 0.02647283856159143,
+      "grad_norm": 4.508693336682513,
+      "learning_rate": 8.80407124681934e-06,
+      "loss": 0.8793,
+      "step": 173
+    },
+    {
+      "epoch": 0.026625860749808723,
+      "grad_norm": 4.233775278965739,
+      "learning_rate": 8.85496183206107e-06,
+      "loss": 0.8819,
+      "step": 174
+    },
+    {
+      "epoch": 0.026778882938026015,
+      "grad_norm": 3.7511210956873544,
+      "learning_rate": 8.9058524173028e-06,
+      "loss": 0.7657,
+      "step": 175
+    },
+    {
+      "epoch": 0.026931905126243307,
+      "grad_norm": 4.143781255150889,
+      "learning_rate": 8.95674300254453e-06,
+      "loss": 0.7149,
+      "step": 176
+    },
+    {
+      "epoch": 0.027084927314460598,
+      "grad_norm": 4.671966201492468,
+      "learning_rate": 9.007633587786259e-06,
+      "loss": 0.9099,
+      "step": 177
+    },
+    {
+      "epoch": 0.02723794950267789,
+      "grad_norm": 3.8699485585110627,
+      "learning_rate": 9.058524173027991e-06,
+      "loss": 0.7838,
+      "step": 178
+    },
+    {
+      "epoch": 0.02739097169089518,
+      "grad_norm": 3.7570884060183305,
+      "learning_rate": 9.109414758269721e-06,
+      "loss": 0.8138,
+      "step": 179
+    },
+    {
+      "epoch": 0.02754399387911247,
+      "grad_norm": 4.294913936841194,
+      "learning_rate": 9.16030534351145e-06,
+      "loss": 0.8253,
+      "step": 180
+    },
+    {
+      "epoch": 0.02769701606732976,
+      "grad_norm": 4.424497607816517,
+      "learning_rate": 9.211195928753182e-06,
+      "loss": 0.7881,
+      "step": 181
+    },
+    {
+      "epoch": 0.027850038255547053,
+      "grad_norm": 3.9878850555273604,
+      "learning_rate": 9.26208651399491e-06,
+      "loss": 0.7683,
+      "step": 182
+    },
+    {
+      "epoch": 0.028003060443764344,
+      "grad_norm": 4.214766952361433,
+      "learning_rate": 9.312977099236641e-06,
+      "loss": 0.811,
+      "step": 183
+    },
+    {
+      "epoch": 0.028156082631981636,
+      "grad_norm": 4.299160854226878,
+      "learning_rate": 9.363867684478373e-06,
+      "loss": 0.8099,
+      "step": 184
+    },
+    {
+      "epoch": 0.028309104820198928,
+      "grad_norm": 4.261601197119955,
+      "learning_rate": 9.414758269720102e-06,
+      "loss": 0.8327,
+      "step": 185
+    },
+    {
+      "epoch": 0.02846212700841622,
+      "grad_norm": 4.348418153113838,
+      "learning_rate": 9.465648854961834e-06,
+      "loss": 0.8351,
+      "step": 186
+    },
+    {
+      "epoch": 0.02861514919663351,
+      "grad_norm": 4.159462607341161,
+      "learning_rate": 9.516539440203563e-06,
+      "loss": 0.8109,
+      "step": 187
+    },
+    {
+      "epoch": 0.028768171384850803,
+      "grad_norm": 4.139615856283388,
+      "learning_rate": 9.567430025445293e-06,
+      "loss": 0.808,
+      "step": 188
+    },
+    {
+      "epoch": 0.028921193573068094,
+      "grad_norm": 3.966142369811451,
+      "learning_rate": 9.618320610687025e-06,
+      "loss": 0.8221,
+      "step": 189
+    },
+    {
+      "epoch": 0.029074215761285386,
+      "grad_norm": 4.2883136258405585,
+      "learning_rate": 9.669211195928754e-06,
+      "loss": 0.7766,
+      "step": 190
+    },
+    {
+      "epoch": 0.029227237949502678,
+      "grad_norm": 3.5038128496057706,
+      "learning_rate": 9.720101781170484e-06,
+      "loss": 0.802,
+      "step": 191
+    },
+    {
+      "epoch": 0.02938026013771997,
+      "grad_norm": 4.018460389006565,
+      "learning_rate": 9.770992366412214e-06,
+      "loss": 0.8793,
+      "step": 192
+    },
+    {
+      "epoch": 0.02953328232593726,
+      "grad_norm": 4.197653542836699,
+      "learning_rate": 9.821882951653945e-06,
+      "loss": 0.7636,
+      "step": 193
+    },
+    {
+      "epoch": 0.029686304514154552,
+      "grad_norm": 4.3491719420032275,
+      "learning_rate": 9.872773536895675e-06,
+      "loss": 0.8217,
+      "step": 194
+    },
+    {
+      "epoch": 0.029839326702371844,
+      "grad_norm": 3.896319127881109,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 0.6789,
+      "step": 195
+    },
+    {
+      "epoch": 0.029992348890589136,
+      "grad_norm": 4.135458574765948,
+      "learning_rate": 9.974554707379136e-06,
+      "loss": 0.7865,
+      "step": 196
+    },
+    {
+      "epoch": 0.030145371078806427,
+      "grad_norm": 4.497645124133756,
+      "learning_rate": 1.0025445292620866e-05,
+      "loss": 0.8283,
+      "step": 197
+    },
+    {
+      "epoch": 0.03029839326702372,
+      "grad_norm": 3.5421499415575246,
+      "learning_rate": 1.0076335877862595e-05,
+      "loss": 0.9806,
+      "step": 198
+    },
+    {
+      "epoch": 0.03045141545524101,
+      "grad_norm": 3.5582689056601446,
+      "learning_rate": 1.0127226463104327e-05,
+      "loss": 0.7229,
+      "step": 199
+    },
+    {
+      "epoch": 0.030604437643458302,
+      "grad_norm": 4.505265157757878,
+      "learning_rate": 1.0178117048346057e-05,
+      "loss": 0.8239,
+      "step": 200
+    },
+    {
+      "epoch": 0.030757459831675594,
+      "grad_norm": 3.785002672226262,
+      "learning_rate": 1.0229007633587786e-05,
+      "loss": 0.8285,
+      "step": 201
+    },
+    {
+      "epoch": 0.030910482019892885,
+      "grad_norm": 4.056563880467215,
+      "learning_rate": 1.0279898218829518e-05,
+      "loss": 0.8093,
+      "step": 202
+    },
+    {
+      "epoch": 0.031063504208110177,
+      "grad_norm": 3.8406342164711114,
+      "learning_rate": 1.0330788804071248e-05,
+      "loss": 0.8487,
+      "step": 203
+    },
+    {
+      "epoch": 0.03121652639632747,
+      "grad_norm": 4.52064054660571,
+      "learning_rate": 1.0381679389312977e-05,
+      "loss": 0.7581,
+      "step": 204
+    },
+    {
+      "epoch": 0.03136954858454476,
+      "grad_norm": 3.6506482696933977,
+      "learning_rate": 1.0432569974554709e-05,
+      "loss": 0.7612,
+      "step": 205
+    },
+    {
+      "epoch": 0.03152257077276205,
+      "grad_norm": 4.383431897980942,
+      "learning_rate": 1.0483460559796439e-05,
+      "loss": 0.8776,
+      "step": 206
+    },
+    {
+      "epoch": 0.03167559296097934,
+      "grad_norm": 3.741616408368712,
+      "learning_rate": 1.0534351145038168e-05,
+      "loss": 0.7634,
+      "step": 207
+    },
+    {
+      "epoch": 0.03182861514919663,
+      "grad_norm": 4.448598087029719,
+      "learning_rate": 1.0585241730279898e-05,
+      "loss": 0.8928,
+      "step": 208
+    },
+    {
+      "epoch": 0.03198163733741392,
+      "grad_norm": 3.7260572298800136,
+      "learning_rate": 1.063613231552163e-05,
+      "loss": 0.7409,
+      "step": 209
+    },
+    {
+      "epoch": 0.032134659525631215,
+      "grad_norm": 4.082446929366836,
+      "learning_rate": 1.068702290076336e-05,
+      "loss": 0.7702,
+      "step": 210
+    },
+    {
+      "epoch": 0.03228768171384851,
+      "grad_norm": 4.276111120838432,
+      "learning_rate": 1.0737913486005089e-05,
+      "loss": 0.8451,
+      "step": 211
+    },
+    {
+      "epoch": 0.0324407039020658,
+      "grad_norm": 3.5982082879998303,
+      "learning_rate": 1.0788804071246821e-05,
+      "loss": 0.7412,
+      "step": 212
+    },
+    {
+      "epoch": 0.03259372609028309,
+      "grad_norm": 3.7182399171777427,
+      "learning_rate": 1.0839694656488552e-05,
+      "loss": 0.7734,
+      "step": 213
+    },
+    {
+      "epoch": 0.03274674827850038,
+      "grad_norm": 4.45090671467595,
+      "learning_rate": 1.089058524173028e-05,
+      "loss": 0.7803,
+      "step": 214
+    },
+    {
+      "epoch": 0.03289977046671767,
+      "grad_norm": 4.10877525262828,
+      "learning_rate": 1.094147582697201e-05,
+      "loss": 0.8259,
+      "step": 215
+    },
+    {
+      "epoch": 0.033052792654934965,
+      "grad_norm": 3.6500837791600693,
+      "learning_rate": 1.0992366412213743e-05,
+      "loss": 0.789,
+      "step": 216
+    },
+    {
+      "epoch": 0.033205814843152257,
+      "grad_norm": 3.5491002374247267,
+      "learning_rate": 1.1043256997455471e-05,
+      "loss": 0.7059,
+      "step": 217
+    },
+    {
+      "epoch": 0.03335883703136955,
+      "grad_norm": 3.4521743275023624,
+      "learning_rate": 1.1094147582697202e-05,
+      "loss": 0.7247,
+      "step": 218
+    },
+    {
+      "epoch": 0.03351185921958684,
+      "grad_norm": 3.8737746868152994,
+      "learning_rate": 1.1145038167938934e-05,
+      "loss": 0.7646,
+      "step": 219
+    },
+    {
+      "epoch": 0.03366488140780413,
+      "grad_norm": 3.8562641801906192,
+      "learning_rate": 1.1195928753180662e-05,
+      "loss": 0.8318,
+      "step": 220
+    },
+    {
+      "epoch": 0.03381790359602142,
+      "grad_norm": 3.912581720585126,
+      "learning_rate": 1.1246819338422393e-05,
+      "loss": 0.8093,
+      "step": 221
+    },
+    {
+      "epoch": 0.033970925784238715,
+      "grad_norm": 4.065770061675511,
+      "learning_rate": 1.1297709923664125e-05,
+      "loss": 0.7937,
+      "step": 222
+    },
+    {
+      "epoch": 0.034123947972456006,
+      "grad_norm": 3.6237024117778986,
+      "learning_rate": 1.1348600508905853e-05,
+      "loss": 0.7211,
+      "step": 223
+    },
+    {
+      "epoch": 0.0342769701606733,
+      "grad_norm": 3.199937206085619,
+      "learning_rate": 1.1399491094147584e-05,
+      "loss": 0.6572,
+      "step": 224
+    },
+    {
+      "epoch": 0.03442999234889059,
+      "grad_norm": 3.9340874687404392,
+      "learning_rate": 1.1450381679389312e-05,
+      "loss": 0.7167,
+      "step": 225
+    },
+    {
+      "epoch": 0.03458301453710788,
+      "grad_norm": 4.280709448985111,
+      "learning_rate": 1.1501272264631044e-05,
+      "loss": 0.8466,
+      "step": 226
+    },
+    {
+      "epoch": 0.03473603672532517,
+      "grad_norm": 4.079222952395199,
+      "learning_rate": 1.1552162849872775e-05,
+      "loss": 0.7506,
+      "step": 227
+    },
+    {
+      "epoch": 0.034889058913542464,
+      "grad_norm": 3.5818500329635308,
+      "learning_rate": 1.1603053435114503e-05,
+      "loss": 0.8351,
+      "step": 228
+    },
+    {
+      "epoch": 0.035042081101759756,
+      "grad_norm": 3.683692326325265,
+      "learning_rate": 1.1653944020356235e-05,
+      "loss": 0.8308,
+      "step": 229
+    },
+    {
+      "epoch": 0.03519510328997705,
+      "grad_norm": 4.314842659628774,
+      "learning_rate": 1.1704834605597966e-05,
+      "loss": 0.863,
+      "step": 230
+    },
+    {
+      "epoch": 0.03534812547819434,
+      "grad_norm": 3.8577935082344057,
+      "learning_rate": 1.1755725190839696e-05,
+      "loss": 0.7947,
+      "step": 231
+    },
+    {
+      "epoch": 0.03550114766641163,
+      "grad_norm": 4.001739000653889,
+      "learning_rate": 1.1806615776081425e-05,
+      "loss": 0.8155,
+      "step": 232
+    },
+    {
+      "epoch": 0.03565416985462892,
+      "grad_norm": 4.413074080050747,
+      "learning_rate": 1.1857506361323157e-05,
+      "loss": 0.8612,
+      "step": 233
+    },
+    {
+      "epoch": 0.035807192042846214,
+      "grad_norm": 3.7562800982675433,
+      "learning_rate": 1.1908396946564887e-05,
+      "loss": 0.7496,
+      "step": 234
+    },
+    {
+      "epoch": 0.035960214231063506,
+      "grad_norm": 4.676164618927764,
+      "learning_rate": 1.1959287531806616e-05,
+      "loss": 0.7961,
+      "step": 235
+    },
+    {
+      "epoch": 0.0361132364192808,
+      "grad_norm": 4.1865326091807376,
+      "learning_rate": 1.2010178117048348e-05,
+      "loss": 0.7285,
+      "step": 236
+    },
+    {
+      "epoch": 0.03626625860749809,
+      "grad_norm": 4.330640395701582,
+      "learning_rate": 1.2061068702290078e-05,
+      "loss": 0.86,
+      "step": 237
+    },
+    {
+      "epoch": 0.03641928079571538,
+      "grad_norm": 4.221924703378774,
+      "learning_rate": 1.2111959287531807e-05,
+      "loss": 0.8088,
+      "step": 238
+    },
+    {
+      "epoch": 0.03657230298393267,
+      "grad_norm": 4.08774124838906,
+      "learning_rate": 1.2162849872773539e-05,
+      "loss": 0.7768,
+      "step": 239
+    },
+    {
+      "epoch": 0.036725325172149964,
+      "grad_norm": 3.6681049073966387,
+      "learning_rate": 1.2213740458015269e-05,
+      "loss": 0.8089,
+      "step": 240
+    },
+    {
+      "epoch": 0.036878347360367256,
+      "grad_norm": 4.339040545843134,
+      "learning_rate": 1.2264631043256998e-05,
+      "loss": 0.9708,
+      "step": 241
+    },
+    {
+      "epoch": 0.03703136954858455,
+      "grad_norm": 4.090377047572124,
+      "learning_rate": 1.2315521628498728e-05,
+      "loss": 0.7973,
+      "step": 242
+    },
+    {
+      "epoch": 0.03718439173680184,
+      "grad_norm": 3.8573977126801324,
+      "learning_rate": 1.236641221374046e-05,
+      "loss": 0.8493,
+      "step": 243
+    },
+    {
+      "epoch": 0.03733741392501913,
+      "grad_norm": 3.8417101718993036,
+      "learning_rate": 1.2417302798982189e-05,
+      "loss": 0.7898,
+      "step": 244
+    },
+    {
+      "epoch": 0.03749043611323642,
+      "grad_norm": 4.167929367567143,
+      "learning_rate": 1.2468193384223919e-05,
+      "loss": 0.8347,
+      "step": 245
+    },
+    {
+      "epoch": 0.037643458301453714,
+      "grad_norm": 4.304125499094029,
+      "learning_rate": 1.2519083969465651e-05,
+      "loss": 0.9077,
+      "step": 246
+    },
+    {
+      "epoch": 0.037796480489671005,
+      "grad_norm": 3.7732211137563523,
+      "learning_rate": 1.256997455470738e-05,
+      "loss": 0.7202,
+      "step": 247
+    },
+    {
+      "epoch": 0.0379495026778883,
+      "grad_norm": 4.028839146378367,
+      "learning_rate": 1.262086513994911e-05,
+      "loss": 0.9032,
+      "step": 248
+    },
+    {
+      "epoch": 0.03810252486610559,
+      "grad_norm": 3.571061110080041,
+      "learning_rate": 1.2671755725190839e-05,
+      "loss": 0.8451,
+      "step": 249
+    },
+    {
+      "epoch": 0.03825554705432287,
+      "grad_norm": 4.231210681726749,
+      "learning_rate": 1.2722646310432571e-05,
+      "loss": 0.8744,
+      "step": 250
+    },
+    {
+      "epoch": 0.038408569242540165,
+      "grad_norm": 3.8963900051843092,
+      "learning_rate": 1.2773536895674301e-05,
+      "loss": 0.8427,
+      "step": 251
+    },
+    {
+      "epoch": 0.03856159143075746,
+      "grad_norm": 3.983079418182341,
+      "learning_rate": 1.2824427480916032e-05,
+      "loss": 0.7799,
+      "step": 252
+    },
+    {
+      "epoch": 0.03871461361897475,
+      "grad_norm": 3.3692269874643967,
+      "learning_rate": 1.2875318066157762e-05,
+      "loss": 0.7349,
+      "step": 253
+    },
+    {
+      "epoch": 0.03886763580719204,
+      "grad_norm": 4.081444797385999,
+      "learning_rate": 1.2926208651399492e-05,
+      "loss": 0.8248,
+      "step": 254
+    },
+    {
+      "epoch": 0.03902065799540933,
+      "grad_norm": 3.5071517749127143,
+      "learning_rate": 1.2977099236641223e-05,
+      "loss": 0.7729,
+      "step": 255
+    },
+    {
+      "epoch": 0.03917368018362662,
+      "grad_norm": 3.983554583470299,
+      "learning_rate": 1.3027989821882953e-05,
+      "loss": 0.7358,
+      "step": 256
+    },
+    {
+      "epoch": 0.039326702371843915,
+      "grad_norm": 4.156162140200395,
+      "learning_rate": 1.3078880407124683e-05,
+      "loss": 0.7827,
+      "step": 257
+    },
+    {
+      "epoch": 0.039479724560061207,
+      "grad_norm": 3.7356062298023005,
+      "learning_rate": 1.3129770992366414e-05,
+      "loss": 0.8187,
+      "step": 258
+    },
+    {
+      "epoch": 0.0396327467482785,
+      "grad_norm": 4.41491523535067,
+      "learning_rate": 1.3180661577608142e-05,
+      "loss": 0.9266,
+      "step": 259
+    },
+    {
+      "epoch": 0.03978576893649579,
+      "grad_norm": 3.823408041477082,
+      "learning_rate": 1.3231552162849874e-05,
+      "loss": 0.7646,
+      "step": 260
+    },
+    {
+      "epoch": 0.03993879112471308,
+      "grad_norm": 4.187964678105524,
+      "learning_rate": 1.3282442748091605e-05,
+      "loss": 0.7775,
+      "step": 261
+    },
+    {
+      "epoch": 0.04009181331293037,
+      "grad_norm": 3.677752698776951,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7735,
+      "step": 262
+    },
+    {
+      "epoch": 0.040244835501147665,
+      "grad_norm": 4.163618150951032,
+      "learning_rate": 1.3384223918575065e-05,
+      "loss": 0.8633,
+      "step": 263
+    },
+    {
+      "epoch": 0.040397857689364956,
+      "grad_norm": 4.340339901550398,
+      "learning_rate": 1.3435114503816796e-05,
+      "loss": 0.8143,
+      "step": 264
+    },
+    {
+      "epoch": 0.04055087987758225,
+      "grad_norm": 3.779345498739458,
+      "learning_rate": 1.3486005089058524e-05,
+      "loss": 0.7737,
+      "step": 265
+    },
+    {
+      "epoch": 0.04070390206579954,
+      "grad_norm": 3.4660216974591944,
+      "learning_rate": 1.3536895674300256e-05,
+      "loss": 0.7818,
+      "step": 266
+    },
+    {
+      "epoch": 0.04085692425401683,
+      "grad_norm": 3.460945209038202,
+      "learning_rate": 1.3587786259541987e-05,
+      "loss": 0.8819,
+      "step": 267
+    },
+    {
+      "epoch": 0.04100994644223412,
+      "grad_norm": 3.174432254917918,
+      "learning_rate": 1.3638676844783715e-05,
+      "loss": 0.8147,
+      "step": 268
+    },
+    {
+      "epoch": 0.041162968630451414,
+      "grad_norm": 3.6606958430717436,
+      "learning_rate": 1.3689567430025446e-05,
+      "loss": 0.7882,
+      "step": 269
+    },
+    {
+      "epoch": 0.041315990818668706,
+      "grad_norm": 3.4598901409929566,
+      "learning_rate": 1.3740458015267178e-05,
+      "loss": 0.7396,
+      "step": 270
+    },
+    {
+      "epoch": 0.041469013006886,
+      "grad_norm": 3.837680859754261,
+      "learning_rate": 1.3791348600508906e-05,
+      "loss": 0.9517,
+      "step": 271
+    },
+    {
+      "epoch": 0.04162203519510329,
+      "grad_norm": 4.205620824295019,
+      "learning_rate": 1.3842239185750637e-05,
+      "loss": 0.8136,
+      "step": 272
+    },
+    {
+      "epoch": 0.04177505738332058,
+      "grad_norm": 4.593353976664115,
+      "learning_rate": 1.3893129770992369e-05,
+      "loss": 0.8764,
+      "step": 273
+    },
+    {
+      "epoch": 0.04192807957153787,
+      "grad_norm": 3.674973105300058,
+      "learning_rate": 1.3944020356234097e-05,
+      "loss": 0.7451,
+      "step": 274
+    },
+    {
+      "epoch": 0.042081101759755164,
+      "grad_norm": 4.080257867421216,
+      "learning_rate": 1.3994910941475828e-05,
+      "loss": 0.938,
+      "step": 275
+    },
+    {
+      "epoch": 0.042234123947972456,
+      "grad_norm": 3.6825330496591513,
+      "learning_rate": 1.4045801526717558e-05,
+      "loss": 0.9145,
+      "step": 276
+    },
+    {
+      "epoch": 0.04238714613618975,
+      "grad_norm": 3.6676110960298263,
+      "learning_rate": 1.4096692111959288e-05,
+      "loss": 0.8281,
+      "step": 277
+    },
+    {
+      "epoch": 0.04254016832440704,
+      "grad_norm": 3.9692318784638876,
+      "learning_rate": 1.4147582697201019e-05,
+      "loss": 0.8391,
+      "step": 278
+    },
+    {
+      "epoch": 0.04269319051262433,
+      "grad_norm": 4.003302405983541,
+      "learning_rate": 1.4198473282442749e-05,
+      "loss": 0.8807,
+      "step": 279
+    },
+    {
+      "epoch": 0.04284621270084162,
+      "grad_norm": 3.6051890207501125,
+      "learning_rate": 1.424936386768448e-05,
+      "loss": 0.8002,
+      "step": 280
+    },
+    {
+      "epoch": 0.042999234889058914,
+      "grad_norm": 3.741199905910224,
+      "learning_rate": 1.430025445292621e-05,
+      "loss": 0.9184,
+      "step": 281
+    },
+    {
+      "epoch": 0.043152257077276206,
+      "grad_norm": 3.989340278328752,
+      "learning_rate": 1.435114503816794e-05,
+      "loss": 0.8947,
+      "step": 282
+    },
+    {
+      "epoch": 0.0433052792654935,
+      "grad_norm": 3.64277436441139,
+      "learning_rate": 1.4402035623409672e-05,
+      "loss": 0.8931,
+      "step": 283
+    },
+    {
+      "epoch": 0.04345830145371079,
+      "grad_norm": 4.1014237450789635,
+      "learning_rate": 1.4452926208651401e-05,
+      "loss": 0.8706,
+      "step": 284
+    },
+    {
+      "epoch": 0.04361132364192808,
+      "grad_norm": 3.4366912250691297,
+      "learning_rate": 1.4503816793893131e-05,
+      "loss": 0.8406,
+      "step": 285
+    },
+    {
+      "epoch": 0.04376434583014537,
+      "grad_norm": 3.580030889617023,
+      "learning_rate": 1.455470737913486e-05,
+      "loss": 0.8221,
+      "step": 286
+    },
+    {
+      "epoch": 0.043917368018362664,
+      "grad_norm": 3.298085781868641,
+      "learning_rate": 1.4605597964376592e-05,
+      "loss": 0.7627,
+      "step": 287
+    },
+    {
+      "epoch": 0.044070390206579955,
+      "grad_norm": 3.6026238179183316,
+      "learning_rate": 1.4656488549618322e-05,
+      "loss": 0.8868,
+      "step": 288
+    },
+    {
+      "epoch": 0.04422341239479725,
+      "grad_norm": 3.5628106762524663,
+      "learning_rate": 1.4707379134860051e-05,
+      "loss": 0.8383,
+      "step": 289
+    },
+    {
+      "epoch": 0.04437643458301454,
+      "grad_norm": 4.098960888954263,
+      "learning_rate": 1.4758269720101783e-05,
+      "loss": 0.8245,
+      "step": 290
+    },
+    {
+      "epoch": 0.04452945677123183,
+      "grad_norm": 4.019679742440305,
+      "learning_rate": 1.4809160305343513e-05,
+      "loss": 0.9617,
+      "step": 291
+    },
+    {
+      "epoch": 0.04468247895944912,
+      "grad_norm": 3.870786617215219,
+      "learning_rate": 1.4860050890585242e-05,
+      "loss": 0.7536,
+      "step": 292
+    },
+    {
+      "epoch": 0.044835501147666414,
+      "grad_norm": 3.624983481704036,
+      "learning_rate": 1.4910941475826972e-05,
+      "loss": 0.7665,
+      "step": 293
+    },
+    {
+      "epoch": 0.044988523335883705,
+      "grad_norm": 3.4579342237989863,
+      "learning_rate": 1.4961832061068704e-05,
+      "loss": 0.7475,
+      "step": 294
+    },
+    {
+      "epoch": 0.045141545524101,
+      "grad_norm": 3.825987018762197,
+      "learning_rate": 1.5012722646310433e-05,
+      "loss": 0.7701,
+      "step": 295
+    },
+    {
+      "epoch": 0.04529456771231829,
+      "grad_norm": 3.925285067712268,
+      "learning_rate": 1.5063613231552163e-05,
+      "loss": 0.8424,
+      "step": 296
+    },
+    {
+      "epoch": 0.04544758990053558,
+      "grad_norm": 4.536531873258212,
+      "learning_rate": 1.5114503816793895e-05,
+      "loss": 0.8194,
+      "step": 297
+    },
+    {
+      "epoch": 0.04560061208875287,
+      "grad_norm": 3.849198877547217,
+      "learning_rate": 1.5165394402035624e-05,
+      "loss": 0.8097,
+      "step": 298
+    },
+    {
+      "epoch": 0.04575363427697016,
+      "grad_norm": 3.4268881311529,
+      "learning_rate": 1.5216284987277354e-05,
+      "loss": 0.7637,
+      "step": 299
+    },
+    {
+      "epoch": 0.045906656465187455,
+      "grad_norm": 3.844837610723275,
+      "learning_rate": 1.5267175572519086e-05,
+      "loss": 0.9146,
+      "step": 300
+    },
+    {
+      "epoch": 0.04605967865340475,
+      "grad_norm": 4.200860333219284,
+      "learning_rate": 1.5318066157760817e-05,
+      "loss": 0.8694,
+      "step": 301
+    },
+    {
+      "epoch": 0.04621270084162204,
+      "grad_norm": 3.708632369987497,
+      "learning_rate": 1.5368956743002547e-05,
+      "loss": 0.7955,
+      "step": 302
+    },
+    {
+      "epoch": 0.04636572302983933,
+      "grad_norm": 4.560895491616818,
+      "learning_rate": 1.5419847328244274e-05,
+      "loss": 0.8735,
+      "step": 303
+    },
+    {
+      "epoch": 0.04651874521805662,
+      "grad_norm": 4.1460005936560576,
+      "learning_rate": 1.5470737913486008e-05,
+      "loss": 0.9344,
+      "step": 304
+    },
+    {
+      "epoch": 0.046671767406273906,
+      "grad_norm": 3.5819446783097875,
+      "learning_rate": 1.5521628498727738e-05,
+      "loss": 0.8728,
+      "step": 305
+    },
+    {
+      "epoch": 0.0468247895944912,
+      "grad_norm": 3.5673030396101755,
+      "learning_rate": 1.5572519083969465e-05,
+      "loss": 0.8558,
+      "step": 306
+    },
+    {
+      "epoch": 0.04697781178270849,
+      "grad_norm": 4.465799436044937,
+      "learning_rate": 1.56234096692112e-05,
+      "loss": 0.9466,
+      "step": 307
+    },
+    {
+      "epoch": 0.04713083397092578,
+      "grad_norm": 3.8186144985338335,
+      "learning_rate": 1.567430025445293e-05,
+      "loss": 0.8226,
+      "step": 308
+    },
+    {
+      "epoch": 0.04728385615914307,
+      "grad_norm": 3.0526556169711796,
+      "learning_rate": 1.5725190839694656e-05,
+      "loss": 0.8292,
+      "step": 309
+    },
+    {
+      "epoch": 0.047436878347360364,
+      "grad_norm": 3.8395154224833963,
+      "learning_rate": 1.577608142493639e-05,
+      "loss": 0.9308,
+      "step": 310
+    },
+    {
+      "epoch": 0.047589900535577656,
+      "grad_norm": 3.3739437105583048,
+      "learning_rate": 1.582697201017812e-05,
+      "loss": 0.8354,
+      "step": 311
+    },
+    {
+      "epoch": 0.04774292272379495,
+      "grad_norm": 3.428552157235765,
+      "learning_rate": 1.5877862595419847e-05,
+      "loss": 0.7738,
+      "step": 312
+    },
+    {
+      "epoch": 0.04789594491201224,
+      "grad_norm": 3.8083642028444857,
+      "learning_rate": 1.5928753180661577e-05,
+      "loss": 0.8379,
+      "step": 313
+    },
+    {
+      "epoch": 0.04804896710022953,
+      "grad_norm": 3.7472390334350854,
+      "learning_rate": 1.597964376590331e-05,
+      "loss": 0.7635,
+      "step": 314
+    },
+    {
+      "epoch": 0.04820198928844682,
+      "grad_norm": 3.5905420969046977,
+      "learning_rate": 1.6030534351145038e-05,
+      "loss": 0.84,
+      "step": 315
+    },
+    {
+      "epoch": 0.048355011476664114,
+      "grad_norm": 3.371134238680956,
+      "learning_rate": 1.608142493638677e-05,
+      "loss": 0.822,
+      "step": 316
+    },
+    {
+      "epoch": 0.048508033664881406,
+      "grad_norm": 3.251211187350607,
+      "learning_rate": 1.6132315521628502e-05,
+      "loss": 0.8663,
+      "step": 317
+    },
+    {
+      "epoch": 0.0486610558530987,
+      "grad_norm": 3.710129923001892,
+      "learning_rate": 1.618320610687023e-05,
+      "loss": 0.885,
+      "step": 318
+    },
+    {
+      "epoch": 0.04881407804131599,
+      "grad_norm": 3.4804407042961243,
+      "learning_rate": 1.623409669211196e-05,
+      "loss": 0.8388,
+      "step": 319
+    },
+    {
+      "epoch": 0.04896710022953328,
+      "grad_norm": 3.3398691087172763,
+      "learning_rate": 1.628498727735369e-05,
+      "loss": 0.8033,
+      "step": 320
+    },
+    {
+      "epoch": 0.04912012241775057,
+      "grad_norm": 3.4336060940387156,
+      "learning_rate": 1.633587786259542e-05,
+      "loss": 0.8656,
+      "step": 321
+    },
+    {
+      "epoch": 0.049273144605967864,
+      "grad_norm": 3.6774669512700693,
+      "learning_rate": 1.638676844783715e-05,
+      "loss": 0.9224,
+      "step": 322
+    },
+    {
+      "epoch": 0.049426166794185156,
+      "grad_norm": 3.707606788471563,
+      "learning_rate": 1.643765903307888e-05,
+      "loss": 0.8309,
+      "step": 323
+    },
+    {
+      "epoch": 0.04957918898240245,
+      "grad_norm": 3.235739424905745,
+      "learning_rate": 1.648854961832061e-05,
+      "loss": 0.7052,
+      "step": 324
+    },
+    {
+      "epoch": 0.04973221117061974,
+      "grad_norm": 3.7080138914780796,
+      "learning_rate": 1.653944020356234e-05,
+      "loss": 0.8157,
+      "step": 325
+    },
+    {
+      "epoch": 0.04988523335883703,
+      "grad_norm": 3.166803594516287,
+      "learning_rate": 1.6590330788804072e-05,
+      "loss": 0.8085,
+      "step": 326
+    },
+    {
+      "epoch": 0.05003825554705432,
+      "grad_norm": 3.7042887553890647,
+      "learning_rate": 1.6641221374045802e-05,
+      "loss": 0.8331,
+      "step": 327
+    },
+    {
+      "epoch": 0.050191277735271614,
+      "grad_norm": 4.6732971094593125,
+      "learning_rate": 1.6692111959287533e-05,
+      "loss": 0.8703,
+      "step": 328
+    },
+    {
+      "epoch": 0.050344299923488905,
+      "grad_norm": 3.5191964993901137,
+      "learning_rate": 1.6743002544529263e-05,
+      "loss": 0.7757,
+      "step": 329
+    },
+    {
+      "epoch": 0.0504973221117062,
+      "grad_norm": 3.62621311940619,
+      "learning_rate": 1.6793893129770993e-05,
+      "loss": 0.9091,
+      "step": 330
+    },
+    {
+      "epoch": 0.05065034429992349,
+      "grad_norm": 3.4566006756009857,
+      "learning_rate": 1.6844783715012724e-05,
+      "loss": 0.8907,
+      "step": 331
+    },
+    {
+      "epoch": 0.05080336648814078,
+      "grad_norm": 4.226884860942899,
+      "learning_rate": 1.6895674300254454e-05,
+      "loss": 0.9568,
+      "step": 332
+    },
+    {
+      "epoch": 0.05095638867635807,
+      "grad_norm": 3.8334891651194445,
+      "learning_rate": 1.6946564885496184e-05,
+      "loss": 0.802,
+      "step": 333
+    },
+    {
+      "epoch": 0.051109410864575364,
+      "grad_norm": 3.643779521200893,
+      "learning_rate": 1.6997455470737915e-05,
+      "loss": 0.7904,
+      "step": 334
+    },
+    {
+      "epoch": 0.051262433052792655,
+      "grad_norm": 3.6109019840921177,
+      "learning_rate": 1.7048346055979645e-05,
+      "loss": 0.9173,
+      "step": 335
+    },
+    {
+      "epoch": 0.05141545524100995,
+      "grad_norm": 3.556033056937106,
+      "learning_rate": 1.7099236641221375e-05,
+      "loss": 0.6736,
+      "step": 336
+    },
+    {
+      "epoch": 0.05156847742922724,
+      "grad_norm": 3.4590858648740457,
+      "learning_rate": 1.7150127226463106e-05,
+      "loss": 0.7459,
+      "step": 337
+    },
+    {
+      "epoch": 0.05172149961744453,
+      "grad_norm": 3.883487644337928,
+      "learning_rate": 1.7201017811704836e-05,
+      "loss": 0.8293,
+      "step": 338
+    },
+    {
+      "epoch": 0.05187452180566182,
+      "grad_norm": 3.6817700556181308,
+      "learning_rate": 1.7251908396946566e-05,
+      "loss": 0.8543,
+      "step": 339
+    },
+    {
+      "epoch": 0.05202754399387911,
+      "grad_norm": 3.6092941197838107,
+      "learning_rate": 1.7302798982188297e-05,
+      "loss": 0.7518,
+      "step": 340
+    },
+    {
+      "epoch": 0.052180566182096405,
+      "grad_norm": 3.521936034926192,
+      "learning_rate": 1.7353689567430027e-05,
+      "loss": 0.7999,
+      "step": 341
+    },
+    {
+      "epoch": 0.0523335883703137,
+      "grad_norm": 3.2423067286052283,
+      "learning_rate": 1.7404580152671757e-05,
+      "loss": 0.8241,
+      "step": 342
+    },
+    {
+      "epoch": 0.05248661055853099,
+      "grad_norm": 3.5515781332359797,
+      "learning_rate": 1.7455470737913488e-05,
+      "loss": 0.8267,
+      "step": 343
+    },
+    {
+      "epoch": 0.05263963274674828,
+      "grad_norm": 3.361362826153268,
+      "learning_rate": 1.7506361323155218e-05,
+      "loss": 0.8923,
+      "step": 344
+    },
+    {
+      "epoch": 0.05279265493496557,
+      "grad_norm": 4.352751368237682,
+      "learning_rate": 1.755725190839695e-05,
+      "loss": 0.8396,
+      "step": 345
+    },
+    {
+      "epoch": 0.05294567712318286,
+      "grad_norm": 3.405295125419368,
+      "learning_rate": 1.760814249363868e-05,
+      "loss": 0.8015,
+      "step": 346
+    },
+    {
+      "epoch": 0.053098699311400155,
+      "grad_norm": 4.0707162407530255,
+      "learning_rate": 1.765903307888041e-05,
+      "loss": 0.9149,
+      "step": 347
+    },
+    {
+      "epoch": 0.053251721499617446,
+      "grad_norm": 3.199089427504116,
+      "learning_rate": 1.770992366412214e-05,
+      "loss": 0.859,
+      "step": 348
+    },
+    {
+      "epoch": 0.05340474368783474,
+      "grad_norm": 3.3830728370139505,
+      "learning_rate": 1.776081424936387e-05,
+      "loss": 0.9282,
+      "step": 349
+    },
+    {
+      "epoch": 0.05355776587605203,
+      "grad_norm": 3.3892952078432086,
+      "learning_rate": 1.78117048346056e-05,
+      "loss": 0.8592,
+      "step": 350
+    },
+    {
+      "epoch": 0.05371078806426932,
+      "grad_norm": 3.3837045648684434,
+      "learning_rate": 1.786259541984733e-05,
+      "loss": 0.8109,
+      "step": 351
+    },
+    {
+      "epoch": 0.05386381025248661,
+      "grad_norm": 4.01981710203223,
+      "learning_rate": 1.791348600508906e-05,
+      "loss": 0.8086,
+      "step": 352
+    },
+    {
+      "epoch": 0.054016832440703905,
+      "grad_norm": 3.268564561162783,
+      "learning_rate": 1.796437659033079e-05,
+      "loss": 0.8009,
+      "step": 353
+    },
+    {
+      "epoch": 0.054169854628921196,
+      "grad_norm": 3.207034469484103,
+      "learning_rate": 1.8015267175572518e-05,
+      "loss": 0.7459,
+      "step": 354
+    },
+    {
+      "epoch": 0.05432287681713849,
+      "grad_norm": 3.7905236555543698,
+      "learning_rate": 1.8066157760814252e-05,
+      "loss": 0.9694,
+      "step": 355
+    },
+    {
+      "epoch": 0.05447589900535578,
+      "grad_norm": 3.899335977334721,
+      "learning_rate": 1.8117048346055982e-05,
+      "loss": 0.8516,
+      "step": 356
+    },
+    {
+      "epoch": 0.05462892119357307,
+      "grad_norm": 3.6683342704748183,
+      "learning_rate": 1.816793893129771e-05,
+      "loss": 0.8954,
+      "step": 357
+    },
+    {
+      "epoch": 0.05478194338179036,
+      "grad_norm": 3.2948530407738317,
+      "learning_rate": 1.8218829516539443e-05,
+      "loss": 0.8569,
+      "step": 358
+    },
+    {
+      "epoch": 0.054934965570007654,
+      "grad_norm": 3.107869632993156,
+      "learning_rate": 1.8269720101781173e-05,
+      "loss": 0.7997,
+      "step": 359
+    },
+    {
+      "epoch": 0.05508798775822494,
+      "grad_norm": 3.740911956061003,
+      "learning_rate": 1.83206106870229e-05,
+      "loss": 0.7949,
+      "step": 360
+    },
+    {
+      "epoch": 0.05524100994644223,
+      "grad_norm": 3.5260469337243086,
+      "learning_rate": 1.8371501272264634e-05,
+      "loss": 0.8832,
+      "step": 361
+    },
+    {
+      "epoch": 0.05539403213465952,
+      "grad_norm": 4.024532626989753,
+      "learning_rate": 1.8422391857506364e-05,
+      "loss": 0.8864,
+      "step": 362
+    },
+    {
+      "epoch": 0.055547054322876814,
+      "grad_norm": 3.7363845688636372,
+      "learning_rate": 1.847328244274809e-05,
+      "loss": 0.7486,
+      "step": 363
+    },
+    {
+      "epoch": 0.055700076511094106,
+      "grad_norm": 2.989844465377914,
+      "learning_rate": 1.852417302798982e-05,
+      "loss": 0.7627,
+      "step": 364
+    },
+    {
+      "epoch": 0.0558530986993114,
+      "grad_norm": 3.4845633849092814,
+      "learning_rate": 1.8575063613231555e-05,
+      "loss": 0.8626,
+      "step": 365
+    },
+    {
+      "epoch": 0.05600612088752869,
+      "grad_norm": 3.482589493869205,
+      "learning_rate": 1.8625954198473282e-05,
+      "loss": 0.7346,
+      "step": 366
+    },
+    {
+      "epoch": 0.05615914307574598,
+      "grad_norm": 3.2941484050459255,
+      "learning_rate": 1.8676844783715013e-05,
+      "loss": 0.8794,
+      "step": 367
+    },
+    {
+      "epoch": 0.05631216526396327,
+      "grad_norm": 3.7695913912244796,
+      "learning_rate": 1.8727735368956746e-05,
+      "loss": 0.8083,
+      "step": 368
+    },
+    {
+      "epoch": 0.056465187452180564,
+      "grad_norm": 3.247852909314503,
+      "learning_rate": 1.8778625954198473e-05,
+      "loss": 0.7414,
+      "step": 369
+    },
+    {
+      "epoch": 0.056618209640397855,
+      "grad_norm": 3.3192936130223103,
+      "learning_rate": 1.8829516539440204e-05,
+      "loss": 0.9026,
+      "step": 370
+    },
+    {
+      "epoch": 0.05677123182861515,
+      "grad_norm": 3.41682048630867,
+      "learning_rate": 1.8880407124681937e-05,
+      "loss": 0.7388,
+      "step": 371
+    },
+    {
+      "epoch": 0.05692425401683244,
+      "grad_norm": 3.623094017822295,
+      "learning_rate": 1.8931297709923668e-05,
+      "loss": 0.9649,
+      "step": 372
+    },
+    {
+      "epoch": 0.05707727620504973,
+      "grad_norm": 3.7002949571305903,
+      "learning_rate": 1.8982188295165395e-05,
+      "loss": 0.8871,
+      "step": 373
+    },
+    {
+      "epoch": 0.05723029839326702,
+      "grad_norm": 3.325309840511273,
+      "learning_rate": 1.9033078880407125e-05,
+      "loss": 0.9166,
+      "step": 374
+    },
+    {
+      "epoch": 0.057383320581484314,
+      "grad_norm": 3.1858695475335383,
+      "learning_rate": 1.908396946564886e-05,
+      "loss": 0.8516,
+      "step": 375
+    },
+    {
+      "epoch": 0.057536342769701605,
+      "grad_norm": 3.507805903591568,
+      "learning_rate": 1.9134860050890586e-05,
+      "loss": 0.9728,
+      "step": 376
+    },
+    {
+      "epoch": 0.0576893649579189,
+      "grad_norm": 2.953880567303911,
+      "learning_rate": 1.9185750636132316e-05,
+      "loss": 0.9371,
+      "step": 377
+    },
+    {
+      "epoch": 0.05784238714613619,
+      "grad_norm": 3.2343974617308273,
+      "learning_rate": 1.923664122137405e-05,
+      "loss": 0.8903,
+      "step": 378
+    },
+    {
+      "epoch": 0.05799540933435348,
+      "grad_norm": 3.267836515525953,
+      "learning_rate": 1.9287531806615777e-05,
+      "loss": 0.8397,
+      "step": 379
+    },
+    {
+      "epoch": 0.05814843152257077,
+      "grad_norm": 3.594078860448196,
+      "learning_rate": 1.9338422391857507e-05,
+      "loss": 0.9597,
+      "step": 380
+    },
+    {
+      "epoch": 0.05830145371078806,
+      "grad_norm": 3.2708415507993034,
+      "learning_rate": 1.9389312977099238e-05,
+      "loss": 0.8571,
+      "step": 381
+    },
+    {
+      "epoch": 0.058454475899005355,
+      "grad_norm": 3.3418553549010057,
+      "learning_rate": 1.9440203562340968e-05,
+      "loss": 0.8239,
+      "step": 382
+    },
+    {
+      "epoch": 0.05860749808722265,
+      "grad_norm": 3.218696674016876,
+      "learning_rate": 1.9491094147582698e-05,
+      "loss": 0.8126,
+      "step": 383
+    },
+    {
+      "epoch": 0.05876052027543994,
+      "grad_norm": 3.171066381208782,
+      "learning_rate": 1.954198473282443e-05,
+      "loss": 0.757,
+      "step": 384
+    },
+    {
+      "epoch": 0.05891354246365723,
+      "grad_norm": 3.8453943663575982,
+      "learning_rate": 1.959287531806616e-05,
+      "loss": 0.9481,
+      "step": 385
+    },
+    {
+      "epoch": 0.05906656465187452,
+      "grad_norm": 3.262482788883625,
+      "learning_rate": 1.964376590330789e-05,
+      "loss": 0.8465,
+      "step": 386
+    },
+    {
+      "epoch": 0.05921958684009181,
+      "grad_norm": 3.303889728308021,
+      "learning_rate": 1.969465648854962e-05,
+      "loss": 0.7714,
+      "step": 387
+    },
+    {
+      "epoch": 0.059372609028309105,
+      "grad_norm": 3.519976681128204,
+      "learning_rate": 1.974554707379135e-05,
+      "loss": 0.832,
+      "step": 388
+    },
+    {
+      "epoch": 0.059525631216526396,
+      "grad_norm": 3.369259096460689,
+      "learning_rate": 1.979643765903308e-05,
+      "loss": 0.8669,
+      "step": 389
+    },
+    {
+      "epoch": 0.05967865340474369,
+      "grad_norm": 3.7927752076141963,
+      "learning_rate": 1.984732824427481e-05,
+      "loss": 0.846,
+      "step": 390
+    },
+    {
+      "epoch": 0.05983167559296098,
+      "grad_norm": 3.7018297119428523,
+      "learning_rate": 1.989821882951654e-05,
+      "loss": 0.819,
+      "step": 391
+    },
+    {
+      "epoch": 0.05998469778117827,
+      "grad_norm": 3.394158558872775,
+      "learning_rate": 1.994910941475827e-05,
+      "loss": 0.7892,
+      "step": 392
+    },
+    {
+      "epoch": 0.06013771996939556,
+      "grad_norm": 3.276133265082445,
+      "learning_rate": 2e-05,
+      "loss": 0.8333,
+      "step": 393
+    },
+    {
+      "epoch": 0.060290742157612855,
+      "grad_norm": 3.2855946364421804,
+      "learning_rate": 1.999999969293044e-05,
+      "loss": 0.815,
+      "step": 394
+    },
+    {
+      "epoch": 0.060443764345830146,
+      "grad_norm": 3.1185758818586318,
+      "learning_rate": 1.999999877172178e-05,
+      "loss": 0.779,
+      "step": 395
+    },
+    {
+      "epoch": 0.06059678653404744,
+      "grad_norm": 3.25149254683972,
+      "learning_rate": 1.9999997236374075e-05,
+      "loss": 0.83,
+      "step": 396
+    },
+    {
+      "epoch": 0.06074980872226473,
+      "grad_norm": 2.951296265260351,
+      "learning_rate": 1.9999995086887418e-05,
+      "loss": 0.8079,
+      "step": 397
+    },
+    {
+      "epoch": 0.06090283091048202,
+      "grad_norm": 3.0992783869324283,
+      "learning_rate": 1.9999992323261942e-05,
+      "loss": 0.8959,
+      "step": 398
+    },
+    {
+      "epoch": 0.06105585309869931,
+      "grad_norm": 3.3456301521826664,
+      "learning_rate": 1.9999988945497816e-05,
+      "loss": 0.9921,
+      "step": 399
+    },
+    {
+      "epoch": 0.061208875286916604,
+      "grad_norm": 3.569679380617617,
+      "learning_rate": 1.9999984953595253e-05,
+      "loss": 0.9131,
+      "step": 400
+    },
+    {
+      "epoch": 0.061361897475133896,
+      "grad_norm": 3.972081462044736,
+      "learning_rate": 1.999998034755449e-05,
+      "loss": 0.8658,
+      "step": 401
+    },
+    {
+      "epoch": 0.06151491966335119,
+      "grad_norm": 3.1594994773921017,
+      "learning_rate": 1.9999975127375815e-05,
+      "loss": 0.7164,
+      "step": 402
+    },
+    {
+      "epoch": 0.06166794185156848,
+      "grad_norm": 3.003388020711569,
+      "learning_rate": 1.999996929305955e-05,
+      "loss": 0.8885,
+      "step": 403
+    },
+    {
+      "epoch": 0.06182096403978577,
+      "grad_norm": 3.9041072645619974,
+      "learning_rate": 1.9999962844606046e-05,
+      "loss": 0.9423,
+      "step": 404
+    },
+    {
+      "epoch": 0.06197398622800306,
+      "grad_norm": 3.604645472820965,
+      "learning_rate": 1.9999955782015706e-05,
+      "loss": 0.8915,
+      "step": 405
+    },
+    {
+      "epoch": 0.062127008416220354,
+      "grad_norm": 3.0395146768409615,
+      "learning_rate": 1.9999948105288963e-05,
+      "loss": 0.8001,
+      "step": 406
+    },
+    {
+      "epoch": 0.062280030604437646,
+      "grad_norm": 3.3622363017187085,
+      "learning_rate": 1.9999939814426283e-05,
+      "loss": 0.9596,
+      "step": 407
+    },
+    {
+      "epoch": 0.06243305279265494,
+      "grad_norm": 3.290114424412209,
+      "learning_rate": 1.9999930909428183e-05,
+      "loss": 0.8567,
+      "step": 408
+    },
+    {
+      "epoch": 0.06258607498087222,
+      "grad_norm": 3.4438483746832245,
+      "learning_rate": 1.9999921390295203e-05,
+      "loss": 0.9297,
+      "step": 409
+    },
+    {
+      "epoch": 0.06273909716908951,
+      "grad_norm": 3.407358852073633,
+      "learning_rate": 1.999991125702793e-05,
+      "loss": 0.9102,
+      "step": 410
+    },
+    {
+      "epoch": 0.0628921193573068,
+      "grad_norm": 3.9536948948238693,
+      "learning_rate": 1.999990050962699e-05,
+      "loss": 0.7925,
+      "step": 411
+    },
+    {
+      "epoch": 0.0630451415455241,
+      "grad_norm": 3.1915535537530517,
+      "learning_rate": 1.9999889148093036e-05,
+      "loss": 0.7997,
+      "step": 412
+    },
+    {
+      "epoch": 0.06319816373374139,
+      "grad_norm": 3.3935055650270645,
+      "learning_rate": 1.9999877172426775e-05,
+      "loss": 0.7743,
+      "step": 413
+    },
+    {
+      "epoch": 0.06335118592195868,
+      "grad_norm": 3.4416667252947377,
+      "learning_rate": 1.9999864582628932e-05,
+      "loss": 0.8931,
+      "step": 414
+    },
+    {
+      "epoch": 0.06350420811017597,
+      "grad_norm": 3.791402724599172,
+      "learning_rate": 1.9999851378700286e-05,
+      "loss": 0.979,
+      "step": 415
+    },
+    {
+      "epoch": 0.06365723029839326,
+      "grad_norm": 3.1724547503182703,
+      "learning_rate": 1.999983756064165e-05,
+      "loss": 0.7635,
+      "step": 416
+    },
+    {
+      "epoch": 0.06381025248661056,
+      "grad_norm": 3.2774243411313253,
+      "learning_rate": 1.999982312845387e-05,
+      "loss": 0.841,
+      "step": 417
+    },
+    {
+      "epoch": 0.06396327467482785,
+      "grad_norm": 3.1958537216114604,
+      "learning_rate": 1.9999808082137828e-05,
+      "loss": 0.9372,
+      "step": 418
+    },
+    {
+      "epoch": 0.06411629686304514,
+      "grad_norm": 3.5937707452395555,
+      "learning_rate": 1.9999792421694454e-05,
+      "loss": 1.0499,
+      "step": 419
+    },
+    {
+      "epoch": 0.06426931905126243,
+      "grad_norm": 3.0929075141412032,
+      "learning_rate": 1.999977614712471e-05,
+      "loss": 0.9361,
+      "step": 420
+    },
+    {
+      "epoch": 0.06442234123947972,
+      "grad_norm": 3.6154685592166524,
+      "learning_rate": 1.9999759258429592e-05,
+      "loss": 0.8855,
+      "step": 421
+    },
+    {
+      "epoch": 0.06457536342769701,
+      "grad_norm": 3.317458456160721,
+      "learning_rate": 1.9999741755610137e-05,
+      "loss": 0.762,
+      "step": 422
+    },
+    {
+      "epoch": 0.0647283856159143,
+      "grad_norm": 3.552919568915116,
+      "learning_rate": 1.9999723638667424e-05,
+      "loss": 0.9853,
+      "step": 423
+    },
+    {
+      "epoch": 0.0648814078041316,
+      "grad_norm": 3.614389994148897,
+      "learning_rate": 1.9999704907602564e-05,
+      "loss": 0.8873,
+      "step": 424
+    },
+    {
+      "epoch": 0.06503442999234889,
+      "grad_norm": 3.5998150028550064,
+      "learning_rate": 1.9999685562416704e-05,
+      "loss": 0.9252,
+      "step": 425
+    },
+    {
+      "epoch": 0.06518745218056618,
+      "grad_norm": 3.049004959216067,
+      "learning_rate": 1.9999665603111035e-05,
+      "loss": 0.8648,
+      "step": 426
+    },
+    {
+      "epoch": 0.06534047436878347,
+      "grad_norm": 3.1745175809636237,
+      "learning_rate": 1.9999645029686784e-05,
+      "loss": 0.8266,
+      "step": 427
+    },
+    {
+      "epoch": 0.06549349655700076,
+      "grad_norm": 3.4882134719187805,
+      "learning_rate": 1.9999623842145212e-05,
+      "loss": 0.7766,
+      "step": 428
+    },
+    {
+      "epoch": 0.06564651874521805,
+      "grad_norm": 3.506680615344111,
+      "learning_rate": 1.999960204048762e-05,
+      "loss": 0.8902,
+      "step": 429
+    },
+    {
+      "epoch": 0.06579954093343535,
+      "grad_norm": 3.343900859921755,
+      "learning_rate": 1.9999579624715347e-05,
+      "loss": 0.8235,
+      "step": 430
+    },
+    {
+      "epoch": 0.06595256312165264,
+      "grad_norm": 3.506632382776108,
+      "learning_rate": 1.9999556594829775e-05,
+      "loss": 0.8132,
+      "step": 431
+    },
+    {
+      "epoch": 0.06610558530986993,
+      "grad_norm": 3.058113063265004,
+      "learning_rate": 1.9999532950832313e-05,
+      "loss": 0.8041,
+      "step": 432
+    },
+    {
+      "epoch": 0.06625860749808722,
+      "grad_norm": 3.3485252385202715,
+      "learning_rate": 1.999950869272441e-05,
+      "loss": 0.8817,
+      "step": 433
+    },
+    {
+      "epoch": 0.06641162968630451,
+      "grad_norm": 3.5781509996663234,
+      "learning_rate": 1.9999483820507562e-05,
+      "loss": 0.9097,
+      "step": 434
+    },
+    {
+      "epoch": 0.0665646518745218,
+      "grad_norm": 3.0972342262671324,
+      "learning_rate": 1.9999458334183296e-05,
+      "loss": 0.7467,
+      "step": 435
+    },
+    {
+      "epoch": 0.0667176740627391,
+      "grad_norm": 3.106512537323738,
+      "learning_rate": 1.9999432233753176e-05,
+      "loss": 0.8818,
+      "step": 436
+    },
+    {
+      "epoch": 0.06687069625095639,
+      "grad_norm": 3.2596638513194662,
+      "learning_rate": 1.9999405519218804e-05,
+      "loss": 0.9509,
+      "step": 437
+    },
+    {
+      "epoch": 0.06702371843917368,
+      "grad_norm": 3.477762465985897,
+      "learning_rate": 1.999937819058182e-05,
+      "loss": 0.8831,
+      "step": 438
+    },
+    {
+      "epoch": 0.06717674062739097,
+      "grad_norm": 3.3767003024179054,
+      "learning_rate": 1.9999350247843904e-05,
+      "loss": 0.9931,
+      "step": 439
+    },
+    {
+      "epoch": 0.06732976281560826,
+      "grad_norm": 3.0165691988373786,
+      "learning_rate": 1.999932169100677e-05,
+      "loss": 0.8623,
+      "step": 440
+    },
+    {
+      "epoch": 0.06748278500382555,
+      "grad_norm": 3.0751006001505594,
+      "learning_rate": 1.9999292520072177e-05,
+      "loss": 0.6962,
+      "step": 441
+    },
+    {
+      "epoch": 0.06763580719204285,
+      "grad_norm": 3.6467793080534756,
+      "learning_rate": 1.999926273504191e-05,
+      "loss": 0.9321,
+      "step": 442
+    },
+    {
+      "epoch": 0.06778882938026014,
+      "grad_norm": 3.0687590959557633,
+      "learning_rate": 1.99992323359178e-05,
+      "loss": 0.8116,
+      "step": 443
+    },
+    {
+      "epoch": 0.06794185156847743,
+      "grad_norm": 3.43496703843989,
+      "learning_rate": 1.9999201322701717e-05,
+      "loss": 0.8043,
+      "step": 444
+    },
+    {
+      "epoch": 0.06809487375669472,
+      "grad_norm": 3.544779183403212,
+      "learning_rate": 1.9999169695395566e-05,
+      "loss": 0.8923,
+      "step": 445
+    },
+    {
+      "epoch": 0.06824789594491201,
+      "grad_norm": 3.4034242125677707,
+      "learning_rate": 1.9999137454001282e-05,
+      "loss": 0.8836,
+      "step": 446
+    },
+    {
+      "epoch": 0.0684009181331293,
+      "grad_norm": 2.9508936139483257,
+      "learning_rate": 1.9999104598520854e-05,
+      "loss": 0.8638,
+      "step": 447
+    },
+    {
+      "epoch": 0.0685539403213466,
+      "grad_norm": 3.109658780002175,
+      "learning_rate": 1.9999071128956294e-05,
+      "loss": 0.7898,
+      "step": 448
+    },
+    {
+      "epoch": 0.06870696250956389,
+      "grad_norm": 3.3036335036928475,
+      "learning_rate": 1.9999037045309663e-05,
+      "loss": 0.9348,
+      "step": 449
+    },
+    {
+      "epoch": 0.06885998469778118,
+      "grad_norm": 2.931573589022011,
+      "learning_rate": 1.9999002347583048e-05,
+      "loss": 0.8075,
+      "step": 450
+    },
+    {
+      "epoch": 0.06901300688599847,
+      "grad_norm": 3.1926862679514265,
+      "learning_rate": 1.999896703577858e-05,
+      "loss": 0.8226,
+      "step": 451
+    },
+    {
+      "epoch": 0.06916602907421576,
+      "grad_norm": 3.455774568574597,
+      "learning_rate": 1.9998931109898433e-05,
+      "loss": 0.872,
+      "step": 452
+    },
+    {
+      "epoch": 0.06931905126243305,
+      "grad_norm": 3.1693621632015807,
+      "learning_rate": 1.999889456994481e-05,
+      "loss": 0.7752,
+      "step": 453
+    },
+    {
+      "epoch": 0.06947207345065035,
+      "grad_norm": 2.9868239567077977,
+      "learning_rate": 1.9998857415919955e-05,
+      "loss": 0.9065,
+      "step": 454
+    },
+    {
+      "epoch": 0.06962509563886764,
+      "grad_norm": 3.0397111537097006,
+      "learning_rate": 1.999881964782615e-05,
+      "loss": 0.7929,
+      "step": 455
+    },
+    {
+      "epoch": 0.06977811782708493,
+      "grad_norm": 3.549361128360661,
+      "learning_rate": 1.9998781265665715e-05,
+      "loss": 0.8541,
+      "step": 456
+    },
+    {
+      "epoch": 0.06993114001530222,
+      "grad_norm": 3.044784684574379,
+      "learning_rate": 1.9998742269441008e-05,
+      "loss": 0.8231,
+      "step": 457
+    },
+    {
+      "epoch": 0.07008416220351951,
+      "grad_norm": 3.1048626907375954,
+      "learning_rate": 1.9998702659154423e-05,
+      "loss": 0.7968,
+      "step": 458
+    },
+    {
+      "epoch": 0.0702371843917368,
+      "grad_norm": 3.2945028895738107,
+      "learning_rate": 1.999866243480839e-05,
+      "loss": 0.8379,
+      "step": 459
+    },
+    {
+      "epoch": 0.0703902065799541,
+      "grad_norm": 3.178438471947671,
+      "learning_rate": 1.9998621596405384e-05,
+      "loss": 0.9277,
+      "step": 460
+    },
+    {
+      "epoch": 0.07054322876817139,
+      "grad_norm": 3.42484137351198,
+      "learning_rate": 1.9998580143947914e-05,
+      "loss": 0.8155,
+      "step": 461
+    },
+    {
+      "epoch": 0.07069625095638868,
+      "grad_norm": 3.0491149098944135,
+      "learning_rate": 1.999853807743852e-05,
+      "loss": 0.7763,
+      "step": 462
+    },
+    {
+      "epoch": 0.07084927314460597,
+      "grad_norm": 3.376002345401871,
+      "learning_rate": 1.9998495396879786e-05,
+      "loss": 0.9263,
+      "step": 463
+    },
+    {
+      "epoch": 0.07100229533282326,
+      "grad_norm": 3.2633106035230766,
+      "learning_rate": 1.9998452102274336e-05,
+      "loss": 0.9468,
+      "step": 464
+    },
+    {
+      "epoch": 0.07115531752104055,
+      "grad_norm": 3.1933206048414293,
+      "learning_rate": 1.9998408193624834e-05,
+      "loss": 0.8901,
+      "step": 465
+    },
+    {
+      "epoch": 0.07130833970925785,
+      "grad_norm": 3.4082992385767206,
+      "learning_rate": 1.999836367093396e-05,
+      "loss": 0.9357,
+      "step": 466
+    },
+    {
+      "epoch": 0.07146136189747514,
+      "grad_norm": 3.4102970874735745,
+      "learning_rate": 1.999831853420447e-05,
+      "loss": 0.9104,
+      "step": 467
+    },
+    {
+      "epoch": 0.07161438408569243,
+      "grad_norm": 3.014808972109428,
+      "learning_rate": 1.9998272783439118e-05,
+      "loss": 0.8472,
+      "step": 468
+    },
+    {
+      "epoch": 0.07176740627390972,
+      "grad_norm": 3.0016946749987983,
+      "learning_rate": 1.9998226418640724e-05,
+      "loss": 0.8353,
+      "step": 469
+    },
+    {
+      "epoch": 0.07192042846212701,
+      "grad_norm": 3.4092515518117836,
+      "learning_rate": 1.999817943981213e-05,
+      "loss": 0.7995,
+      "step": 470
+    },
+    {
+      "epoch": 0.0720734506503443,
+      "grad_norm": 2.9282228456092887,
+      "learning_rate": 1.9998131846956225e-05,
+      "loss": 0.899,
+      "step": 471
+    },
+    {
+      "epoch": 0.0722264728385616,
+      "grad_norm": 3.4435930622433037,
+      "learning_rate": 1.9998083640075935e-05,
+      "loss": 0.9694,
+      "step": 472
+    },
+    {
+      "epoch": 0.07237949502677889,
+      "grad_norm": 3.284470938610737,
+      "learning_rate": 1.9998034819174207e-05,
+      "loss": 0.878,
+      "step": 473
+    },
+    {
+      "epoch": 0.07253251721499618,
+      "grad_norm": 3.2453744318117943,
+      "learning_rate": 1.9997985384254057e-05,
+      "loss": 0.7952,
+      "step": 474
+    },
+    {
+      "epoch": 0.07268553940321347,
+      "grad_norm": 3.577377408891758,
+      "learning_rate": 1.9997935335318507e-05,
+      "loss": 0.8137,
+      "step": 475
+    },
+    {
+      "epoch": 0.07283856159143076,
+      "grad_norm": 3.2705975222934556,
+      "learning_rate": 1.9997884672370637e-05,
+      "loss": 0.918,
+      "step": 476
+    },
+    {
+      "epoch": 0.07299158377964805,
+      "grad_norm": 3.4645999740507385,
+      "learning_rate": 1.9997833395413554e-05,
+      "loss": 0.7966,
+      "step": 477
+    },
+    {
+      "epoch": 0.07314460596786534,
+      "grad_norm": 3.241023758964187,
+      "learning_rate": 1.9997781504450416e-05,
+      "loss": 0.7054,
+      "step": 478
+    },
+    {
+      "epoch": 0.07329762815608264,
+      "grad_norm": 3.1306221791660174,
+      "learning_rate": 1.99977289994844e-05,
+      "loss": 0.8397,
+      "step": 479
+    },
+    {
+      "epoch": 0.07345065034429993,
+      "grad_norm": 3.5542684374846676,
+      "learning_rate": 1.999767588051874e-05,
+      "loss": 0.9649,
+      "step": 480
+    },
+    {
+      "epoch": 0.07360367253251722,
+      "grad_norm": 3.1099423042693375,
+      "learning_rate": 1.9997622147556687e-05,
+      "loss": 0.8699,
+      "step": 481
+    },
+    {
+      "epoch": 0.07375669472073451,
+      "grad_norm": 3.092152194487772,
+      "learning_rate": 1.9997567800601548e-05,
+      "loss": 0.9186,
+      "step": 482
+    },
+    {
+      "epoch": 0.0739097169089518,
+      "grad_norm": 3.402202777509611,
+      "learning_rate": 1.9997512839656665e-05,
+      "loss": 0.8209,
+      "step": 483
+    },
+    {
+      "epoch": 0.0740627390971691,
+      "grad_norm": 2.9907255336981184,
+      "learning_rate": 1.9997457264725404e-05,
+      "loss": 0.9086,
+      "step": 484
+    },
+    {
+      "epoch": 0.07421576128538639,
+      "grad_norm": 3.131981310207802,
+      "learning_rate": 1.9997401075811182e-05,
+      "loss": 0.8904,
+      "step": 485
+    },
+    {
+      "epoch": 0.07436878347360368,
+      "grad_norm": 3.209634258627602,
+      "learning_rate": 1.999734427291745e-05,
+      "loss": 0.7896,
+      "step": 486
+    },
+    {
+      "epoch": 0.07452180566182097,
+      "grad_norm": 3.340921344451802,
+      "learning_rate": 1.9997286856047696e-05,
+      "loss": 0.8265,
+      "step": 487
+    },
+    {
+      "epoch": 0.07467482785003826,
+      "grad_norm": 3.13565494738022,
+      "learning_rate": 1.999722882520545e-05,
+      "loss": 0.8693,
+      "step": 488
+    },
+    {
+      "epoch": 0.07482785003825555,
+      "grad_norm": 3.02235992945377,
+      "learning_rate": 1.999717018039427e-05,
+      "loss": 0.8724,
+      "step": 489
+    },
+    {
+      "epoch": 0.07498087222647284,
+      "grad_norm": 3.4023434406169795,
+      "learning_rate": 1.999711092161776e-05,
+      "loss": 0.9148,
+      "step": 490
+    },
+    {
+      "epoch": 0.07513389441469014,
+      "grad_norm": 3.343264311302944,
+      "learning_rate": 1.999705104887956e-05,
+      "loss": 0.8204,
+      "step": 491
+    },
+    {
+      "epoch": 0.07528691660290743,
+      "grad_norm": 3.01072439243528,
+      "learning_rate": 1.9996990562183344e-05,
+      "loss": 0.8356,
+      "step": 492
+    },
+    {
+      "epoch": 0.07543993879112472,
+      "grad_norm": 3.0677819421014543,
+      "learning_rate": 1.999692946153283e-05,
+      "loss": 0.9457,
+      "step": 493
+    },
+    {
+      "epoch": 0.07559296097934201,
+      "grad_norm": 3.049481965152659,
+      "learning_rate": 1.999686774693177e-05,
+      "loss": 0.8406,
+      "step": 494
+    },
+    {
+      "epoch": 0.0757459831675593,
+      "grad_norm": 3.1273680772157997,
+      "learning_rate": 1.9996805418383957e-05,
+      "loss": 0.8543,
+      "step": 495
+    },
+    {
+      "epoch": 0.0758990053557766,
+      "grad_norm": 3.2658135510698543,
+      "learning_rate": 1.999674247589321e-05,
+      "loss": 0.801,
+      "step": 496
+    },
+    {
+      "epoch": 0.07605202754399389,
+      "grad_norm": 3.2022841129021966,
+      "learning_rate": 1.9996678919463403e-05,
+      "loss": 0.9407,
+      "step": 497
+    },
+    {
+      "epoch": 0.07620504973221118,
+      "grad_norm": 3.2241072818356846,
+      "learning_rate": 1.9996614749098438e-05,
+      "loss": 0.9081,
+      "step": 498
+    },
+    {
+      "epoch": 0.07635807192042846,
+      "grad_norm": 3.141672970165098,
+      "learning_rate": 1.999654996480225e-05,
+      "loss": 0.8296,
+      "step": 499
+    },
+    {
+      "epoch": 0.07651109410864575,
+      "grad_norm": 2.8655818640002884,
+      "learning_rate": 1.9996484566578824e-05,
+      "loss": 0.7494,
+      "step": 500
+    },
+    {
+      "epoch": 0.07666411629686304,
+      "grad_norm": 3.005360495550959,
+      "learning_rate": 1.9996418554432175e-05,
+      "loss": 0.7814,
+      "step": 501
+    },
+    {
+      "epoch": 0.07681713848508033,
+      "grad_norm": 3.3007083879137364,
+      "learning_rate": 1.999635192836636e-05,
+      "loss": 0.8228,
+      "step": 502
+    },
+    {
+      "epoch": 0.07697016067329762,
+      "grad_norm": 3.239789821398647,
+      "learning_rate": 1.9996284688385458e-05,
+      "loss": 0.8911,
+      "step": 503
+    },
+    {
+      "epoch": 0.07712318286151491,
+      "grad_norm": 3.16801784653872,
+      "learning_rate": 1.9996216834493614e-05,
+      "loss": 0.8827,
+      "step": 504
+    },
+    {
+      "epoch": 0.0772762050497322,
+      "grad_norm": 3.2341267297113223,
+      "learning_rate": 1.9996148366694987e-05,
+      "loss": 0.8773,
+      "step": 505
+    },
+    {
+      "epoch": 0.0774292272379495,
+      "grad_norm": 3.0993619034994526,
+      "learning_rate": 1.9996079284993785e-05,
+      "loss": 0.898,
+      "step": 506
+    },
+    {
+      "epoch": 0.07758224942616679,
+      "grad_norm": 2.9243224608176637,
+      "learning_rate": 1.999600958939425e-05,
+      "loss": 0.8461,
+      "step": 507
+    },
+    {
+      "epoch": 0.07773527161438408,
+      "grad_norm": 3.307290573644802,
+      "learning_rate": 1.9995939279900658e-05,
+      "loss": 0.91,
+      "step": 508
+    },
+    {
+      "epoch": 0.07788829380260137,
+      "grad_norm": 2.9365920881459986,
+      "learning_rate": 1.999586835651733e-05,
+      "loss": 0.9606,
+      "step": 509
+    },
+    {
+      "epoch": 0.07804131599081866,
+      "grad_norm": 2.724422019837075,
+      "learning_rate": 1.999579681924863e-05,
+      "loss": 0.8498,
+      "step": 510
+    },
+    {
+      "epoch": 0.07819433817903595,
+      "grad_norm": 2.9876304383699326,
+      "learning_rate": 1.9995724668098936e-05,
+      "loss": 0.7418,
+      "step": 511
+    },
+    {
+      "epoch": 0.07834736036725325,
+      "grad_norm": 2.864681589216531,
+      "learning_rate": 1.999565190307269e-05,
+      "loss": 0.8075,
+      "step": 512
+    },
+    {
+      "epoch": 0.07850038255547054,
+      "grad_norm": 2.8644979232121517,
+      "learning_rate": 1.9995578524174354e-05,
+      "loss": 0.7353,
+      "step": 513
+    },
+    {
+      "epoch": 0.07865340474368783,
+      "grad_norm": 3.2377935340367707,
+      "learning_rate": 1.999550453140844e-05,
+      "loss": 0.8998,
+      "step": 514
+    },
+    {
+      "epoch": 0.07880642693190512,
+      "grad_norm": 3.0349644799969457,
+      "learning_rate": 1.999542992477949e-05,
+      "loss": 0.8999,
+      "step": 515
+    },
+    {
+      "epoch": 0.07895944912012241,
+      "grad_norm": 3.187108910079354,
+      "learning_rate": 1.9995354704292086e-05,
+      "loss": 0.9367,
+      "step": 516
+    },
+    {
+      "epoch": 0.0791124713083397,
+      "grad_norm": 3.6556676597386955,
+      "learning_rate": 1.9995278869950848e-05,
+      "loss": 0.8974,
+      "step": 517
+    },
+    {
+      "epoch": 0.079265493496557,
+      "grad_norm": 3.149907651192199,
+      "learning_rate": 1.9995202421760432e-05,
+      "loss": 0.8972,
+      "step": 518
+    },
+    {
+      "epoch": 0.07941851568477429,
+      "grad_norm": 3.1073516404536368,
+      "learning_rate": 1.9995125359725534e-05,
+      "loss": 0.841,
+      "step": 519
+    },
+    {
+      "epoch": 0.07957153787299158,
+      "grad_norm": 3.3022754605273668,
+      "learning_rate": 1.9995047683850887e-05,
+      "loss": 0.9602,
+      "step": 520
+    },
+    {
+      "epoch": 0.07972456006120887,
+      "grad_norm": 3.0277051105867216,
+      "learning_rate": 1.999496939414126e-05,
+      "loss": 0.8321,
+      "step": 521
+    },
+    {
+      "epoch": 0.07987758224942616,
+      "grad_norm": 3.2127384679862416,
+      "learning_rate": 1.9994890490601463e-05,
+      "loss": 0.7626,
+      "step": 522
+    },
+    {
+      "epoch": 0.08003060443764345,
+      "grad_norm": 3.1231645870718814,
+      "learning_rate": 1.999481097323634e-05,
+      "loss": 0.8598,
+      "step": 523
+    },
+    {
+      "epoch": 0.08018362662586075,
+      "grad_norm": 2.919715281820122,
+      "learning_rate": 1.9994730842050776e-05,
+      "loss": 0.8244,
+      "step": 524
+    },
+    {
+      "epoch": 0.08033664881407804,
+      "grad_norm": 2.822749796794928,
+      "learning_rate": 1.999465009704969e-05,
+      "loss": 0.8562,
+      "step": 525
+    },
+    {
+      "epoch": 0.08048967100229533,
+      "grad_norm": 2.91444240792536,
+      "learning_rate": 1.9994568738238046e-05,
+      "loss": 1.0529,
+      "step": 526
+    },
+    {
+      "epoch": 0.08064269319051262,
+      "grad_norm": 3.210762650851107,
+      "learning_rate": 1.9994486765620834e-05,
+      "loss": 0.9525,
+      "step": 527
+    },
+    {
+      "epoch": 0.08079571537872991,
+      "grad_norm": 3.1679560833629528,
+      "learning_rate": 1.9994404179203092e-05,
+      "loss": 0.8337,
+      "step": 528
+    },
+    {
+      "epoch": 0.0809487375669472,
+      "grad_norm": 2.833453068833028,
+      "learning_rate": 1.9994320978989887e-05,
+      "loss": 0.9588,
+      "step": 529
+    },
+    {
+      "epoch": 0.0811017597551645,
+      "grad_norm": 3.2638311444981998,
+      "learning_rate": 1.9994237164986338e-05,
+      "loss": 0.9316,
+      "step": 530
+    },
+    {
+      "epoch": 0.08125478194338179,
+      "grad_norm": 3.4811522175868608,
+      "learning_rate": 1.9994152737197586e-05,
+      "loss": 0.9002,
+      "step": 531
+    },
+    {
+      "epoch": 0.08140780413159908,
+      "grad_norm": 3.0731278665818036,
+      "learning_rate": 1.9994067695628815e-05,
+      "loss": 0.8697,
+      "step": 532
+    },
+    {
+      "epoch": 0.08156082631981637,
+      "grad_norm": 2.8689492083403647,
+      "learning_rate": 1.999398204028525e-05,
+      "loss": 0.8766,
+      "step": 533
+    },
+    {
+      "epoch": 0.08171384850803366,
+      "grad_norm": 2.972650733628086,
+      "learning_rate": 1.9993895771172153e-05,
+      "loss": 0.7181,
+      "step": 534
+    },
+    {
+      "epoch": 0.08186687069625095,
+      "grad_norm": 2.880097853744865,
+      "learning_rate": 1.9993808888294816e-05,
+      "loss": 0.8629,
+      "step": 535
+    },
+    {
+      "epoch": 0.08201989288446825,
+      "grad_norm": 2.8699706044768116,
+      "learning_rate": 1.9993721391658584e-05,
+      "loss": 0.8897,
+      "step": 536
+    },
+    {
+      "epoch": 0.08217291507268554,
+      "grad_norm": 3.2307318749796448,
+      "learning_rate": 1.9993633281268825e-05,
+      "loss": 0.79,
+      "step": 537
+    },
+    {
+      "epoch": 0.08232593726090283,
+      "grad_norm": 3.4903511369450237,
+      "learning_rate": 1.999354455713095e-05,
+      "loss": 0.968,
+      "step": 538
+    },
+    {
+      "epoch": 0.08247895944912012,
+      "grad_norm": 2.487957846673477,
+      "learning_rate": 1.9993455219250407e-05,
+      "loss": 0.8133,
+      "step": 539
+    },
+    {
+      "epoch": 0.08263198163733741,
+      "grad_norm": 3.022484553326492,
+      "learning_rate": 1.999336526763269e-05,
+      "loss": 0.8825,
+      "step": 540
+    },
+    {
+      "epoch": 0.0827850038255547,
+      "grad_norm": 2.8453697669895015,
+      "learning_rate": 1.9993274702283313e-05,
+      "loss": 0.8821,
+      "step": 541
+    },
+    {
+      "epoch": 0.082938026013772,
+      "grad_norm": 3.3972904425220944,
+      "learning_rate": 1.999318352320784e-05,
+      "loss": 0.9181,
+      "step": 542
+    },
+    {
+      "epoch": 0.08309104820198929,
+      "grad_norm": 3.389004916540709,
+      "learning_rate": 1.999309173041188e-05,
+      "loss": 0.9292,
+      "step": 543
+    },
+    {
+      "epoch": 0.08324407039020658,
+      "grad_norm": 3.0472330893944806,
+      "learning_rate": 1.999299932390106e-05,
+      "loss": 0.7503,
+      "step": 544
+    },
+    {
+      "epoch": 0.08339709257842387,
+      "grad_norm": 3.062998869061395,
+      "learning_rate": 1.9992906303681057e-05,
+      "loss": 0.9149,
+      "step": 545
+    },
+    {
+      "epoch": 0.08355011476664116,
+      "grad_norm": 3.662910791698355,
+      "learning_rate": 1.999281266975759e-05,
+      "loss": 0.8746,
+      "step": 546
+    },
+    {
+      "epoch": 0.08370313695485845,
+      "grad_norm": 2.9445135400870637,
+      "learning_rate": 1.99927184221364e-05,
+      "loss": 0.8416,
+      "step": 547
+    },
+    {
+      "epoch": 0.08385615914307575,
+      "grad_norm": 2.9881136587386026,
+      "learning_rate": 1.9992623560823284e-05,
+      "loss": 0.86,
+      "step": 548
+    },
+    {
+      "epoch": 0.08400918133129304,
+      "grad_norm": 2.8957772667517085,
+      "learning_rate": 1.999252808582406e-05,
+      "loss": 0.8825,
+      "step": 549
+    },
+    {
+      "epoch": 0.08416220351951033,
+      "grad_norm": 2.7973789830398883,
+      "learning_rate": 1.99924319971446e-05,
+      "loss": 0.8052,
+      "step": 550
+    },
+    {
+      "epoch": 0.08431522570772762,
+      "grad_norm": 2.778130072053958,
+      "learning_rate": 1.9992335294790797e-05,
+      "loss": 0.7703,
+      "step": 551
+    },
+    {
+      "epoch": 0.08446824789594491,
+      "grad_norm": 3.1327192932994405,
+      "learning_rate": 1.9992237978768593e-05,
+      "loss": 0.9049,
+      "step": 552
+    },
+    {
+      "epoch": 0.0846212700841622,
+      "grad_norm": 3.1565650717986977,
+      "learning_rate": 1.9992140049083968e-05,
+      "loss": 0.8519,
+      "step": 553
+    },
+    {
+      "epoch": 0.0847742922723795,
+      "grad_norm": 3.1756761607270194,
+      "learning_rate": 1.999204150574293e-05,
+      "loss": 1.0047,
+      "step": 554
+    },
+    {
+      "epoch": 0.08492731446059679,
+      "grad_norm": 2.6052032947874864,
+      "learning_rate": 1.9991942348751534e-05,
+      "loss": 0.8446,
+      "step": 555
+    },
+    {
+      "epoch": 0.08508033664881408,
+      "grad_norm": 3.3747384938061846,
+      "learning_rate": 1.9991842578115872e-05,
+      "loss": 0.9783,
+      "step": 556
+    },
+    {
+      "epoch": 0.08523335883703137,
+      "grad_norm": 3.1466121416772532,
+      "learning_rate": 1.999174219384207e-05,
+      "loss": 0.9396,
+      "step": 557
+    },
+    {
+      "epoch": 0.08538638102524866,
+      "grad_norm": 3.203806072282312,
+      "learning_rate": 1.999164119593629e-05,
+      "loss": 0.8774,
+      "step": 558
+    },
+    {
+      "epoch": 0.08553940321346595,
+      "grad_norm": 2.8295832347751544,
+      "learning_rate": 1.9991539584404734e-05,
+      "loss": 0.7787,
+      "step": 559
+    },
+    {
+      "epoch": 0.08569242540168324,
+      "grad_norm": 2.9986481486485315,
+      "learning_rate": 1.9991437359253647e-05,
+      "loss": 0.8303,
+      "step": 560
+    },
+    {
+      "epoch": 0.08584544758990054,
+      "grad_norm": 3.1163102632937525,
+      "learning_rate": 1.9991334520489304e-05,
+      "loss": 0.8202,
+      "step": 561
+    },
+    {
+      "epoch": 0.08599846977811783,
+      "grad_norm": 3.3672338761393616,
+      "learning_rate": 1.9991231068118024e-05,
+      "loss": 0.8453,
+      "step": 562
+    },
+    {
+      "epoch": 0.08615149196633512,
+      "grad_norm": 2.9326058066627123,
+      "learning_rate": 1.9991127002146157e-05,
+      "loss": 0.816,
+      "step": 563
+    },
+    {
+      "epoch": 0.08630451415455241,
+      "grad_norm": 2.9286950810713277,
+      "learning_rate": 1.9991022322580096e-05,
+      "loss": 0.893,
+      "step": 564
+    },
+    {
+      "epoch": 0.0864575363427697,
+      "grad_norm": 3.390923338954129,
+      "learning_rate": 1.999091702942627e-05,
+      "loss": 0.942,
+      "step": 565
+    },
+    {
+      "epoch": 0.086610558530987,
+      "grad_norm": 3.100849028861991,
+      "learning_rate": 1.9990811122691142e-05,
+      "loss": 0.8572,
+      "step": 566
+    },
+    {
+      "epoch": 0.08676358071920429,
+      "grad_norm": 3.024939074035441,
+      "learning_rate": 1.9990704602381222e-05,
+      "loss": 0.9249,
+      "step": 567
+    },
+    {
+      "epoch": 0.08691660290742158,
+      "grad_norm": 3.086292608040693,
+      "learning_rate": 1.9990597468503044e-05,
+      "loss": 0.7923,
+      "step": 568
+    },
+    {
+      "epoch": 0.08706962509563887,
+      "grad_norm": 2.8094297793401397,
+      "learning_rate": 1.9990489721063194e-05,
+      "loss": 0.7916,
+      "step": 569
+    },
+    {
+      "epoch": 0.08722264728385616,
+      "grad_norm": 2.906165406278737,
+      "learning_rate": 1.999038136006829e-05,
+      "loss": 0.8567,
+      "step": 570
+    },
+    {
+      "epoch": 0.08737566947207345,
+      "grad_norm": 3.027894811082444,
+      "learning_rate": 1.9990272385524983e-05,
+      "loss": 0.8776,
+      "step": 571
+    },
+    {
+      "epoch": 0.08752869166029074,
+      "grad_norm": 2.951885113296094,
+      "learning_rate": 1.9990162797439964e-05,
+      "loss": 0.7175,
+      "step": 572
+    },
+    {
+      "epoch": 0.08768171384850804,
+      "grad_norm": 2.8448885126107464,
+      "learning_rate": 1.9990052595819966e-05,
+      "loss": 0.7836,
+      "step": 573
+    },
+    {
+      "epoch": 0.08783473603672533,
+      "grad_norm": 3.0008723156763635,
+      "learning_rate": 1.998994178067176e-05,
+      "loss": 0.9061,
+      "step": 574
+    },
+    {
+      "epoch": 0.08798775822494262,
+      "grad_norm": 2.766808323475977,
+      "learning_rate": 1.9989830352002144e-05,
+      "loss": 0.7398,
+      "step": 575
+    },
+    {
+      "epoch": 0.08814078041315991,
+      "grad_norm": 3.089772730908918,
+      "learning_rate": 1.998971830981797e-05,
+      "loss": 0.9764,
+      "step": 576
+    },
+    {
+      "epoch": 0.0882938026013772,
+      "grad_norm": 3.2210205902228903,
+      "learning_rate": 1.9989605654126113e-05,
+      "loss": 0.88,
+      "step": 577
+    },
+    {
+      "epoch": 0.0884468247895945,
+      "grad_norm": 3.3361093624428526,
+      "learning_rate": 1.9989492384933493e-05,
+      "loss": 0.9849,
+      "step": 578
+    },
+    {
+      "epoch": 0.08859984697781179,
+      "grad_norm": 2.9746314886323773,
+      "learning_rate": 1.9989378502247067e-05,
+      "loss": 0.8476,
+      "step": 579
+    },
+    {
+      "epoch": 0.08875286916602908,
+      "grad_norm": 2.8992176182889104,
+      "learning_rate": 1.9989264006073826e-05,
+      "loss": 0.7202,
+      "step": 580
+    },
+    {
+      "epoch": 0.08890589135424637,
+      "grad_norm": 2.7055032415422726,
+      "learning_rate": 1.998914889642081e-05,
+      "loss": 0.7931,
+      "step": 581
+    },
+    {
+      "epoch": 0.08905891354246366,
+      "grad_norm": 2.97229656950559,
+      "learning_rate": 1.9989033173295075e-05,
+      "loss": 0.7473,
+      "step": 582
+    },
+    {
+      "epoch": 0.08921193573068095,
+      "grad_norm": 3.013939775038768,
+      "learning_rate": 1.9988916836703738e-05,
+      "loss": 0.8688,
+      "step": 583
+    },
+    {
+      "epoch": 0.08936495791889824,
+      "grad_norm": 3.2961620765077106,
+      "learning_rate": 1.998879988665394e-05,
+      "loss": 0.9014,
+      "step": 584
+    },
+    {
+      "epoch": 0.08951798010711554,
+      "grad_norm": 3.136110348623102,
+      "learning_rate": 1.998868232315287e-05,
+      "loss": 0.8517,
+      "step": 585
+    },
+    {
+      "epoch": 0.08967100229533283,
+      "grad_norm": 2.873765409966888,
+      "learning_rate": 1.998856414620774e-05,
+      "loss": 0.8392,
+      "step": 586
+    },
+    {
+      "epoch": 0.08982402448355012,
+      "grad_norm": 3.0620369003779997,
+      "learning_rate": 1.9988445355825808e-05,
+      "loss": 0.9311,
+      "step": 587
+    },
+    {
+      "epoch": 0.08997704667176741,
+      "grad_norm": 2.9488999826259583,
+      "learning_rate": 1.9988325952014375e-05,
+      "loss": 0.7894,
+      "step": 588
+    },
+    {
+      "epoch": 0.0901300688599847,
+      "grad_norm": 3.364620998306335,
+      "learning_rate": 1.998820593478077e-05,
+      "loss": 0.9131,
+      "step": 589
+    },
+    {
+      "epoch": 0.090283091048202,
+      "grad_norm": 3.486214934509759,
+      "learning_rate": 1.9988085304132362e-05,
+      "loss": 0.9463,
+      "step": 590
+    },
+    {
+      "epoch": 0.09043611323641929,
+      "grad_norm": 3.103071840216333,
+      "learning_rate": 1.9987964060076565e-05,
+      "loss": 0.9319,
+      "step": 591
+    },
+    {
+      "epoch": 0.09058913542463658,
+      "grad_norm": 2.7855301279351736,
+      "learning_rate": 1.998784220262082e-05,
+      "loss": 0.8917,
+      "step": 592
+    },
+    {
+      "epoch": 0.09074215761285387,
+      "grad_norm": 2.628571920591738,
+      "learning_rate": 1.998771973177261e-05,
+      "loss": 0.8585,
+      "step": 593
+    },
+    {
+      "epoch": 0.09089517980107116,
+      "grad_norm": 3.0308283014483135,
+      "learning_rate": 1.9987596647539464e-05,
+      "loss": 0.8611,
+      "step": 594
+    },
+    {
+      "epoch": 0.09104820198928845,
+      "grad_norm": 3.395921773906306,
+      "learning_rate": 1.9987472949928936e-05,
+      "loss": 0.8705,
+      "step": 595
+    },
+    {
+      "epoch": 0.09120122417750574,
+      "grad_norm": 3.185931373181077,
+      "learning_rate": 1.9987348638948623e-05,
+      "loss": 0.8794,
+      "step": 596
+    },
+    {
+      "epoch": 0.09135424636572304,
+      "grad_norm": 3.2153230873365857,
+      "learning_rate": 1.9987223714606156e-05,
+      "loss": 0.9029,
+      "step": 597
+    },
+    {
+      "epoch": 0.09150726855394033,
+      "grad_norm": 2.615904068828914,
+      "learning_rate": 1.9987098176909213e-05,
+      "loss": 0.8384,
+      "step": 598
+    },
+    {
+      "epoch": 0.09166029074215762,
+      "grad_norm": 3.1448963558441023,
+      "learning_rate": 1.99869720258655e-05,
+      "loss": 0.8621,
+      "step": 599
+    },
+    {
+      "epoch": 0.09181331293037491,
+      "grad_norm": 2.848509682376616,
+      "learning_rate": 1.9986845261482767e-05,
+      "loss": 0.8979,
+      "step": 600
+    },
+    {
+      "epoch": 0.0919663351185922,
+      "grad_norm": 2.8631042534970117,
+      "learning_rate": 1.9986717883768796e-05,
+      "loss": 0.7573,
+      "step": 601
+    },
+    {
+      "epoch": 0.0921193573068095,
+      "grad_norm": 2.877288020891183,
+      "learning_rate": 1.998658989273141e-05,
+      "loss": 0.8908,
+      "step": 602
+    },
+    {
+      "epoch": 0.09227237949502679,
+      "grad_norm": 2.9466061917507855,
+      "learning_rate": 1.9986461288378475e-05,
+      "loss": 0.8146,
+      "step": 603
+    },
+    {
+      "epoch": 0.09242540168324408,
+      "grad_norm": 3.2326794041122344,
+      "learning_rate": 1.9986332070717882e-05,
+      "loss": 0.866,
+      "step": 604
+    },
+    {
+      "epoch": 0.09257842387146137,
+      "grad_norm": 3.2136492245233987,
+      "learning_rate": 1.9986202239757572e-05,
+      "loss": 0.9044,
+      "step": 605
+    },
+    {
+      "epoch": 0.09273144605967866,
+      "grad_norm": 2.9185939116175796,
+      "learning_rate": 1.9986071795505516e-05,
+      "loss": 0.799,
+      "step": 606
+    },
+    {
+      "epoch": 0.09288446824789595,
+      "grad_norm": 2.5518184921354545,
+      "learning_rate": 1.9985940737969724e-05,
+      "loss": 0.7654,
+      "step": 607
+    },
+    {
+      "epoch": 0.09303749043611324,
+      "grad_norm": 3.4896794353885157,
+      "learning_rate": 1.9985809067158246e-05,
+      "loss": 0.8924,
+      "step": 608
+    },
+    {
+      "epoch": 0.09319051262433053,
+      "grad_norm": 2.9217667565577408,
+      "learning_rate": 1.998567678307917e-05,
+      "loss": 0.7938,
+      "step": 609
+    },
+    {
+      "epoch": 0.09334353481254781,
+      "grad_norm": 2.9217531851694076,
+      "learning_rate": 1.9985543885740616e-05,
+      "loss": 0.8824,
+      "step": 610
+    },
+    {
+      "epoch": 0.0934965570007651,
+      "grad_norm": 3.466534744413211,
+      "learning_rate": 1.998541037515075e-05,
+      "loss": 0.8839,
+      "step": 611
+    },
+    {
+      "epoch": 0.0936495791889824,
+      "grad_norm": 2.8647068519856473,
+      "learning_rate": 1.998527625131777e-05,
+      "loss": 0.7641,
+      "step": 612
+    },
+    {
+      "epoch": 0.09380260137719969,
+      "grad_norm": 2.990874053782021,
+      "learning_rate": 1.9985141514249913e-05,
+      "loss": 0.8285,
+      "step": 613
+    },
+    {
+      "epoch": 0.09395562356541698,
+      "grad_norm": 3.213841555154716,
+      "learning_rate": 1.9985006163955454e-05,
+      "loss": 0.8348,
+      "step": 614
+    },
+    {
+      "epoch": 0.09410864575363427,
+      "grad_norm": 2.897833932939109,
+      "learning_rate": 1.9984870200442704e-05,
+      "loss": 0.8974,
+      "step": 615
+    },
+    {
+      "epoch": 0.09426166794185156,
+      "grad_norm": 3.255093140463334,
+      "learning_rate": 1.998473362372001e-05,
+      "loss": 0.77,
+      "step": 616
+    },
+    {
+      "epoch": 0.09441469013006885,
+      "grad_norm": 3.5541022603639654,
+      "learning_rate": 1.9984596433795768e-05,
+      "loss": 0.9675,
+      "step": 617
+    },
+    {
+      "epoch": 0.09456771231828615,
+      "grad_norm": 3.156911376432198,
+      "learning_rate": 1.99844586306784e-05,
+      "loss": 0.7643,
+      "step": 618
+    },
+    {
+      "epoch": 0.09472073450650344,
+      "grad_norm": 3.0327127020510014,
+      "learning_rate": 1.9984320214376367e-05,
+      "loss": 0.9454,
+      "step": 619
+    },
+    {
+      "epoch": 0.09487375669472073,
+      "grad_norm": 2.7145164111445763,
+      "learning_rate": 1.998418118489817e-05,
+      "loss": 0.78,
+      "step": 620
+    },
+    {
+      "epoch": 0.09502677888293802,
+      "grad_norm": 3.042572988904212,
+      "learning_rate": 1.998404154225235e-05,
+      "loss": 0.8324,
+      "step": 621
+    },
+    {
+      "epoch": 0.09517980107115531,
+      "grad_norm": 2.590002583866103,
+      "learning_rate": 1.9983901286447477e-05,
+      "loss": 0.7565,
+      "step": 622
+    },
+    {
+      "epoch": 0.0953328232593726,
+      "grad_norm": 3.2287495750309696,
+      "learning_rate": 1.9983760417492173e-05,
+      "loss": 0.7237,
+      "step": 623
+    },
+    {
+      "epoch": 0.0954858454475899,
+      "grad_norm": 2.9615171389313732,
+      "learning_rate": 1.9983618935395084e-05,
+      "loss": 0.8673,
+      "step": 624
+    },
+    {
+      "epoch": 0.09563886763580719,
+      "grad_norm": 2.975886701578284,
+      "learning_rate": 1.9983476840164896e-05,
+      "loss": 0.8856,
+      "step": 625
+    },
+    {
+      "epoch": 0.09579188982402448,
+      "grad_norm": 3.3574876890478054,
+      "learning_rate": 1.9983334131810346e-05,
+      "loss": 0.9716,
+      "step": 626
+    },
+    {
+      "epoch": 0.09594491201224177,
+      "grad_norm": 2.7507257274616648,
+      "learning_rate": 1.9983190810340186e-05,
+      "loss": 0.9263,
+      "step": 627
+    },
+    {
+      "epoch": 0.09609793420045906,
+      "grad_norm": 3.191008180869532,
+      "learning_rate": 1.9983046875763228e-05,
+      "loss": 0.931,
+      "step": 628
+    },
+    {
+      "epoch": 0.09625095638867635,
+      "grad_norm": 2.853979783408991,
+      "learning_rate": 1.998290232808831e-05,
+      "loss": 0.786,
+      "step": 629
+    },
+    {
+      "epoch": 0.09640397857689365,
+      "grad_norm": 3.0623527098455945,
+      "learning_rate": 1.99827571673243e-05,
+      "loss": 0.9594,
+      "step": 630
+    },
+    {
+      "epoch": 0.09655700076511094,
+      "grad_norm": 3.0612621394357387,
+      "learning_rate": 1.9982611393480124e-05,
+      "loss": 0.8467,
+      "step": 631
+    },
+    {
+      "epoch": 0.09671002295332823,
+      "grad_norm": 3.145922754185688,
+      "learning_rate": 1.9982465006564727e-05,
+      "loss": 0.837,
+      "step": 632
+    },
+    {
+      "epoch": 0.09686304514154552,
+      "grad_norm": 3.0267411008622602,
+      "learning_rate": 1.9982318006587107e-05,
+      "loss": 0.9817,
+      "step": 633
+    },
+    {
+      "epoch": 0.09701606732976281,
+      "grad_norm": 2.9130437517324594,
+      "learning_rate": 1.9982170393556282e-05,
+      "loss": 0.7791,
+      "step": 634
+    },
+    {
+      "epoch": 0.0971690895179801,
+      "grad_norm": 2.8442538652688922,
+      "learning_rate": 1.9982022167481324e-05,
+      "loss": 0.9116,
+      "step": 635
+    },
+    {
+      "epoch": 0.0973221117061974,
+      "grad_norm": 2.912158188776152,
+      "learning_rate": 1.9981873328371338e-05,
+      "loss": 0.8294,
+      "step": 636
+    },
+    {
+      "epoch": 0.09747513389441469,
+      "grad_norm": 3.099465213513769,
+      "learning_rate": 1.9981723876235457e-05,
+      "loss": 0.8317,
+      "step": 637
+    },
+    {
+      "epoch": 0.09762815608263198,
+      "grad_norm": 2.9188799056701296,
+      "learning_rate": 1.9981573811082868e-05,
+      "loss": 0.8401,
+      "step": 638
+    },
+    {
+      "epoch": 0.09778117827084927,
+      "grad_norm": 2.819614664399572,
+      "learning_rate": 1.998142313292278e-05,
+      "loss": 0.8368,
+      "step": 639
+    },
+    {
+      "epoch": 0.09793420045906656,
+      "grad_norm": 2.6738959579696577,
+      "learning_rate": 1.9981271841764452e-05,
+      "loss": 0.9057,
+      "step": 640
+    },
+    {
+      "epoch": 0.09808722264728385,
+      "grad_norm": 2.9601321974365735,
+      "learning_rate": 1.9981119937617174e-05,
+      "loss": 0.8107,
+      "step": 641
+    },
+    {
+      "epoch": 0.09824024483550114,
+      "grad_norm": 3.0233216111788472,
+      "learning_rate": 1.9980967420490273e-05,
+      "loss": 0.9779,
+      "step": 642
+    },
+    {
+      "epoch": 0.09839326702371844,
+      "grad_norm": 2.7557734535440166,
+      "learning_rate": 1.9980814290393115e-05,
+      "loss": 0.838,
+      "step": 643
+    },
+    {
+      "epoch": 0.09854628921193573,
+      "grad_norm": 2.704043797297548,
+      "learning_rate": 1.998066054733511e-05,
+      "loss": 0.8751,
+      "step": 644
+    },
+    {
+      "epoch": 0.09869931140015302,
+      "grad_norm": 3.1041651898470657,
+      "learning_rate": 1.9980506191325694e-05,
+      "loss": 0.8846,
+      "step": 645
+    },
+    {
+      "epoch": 0.09885233358837031,
+      "grad_norm": 2.6868333111971476,
+      "learning_rate": 1.9980351222374347e-05,
+      "loss": 0.8375,
+      "step": 646
+    },
+    {
+      "epoch": 0.0990053557765876,
+      "grad_norm": 2.777208715705951,
+      "learning_rate": 1.9980195640490592e-05,
+      "loss": 0.8029,
+      "step": 647
+    },
+    {
+      "epoch": 0.0991583779648049,
+      "grad_norm": 3.4013194486190548,
+      "learning_rate": 1.9980039445683978e-05,
+      "loss": 0.8727,
+      "step": 648
+    },
+    {
+      "epoch": 0.09931140015302219,
+      "grad_norm": 3.0668546286557543,
+      "learning_rate": 1.99798826379641e-05,
+      "loss": 0.9092,
+      "step": 649
+    },
+    {
+      "epoch": 0.09946442234123948,
+      "grad_norm": 3.0453985288017904,
+      "learning_rate": 1.9979725217340587e-05,
+      "loss": 0.8486,
+      "step": 650
+    },
+    {
+      "epoch": 0.09961744452945677,
+      "grad_norm": 2.7942211304315143,
+      "learning_rate": 1.9979567183823108e-05,
+      "loss": 0.8158,
+      "step": 651
+    },
+    {
+      "epoch": 0.09977046671767406,
+      "grad_norm": 3.3292124901062206,
+      "learning_rate": 1.9979408537421367e-05,
+      "loss": 0.9559,
+      "step": 652
+    },
+    {
+      "epoch": 0.09992348890589135,
+      "grad_norm": 3.134527582640561,
+      "learning_rate": 1.997924927814511e-05,
+      "loss": 0.8503,
+      "step": 653
+    },
+    {
+      "epoch": 0.10007651109410864,
+      "grad_norm": 2.9257436635589857,
+      "learning_rate": 1.9979089406004115e-05,
+      "loss": 0.7602,
+      "step": 654
+    },
+    {
+      "epoch": 0.10022953328232594,
+      "grad_norm": 2.9114862208984373,
+      "learning_rate": 1.99789289210082e-05,
+      "loss": 0.8381,
+      "step": 655
+    },
+    {
+      "epoch": 0.10038255547054323,
+      "grad_norm": 3.281387510746206,
+      "learning_rate": 1.9978767823167224e-05,
+      "loss": 0.8544,
+      "step": 656
+    },
+    {
+      "epoch": 0.10053557765876052,
+      "grad_norm": 2.5934944798509774,
+      "learning_rate": 1.9978606112491076e-05,
+      "loss": 0.8238,
+      "step": 657
+    },
+    {
+      "epoch": 0.10068859984697781,
+      "grad_norm": 2.803648838613119,
+      "learning_rate": 1.9978443788989695e-05,
+      "loss": 0.8357,
+      "step": 658
+    },
+    {
+      "epoch": 0.1008416220351951,
+      "grad_norm": 3.2256668016008705,
+      "learning_rate": 1.9978280852673038e-05,
+      "loss": 0.9185,
+      "step": 659
+    },
+    {
+      "epoch": 0.1009946442234124,
+      "grad_norm": 2.9180166328828108,
+      "learning_rate": 1.9978117303551127e-05,
+      "loss": 0.9,
+      "step": 660
+    },
+    {
+      "epoch": 0.10114766641162969,
+      "grad_norm": 2.759551164175027,
+      "learning_rate": 1.997795314163399e-05,
+      "loss": 0.8243,
+      "step": 661
+    },
+    {
+      "epoch": 0.10130068859984698,
+      "grad_norm": 2.8452679100467564,
+      "learning_rate": 1.997778836693172e-05,
+      "loss": 0.8834,
+      "step": 662
+    },
+    {
+      "epoch": 0.10145371078806427,
+      "grad_norm": 2.85925562143535,
+      "learning_rate": 1.9977622979454433e-05,
+      "loss": 0.8022,
+      "step": 663
+    },
+    {
+      "epoch": 0.10160673297628156,
+      "grad_norm": 2.925375455401089,
+      "learning_rate": 1.9977456979212286e-05,
+      "loss": 0.8822,
+      "step": 664
+    },
+    {
+      "epoch": 0.10175975516449885,
+      "grad_norm": 2.725147333625341,
+      "learning_rate": 1.9977290366215473e-05,
+      "loss": 0.8138,
+      "step": 665
+    },
+    {
+      "epoch": 0.10191277735271614,
+      "grad_norm": 3.1953282021420013,
+      "learning_rate": 1.997712314047423e-05,
+      "loss": 0.857,
+      "step": 666
+    },
+    {
+      "epoch": 0.10206579954093344,
+      "grad_norm": 3.0746423357238943,
+      "learning_rate": 1.9976955301998822e-05,
+      "loss": 0.8515,
+      "step": 667
+    },
+    {
+      "epoch": 0.10221882172915073,
+      "grad_norm": 2.900938526635504,
+      "learning_rate": 1.997678685079956e-05,
+      "loss": 0.8601,
+      "step": 668
+    },
+    {
+      "epoch": 0.10237184391736802,
+      "grad_norm": 2.689075444923604,
+      "learning_rate": 1.997661778688679e-05,
+      "loss": 0.8373,
+      "step": 669
+    },
+    {
+      "epoch": 0.10252486610558531,
+      "grad_norm": 2.859953745768656,
+      "learning_rate": 1.9976448110270888e-05,
+      "loss": 0.7555,
+      "step": 670
+    },
+    {
+      "epoch": 0.1026778882938026,
+      "grad_norm": 3.145897880802856,
+      "learning_rate": 1.997627782096228e-05,
+      "loss": 0.9535,
+      "step": 671
+    },
+    {
+      "epoch": 0.1028309104820199,
+      "grad_norm": 2.969611797938469,
+      "learning_rate": 1.9976106918971428e-05,
+      "loss": 0.8239,
+      "step": 672
+    },
+    {
+      "epoch": 0.10298393267023719,
+      "grad_norm": 2.9387550683170014,
+      "learning_rate": 1.9975935404308818e-05,
+      "loss": 0.8714,
+      "step": 673
+    },
+    {
+      "epoch": 0.10313695485845448,
+      "grad_norm": 2.9648398626987444,
+      "learning_rate": 1.9975763276984993e-05,
+      "loss": 0.8172,
+      "step": 674
+    },
+    {
+      "epoch": 0.10328997704667177,
+      "grad_norm": 3.295868102675818,
+      "learning_rate": 1.9975590537010515e-05,
+      "loss": 0.9202,
+      "step": 675
+    },
+    {
+      "epoch": 0.10344299923488906,
+      "grad_norm": 3.22903200685603,
+      "learning_rate": 1.9975417184396005e-05,
+      "loss": 0.9028,
+      "step": 676
+    },
+    {
+      "epoch": 0.10359602142310635,
+      "grad_norm": 2.825966414769792,
+      "learning_rate": 1.9975243219152095e-05,
+      "loss": 0.9263,
+      "step": 677
+    },
+    {
+      "epoch": 0.10374904361132364,
+      "grad_norm": 3.030014157036485,
+      "learning_rate": 1.9975068641289478e-05,
+      "loss": 0.718,
+      "step": 678
+    },
+    {
+      "epoch": 0.10390206579954094,
+      "grad_norm": 3.475541487458047,
+      "learning_rate": 1.9974893450818875e-05,
+      "loss": 0.9376,
+      "step": 679
+    },
+    {
+      "epoch": 0.10405508798775823,
+      "grad_norm": 3.2549605814152356,
+      "learning_rate": 1.997471764775104e-05,
+      "loss": 0.7524,
+      "step": 680
+    },
+    {
+      "epoch": 0.10420811017597552,
+      "grad_norm": 3.126998944683233,
+      "learning_rate": 1.9974541232096774e-05,
+      "loss": 0.9323,
+      "step": 681
+    },
+    {
+      "epoch": 0.10436113236419281,
+      "grad_norm": 3.0013467824393025,
+      "learning_rate": 1.997436420386691e-05,
+      "loss": 0.806,
+      "step": 682
+    },
+    {
+      "epoch": 0.1045141545524101,
+      "grad_norm": 3.0259375619997253,
+      "learning_rate": 1.997418656307232e-05,
+      "loss": 0.8783,
+      "step": 683
+    },
+    {
+      "epoch": 0.1046671767406274,
+      "grad_norm": 3.0450368516418447,
+      "learning_rate": 1.9974008309723917e-05,
+      "loss": 0.9218,
+      "step": 684
+    },
+    {
+      "epoch": 0.10482019892884469,
+      "grad_norm": 2.825921818621975,
+      "learning_rate": 1.997382944383264e-05,
+      "loss": 0.8584,
+      "step": 685
+    },
+    {
+      "epoch": 0.10497322111706198,
+      "grad_norm": 2.912770169006268,
+      "learning_rate": 1.9973649965409483e-05,
+      "loss": 0.8629,
+      "step": 686
+    },
+    {
+      "epoch": 0.10512624330527927,
+      "grad_norm": 2.5954835007580477,
+      "learning_rate": 1.9973469874465464e-05,
+      "loss": 0.8466,
+      "step": 687
+    },
+    {
+      "epoch": 0.10527926549349656,
+      "grad_norm": 2.917363261917072,
+      "learning_rate": 1.997328917101164e-05,
+      "loss": 0.8829,
+      "step": 688
+    },
+    {
+      "epoch": 0.10543228768171385,
+      "grad_norm": 2.812270404547237,
+      "learning_rate": 1.9973107855059116e-05,
+      "loss": 0.8442,
+      "step": 689
+    },
+    {
+      "epoch": 0.10558530986993114,
+      "grad_norm": 3.0483780333514257,
+      "learning_rate": 1.9972925926619023e-05,
+      "loss": 0.8695,
+      "step": 690
+    },
+    {
+      "epoch": 0.10573833205814843,
+      "grad_norm": 2.9664567475153074,
+      "learning_rate": 1.9972743385702535e-05,
+      "loss": 0.9386,
+      "step": 691
+    },
+    {
+      "epoch": 0.10589135424636573,
+      "grad_norm": 2.7137242073331804,
+      "learning_rate": 1.9972560232320863e-05,
+      "loss": 0.7839,
+      "step": 692
+    },
+    {
+      "epoch": 0.10604437643458302,
+      "grad_norm": 3.784136448397792,
+      "learning_rate": 1.9972376466485252e-05,
+      "loss": 0.9644,
+      "step": 693
+    },
+    {
+      "epoch": 0.10619739862280031,
+      "grad_norm": 2.6044337361202183,
+      "learning_rate": 1.997219208820699e-05,
+      "loss": 0.8399,
+      "step": 694
+    },
+    {
+      "epoch": 0.1063504208110176,
+      "grad_norm": 3.106258109412711,
+      "learning_rate": 1.99720070974974e-05,
+      "loss": 0.8544,
+      "step": 695
+    },
+    {
+      "epoch": 0.10650344299923489,
+      "grad_norm": 3.1386122334768376,
+      "learning_rate": 1.9971821494367844e-05,
+      "loss": 0.9707,
+      "step": 696
+    },
+    {
+      "epoch": 0.10665646518745218,
+      "grad_norm": 2.912783268372566,
+      "learning_rate": 1.997163527882972e-05,
+      "loss": 0.8637,
+      "step": 697
+    },
+    {
+      "epoch": 0.10680948737566948,
+      "grad_norm": 2.794401741879464,
+      "learning_rate": 1.9971448450894467e-05,
+      "loss": 0.7666,
+      "step": 698
+    },
+    {
+      "epoch": 0.10696250956388677,
+      "grad_norm": 2.787277032267778,
+      "learning_rate": 1.9971261010573553e-05,
+      "loss": 0.9049,
+      "step": 699
+    },
+    {
+      "epoch": 0.10711553175210406,
+      "grad_norm": 3.1363693151187535,
+      "learning_rate": 1.9971072957878494e-05,
+      "loss": 0.8289,
+      "step": 700
+    },
+    {
+      "epoch": 0.10726855394032135,
+      "grad_norm": 3.016667481448861,
+      "learning_rate": 1.9970884292820837e-05,
+      "loss": 0.7704,
+      "step": 701
+    },
+    {
+      "epoch": 0.10742157612853864,
+      "grad_norm": 3.2062104544910355,
+      "learning_rate": 1.997069501541217e-05,
+      "loss": 0.8958,
+      "step": 702
+    },
+    {
+      "epoch": 0.10757459831675593,
+      "grad_norm": 2.680574521626054,
+      "learning_rate": 1.9970505125664116e-05,
+      "loss": 0.8637,
+      "step": 703
+    },
+    {
+      "epoch": 0.10772762050497323,
+      "grad_norm": 2.674813823925464,
+      "learning_rate": 1.9970314623588335e-05,
+      "loss": 0.7913,
+      "step": 704
+    },
+    {
+      "epoch": 0.10788064269319052,
+      "grad_norm": 2.9406076887807706,
+      "learning_rate": 1.9970123509196533e-05,
+      "loss": 0.9162,
+      "step": 705
+    },
+    {
+      "epoch": 0.10803366488140781,
+      "grad_norm": 3.261303730361806,
+      "learning_rate": 1.996993178250044e-05,
+      "loss": 0.8425,
+      "step": 706
+    },
+    {
+      "epoch": 0.1081866870696251,
+      "grad_norm": 2.8193883845987737,
+      "learning_rate": 1.9969739443511835e-05,
+      "loss": 0.748,
+      "step": 707
+    },
+    {
+      "epoch": 0.10833970925784239,
+      "grad_norm": 2.7504729360532973,
+      "learning_rate": 1.996954649224253e-05,
+      "loss": 0.8859,
+      "step": 708
+    },
+    {
+      "epoch": 0.10849273144605968,
+      "grad_norm": 2.7955831501250716,
+      "learning_rate": 1.996935292870437e-05,
+      "loss": 0.8175,
+      "step": 709
+    },
+    {
+      "epoch": 0.10864575363427698,
+      "grad_norm": 2.9447725337658444,
+      "learning_rate": 1.9969158752909247e-05,
+      "loss": 0.7753,
+      "step": 710
+    },
+    {
+      "epoch": 0.10879877582249427,
+      "grad_norm": 2.78759682667099,
+      "learning_rate": 1.9968963964869088e-05,
+      "loss": 0.958,
+      "step": 711
+    },
+    {
+      "epoch": 0.10895179801071156,
+      "grad_norm": 3.074769801484582,
+      "learning_rate": 1.9968768564595856e-05,
+      "loss": 0.9086,
+      "step": 712
+    },
+    {
+      "epoch": 0.10910482019892885,
+      "grad_norm": 2.9070592094240797,
+      "learning_rate": 1.9968572552101544e-05,
+      "loss": 0.9174,
+      "step": 713
+    },
+    {
+      "epoch": 0.10925784238714614,
+      "grad_norm": 3.233405582020585,
+      "learning_rate": 1.9968375927398195e-05,
+      "loss": 0.9356,
+      "step": 714
+    },
+    {
+      "epoch": 0.10941086457536343,
+      "grad_norm": 3.022272822063928,
+      "learning_rate": 1.9968178690497884e-05,
+      "loss": 0.851,
+      "step": 715
+    },
+    {
+      "epoch": 0.10956388676358073,
+      "grad_norm": 2.8195304743616285,
+      "learning_rate": 1.9967980841412722e-05,
+      "loss": 0.7116,
+      "step": 716
+    },
+    {
+      "epoch": 0.10971690895179802,
+      "grad_norm": 2.9245399474491602,
+      "learning_rate": 1.9967782380154864e-05,
+      "loss": 0.8679,
+      "step": 717
+    },
+    {
+      "epoch": 0.10986993114001531,
+      "grad_norm": 2.988694346304424,
+      "learning_rate": 1.9967583306736494e-05,
+      "loss": 0.7599,
+      "step": 718
+    },
+    {
+      "epoch": 0.1100229533282326,
+      "grad_norm": 3.1688493992216618,
+      "learning_rate": 1.996738362116984e-05,
+      "loss": 0.9484,
+      "step": 719
+    },
+    {
+      "epoch": 0.11017597551644988,
+      "grad_norm": 2.9048182310387776,
+      "learning_rate": 1.996718332346717e-05,
+      "loss": 0.7532,
+      "step": 720
+    },
+    {
+      "epoch": 0.11032899770466717,
+      "grad_norm": 2.637213696698448,
+      "learning_rate": 1.9966982413640772e-05,
+      "loss": 0.8024,
+      "step": 721
+    },
+    {
+      "epoch": 0.11048201989288446,
+      "grad_norm": 2.89859467937182,
+      "learning_rate": 1.9966780891703e-05,
+      "loss": 0.9697,
+      "step": 722
+    },
+    {
+      "epoch": 0.11063504208110175,
+      "grad_norm": 2.836160191245258,
+      "learning_rate": 1.996657875766622e-05,
+      "loss": 0.8161,
+      "step": 723
+    },
+    {
+      "epoch": 0.11078806426931904,
+      "grad_norm": 3.0972974916050564,
+      "learning_rate": 1.996637601154285e-05,
+      "loss": 0.905,
+      "step": 724
+    },
+    {
+      "epoch": 0.11094108645753634,
+      "grad_norm": 2.746119102608403,
+      "learning_rate": 1.9966172653345337e-05,
+      "loss": 0.7602,
+      "step": 725
+    },
+    {
+      "epoch": 0.11109410864575363,
+      "grad_norm": 2.6662164718564765,
+      "learning_rate": 1.9965968683086177e-05,
+      "loss": 0.8078,
+      "step": 726
+    },
+    {
+      "epoch": 0.11124713083397092,
+      "grad_norm": 3.135727407852158,
+      "learning_rate": 1.9965764100777892e-05,
+      "loss": 0.8687,
+      "step": 727
+    },
+    {
+      "epoch": 0.11140015302218821,
+      "grad_norm": 2.902712664303539,
+      "learning_rate": 1.996555890643305e-05,
+      "loss": 0.8878,
+      "step": 728
+    },
+    {
+      "epoch": 0.1115531752104055,
+      "grad_norm": 3.2232593590832535,
+      "learning_rate": 1.9965353100064246e-05,
+      "loss": 0.8875,
+      "step": 729
+    },
+    {
+      "epoch": 0.1117061973986228,
+      "grad_norm": 3.098086025014026,
+      "learning_rate": 1.9965146681684126e-05,
+      "loss": 0.8271,
+      "step": 730
+    },
+    {
+      "epoch": 0.11185921958684009,
+      "grad_norm": 2.852123909711368,
+      "learning_rate": 1.996493965130536e-05,
+      "loss": 0.7657,
+      "step": 731
+    },
+    {
+      "epoch": 0.11201224177505738,
+      "grad_norm": 3.08617436112628,
+      "learning_rate": 1.9964732008940673e-05,
+      "loss": 0.9049,
+      "step": 732
+    },
+    {
+      "epoch": 0.11216526396327467,
+      "grad_norm": 3.0408480858160787,
+      "learning_rate": 1.996452375460281e-05,
+      "loss": 0.9891,
+      "step": 733
+    },
+    {
+      "epoch": 0.11231828615149196,
+      "grad_norm": 2.5643968058230313,
+      "learning_rate": 1.9964314888304563e-05,
+      "loss": 0.8169,
+      "step": 734
+    },
+    {
+      "epoch": 0.11247130833970925,
+      "grad_norm": 3.2776028562449184,
+      "learning_rate": 1.9964105410058754e-05,
+      "loss": 0.8984,
+      "step": 735
+    },
+    {
+      "epoch": 0.11262433052792654,
+      "grad_norm": 3.1486146940763375,
+      "learning_rate": 1.9963895319878252e-05,
+      "loss": 0.8662,
+      "step": 736
+    },
+    {
+      "epoch": 0.11277735271614384,
+      "grad_norm": 3.0577728216797855,
+      "learning_rate": 1.996368461777596e-05,
+      "loss": 1.0304,
+      "step": 737
+    },
+    {
+      "epoch": 0.11293037490436113,
+      "grad_norm": 2.558019479134406,
+      "learning_rate": 1.996347330376482e-05,
+      "loss": 0.7169,
+      "step": 738
+    },
+    {
+      "epoch": 0.11308339709257842,
+      "grad_norm": 2.7971191724016577,
+      "learning_rate": 1.9963261377857805e-05,
+      "loss": 0.8585,
+      "step": 739
+    },
+    {
+      "epoch": 0.11323641928079571,
+      "grad_norm": 2.7918459247533183,
+      "learning_rate": 1.996304884006793e-05,
+      "loss": 0.8193,
+      "step": 740
+    },
+    {
+      "epoch": 0.113389441469013,
+      "grad_norm": 3.0171317643440276,
+      "learning_rate": 1.9962835690408255e-05,
+      "loss": 0.9036,
+      "step": 741
+    },
+    {
+      "epoch": 0.1135424636572303,
+      "grad_norm": 2.825490934258436,
+      "learning_rate": 1.9962621928891863e-05,
+      "loss": 0.6977,
+      "step": 742
+    },
+    {
+      "epoch": 0.11369548584544759,
+      "grad_norm": 2.946410812576731,
+      "learning_rate": 1.9962407555531884e-05,
+      "loss": 0.9062,
+      "step": 743
+    },
+    {
+      "epoch": 0.11384850803366488,
+      "grad_norm": 2.63568342518963,
+      "learning_rate": 1.9962192570341485e-05,
+      "loss": 0.8382,
+      "step": 744
+    },
+    {
+      "epoch": 0.11400153022188217,
+      "grad_norm": 2.6540164369542145,
+      "learning_rate": 1.9961976973333868e-05,
+      "loss": 0.8843,
+      "step": 745
+    },
+    {
+      "epoch": 0.11415455241009946,
+      "grad_norm": 3.1360366019870254,
+      "learning_rate": 1.996176076452227e-05,
+      "loss": 0.9828,
+      "step": 746
+    },
+    {
+      "epoch": 0.11430757459831675,
+      "grad_norm": 2.4591066537053603,
+      "learning_rate": 1.996154394391998e-05,
+      "loss": 0.8476,
+      "step": 747
+    },
+    {
+      "epoch": 0.11446059678653404,
+      "grad_norm": 2.9487269261454645,
+      "learning_rate": 1.9961326511540303e-05,
+      "loss": 0.907,
+      "step": 748
+    },
+    {
+      "epoch": 0.11461361897475134,
+      "grad_norm": 2.9947147155525338,
+      "learning_rate": 1.996110846739659e-05,
+      "loss": 0.8579,
+      "step": 749
+    },
+    {
+      "epoch": 0.11476664116296863,
+      "grad_norm": 3.037066986180684,
+      "learning_rate": 1.9960889811502247e-05,
+      "loss": 0.8599,
+      "step": 750
+    },
+    {
+      "epoch": 0.11491966335118592,
+      "grad_norm": 3.051458512563892,
+      "learning_rate": 1.9960670543870692e-05,
+      "loss": 0.8731,
+      "step": 751
+    },
+    {
+      "epoch": 0.11507268553940321,
+      "grad_norm": 2.8398841037795317,
+      "learning_rate": 1.996045066451539e-05,
+      "loss": 0.8207,
+      "step": 752
+    },
+    {
+      "epoch": 0.1152257077276205,
+      "grad_norm": 2.7469614353497955,
+      "learning_rate": 1.9960230173449845e-05,
+      "loss": 0.9335,
+      "step": 753
+    },
+    {
+      "epoch": 0.1153787299158378,
+      "grad_norm": 2.935660071595326,
+      "learning_rate": 1.9960009070687603e-05,
+      "loss": 0.8881,
+      "step": 754
+    },
+    {
+      "epoch": 0.11553175210405509,
+      "grad_norm": 2.5830086234835994,
+      "learning_rate": 1.9959787356242243e-05,
+      "loss": 0.8157,
+      "step": 755
+    },
+    {
+      "epoch": 0.11568477429227238,
+      "grad_norm": 2.902450900848275,
+      "learning_rate": 1.9959565030127375e-05,
+      "loss": 0.8613,
+      "step": 756
+    },
+    {
+      "epoch": 0.11583779648048967,
+      "grad_norm": 3.1217190149525544,
+      "learning_rate": 1.9959342092356656e-05,
+      "loss": 0.8571,
+      "step": 757
+    },
+    {
+      "epoch": 0.11599081866870696,
+      "grad_norm": 3.0119235092110737,
+      "learning_rate": 1.995911854294378e-05,
+      "loss": 0.7914,
+      "step": 758
+    },
+    {
+      "epoch": 0.11614384085692425,
+      "grad_norm": 3.270646329337315,
+      "learning_rate": 1.9958894381902473e-05,
+      "loss": 0.9937,
+      "step": 759
+    },
+    {
+      "epoch": 0.11629686304514154,
+      "grad_norm": 2.6432702250244597,
+      "learning_rate": 1.99586696092465e-05,
+      "loss": 0.9079,
+      "step": 760
+    },
+    {
+      "epoch": 0.11644988523335884,
+      "grad_norm": 2.86168615726239,
+      "learning_rate": 1.9958444224989673e-05,
+      "loss": 0.7816,
+      "step": 761
+    },
+    {
+      "epoch": 0.11660290742157613,
+      "grad_norm": 3.1978664221309447,
+      "learning_rate": 1.9958218229145828e-05,
+      "loss": 0.8444,
+      "step": 762
+    },
+    {
+      "epoch": 0.11675592960979342,
+      "grad_norm": 2.932323475678271,
+      "learning_rate": 1.995799162172884e-05,
+      "loss": 0.9456,
+      "step": 763
+    },
+    {
+      "epoch": 0.11690895179801071,
+      "grad_norm": 3.038083319492739,
+      "learning_rate": 1.9957764402752632e-05,
+      "loss": 0.9396,
+      "step": 764
+    },
+    {
+      "epoch": 0.117061973986228,
+      "grad_norm": 2.848333848765862,
+      "learning_rate": 1.995753657223116e-05,
+      "loss": 0.9366,
+      "step": 765
+    },
+    {
+      "epoch": 0.1172149961744453,
+      "grad_norm": 3.0057853013541935,
+      "learning_rate": 1.995730813017841e-05,
+      "loss": 0.9175,
+      "step": 766
+    },
+    {
+      "epoch": 0.11736801836266259,
+      "grad_norm": 2.872981414780168,
+      "learning_rate": 1.9957079076608416e-05,
+      "loss": 0.904,
+      "step": 767
+    },
+    {
+      "epoch": 0.11752104055087988,
+      "grad_norm": 2.77278331612368,
+      "learning_rate": 1.9956849411535243e-05,
+      "loss": 0.8077,
+      "step": 768
+    },
+    {
+      "epoch": 0.11767406273909717,
+      "grad_norm": 2.809951235769412,
+      "learning_rate": 1.9956619134973e-05,
+      "loss": 0.8337,
+      "step": 769
+    },
+    {
+      "epoch": 0.11782708492731446,
+      "grad_norm": 2.5883083215594413,
+      "learning_rate": 1.995638824693582e-05,
+      "loss": 0.8807,
+      "step": 770
+    },
+    {
+      "epoch": 0.11798010711553175,
+      "grad_norm": 2.8036993992336203,
+      "learning_rate": 1.9956156747437892e-05,
+      "loss": 0.8703,
+      "step": 771
+    },
+    {
+      "epoch": 0.11813312930374904,
+      "grad_norm": 3.3243628707668234,
+      "learning_rate": 1.9955924636493427e-05,
+      "loss": 0.7931,
+      "step": 772
+    },
+    {
+      "epoch": 0.11828615149196633,
+      "grad_norm": 2.873881310062135,
+      "learning_rate": 1.995569191411668e-05,
+      "loss": 0.9049,
+      "step": 773
+    },
+    {
+      "epoch": 0.11843917368018363,
+      "grad_norm": 2.7069595333520895,
+      "learning_rate": 1.995545858032195e-05,
+      "loss": 0.8754,
+      "step": 774
+    },
+    {
+      "epoch": 0.11859219586840092,
+      "grad_norm": 2.9096351619045002,
+      "learning_rate": 1.9955224635123563e-05,
+      "loss": 0.7806,
+      "step": 775
+    },
+    {
+      "epoch": 0.11874521805661821,
+      "grad_norm": 2.8827356138086566,
+      "learning_rate": 1.9954990078535882e-05,
+      "loss": 0.8809,
+      "step": 776
+    },
+    {
+      "epoch": 0.1188982402448355,
+      "grad_norm": 2.826962196189732,
+      "learning_rate": 1.9954754910573322e-05,
+      "loss": 0.87,
+      "step": 777
+    },
+    {
+      "epoch": 0.11905126243305279,
+      "grad_norm": 2.8069784613392637,
+      "learning_rate": 1.9954519131250315e-05,
+      "loss": 0.8496,
+      "step": 778
+    },
+    {
+      "epoch": 0.11920428462127008,
+      "grad_norm": 2.971601835723486,
+      "learning_rate": 1.9954282740581347e-05,
+      "loss": 0.8614,
+      "step": 779
+    },
+    {
+      "epoch": 0.11935730680948738,
+      "grad_norm": 2.9075497971975746,
+      "learning_rate": 1.9954045738580935e-05,
+      "loss": 0.8203,
+      "step": 780
+    },
+    {
+      "epoch": 0.11951032899770467,
+      "grad_norm": 2.9402914645046336,
+      "learning_rate": 1.9953808125263634e-05,
+      "loss": 0.9068,
+      "step": 781
+    },
+    {
+      "epoch": 0.11966335118592196,
+      "grad_norm": 2.7248354648371245,
+      "learning_rate": 1.9953569900644038e-05,
+      "loss": 0.8354,
+      "step": 782
+    },
+    {
+      "epoch": 0.11981637337413925,
+      "grad_norm": 2.738997062960359,
+      "learning_rate": 1.9953331064736772e-05,
+      "loss": 0.8416,
+      "step": 783
+    },
+    {
+      "epoch": 0.11996939556235654,
+      "grad_norm": 2.408678484778986,
+      "learning_rate": 1.9953091617556508e-05,
+      "loss": 0.9647,
+      "step": 784
+    },
+    {
+      "epoch": 0.12012241775057383,
+      "grad_norm": 3.397543310223297,
+      "learning_rate": 1.995285155911795e-05,
+      "loss": 0.9912,
+      "step": 785
+    },
+    {
+      "epoch": 0.12027543993879113,
+      "grad_norm": 2.492146839514386,
+      "learning_rate": 1.9952610889435847e-05,
+      "loss": 0.7809,
+      "step": 786
+    },
+    {
+      "epoch": 0.12042846212700842,
+      "grad_norm": 3.186689185906449,
+      "learning_rate": 1.995236960852497e-05,
+      "loss": 0.8792,
+      "step": 787
+    },
+    {
+      "epoch": 0.12058148431522571,
+      "grad_norm": 2.6501731876923884,
+      "learning_rate": 1.9952127716400147e-05,
+      "loss": 0.9151,
+      "step": 788
+    },
+    {
+      "epoch": 0.120734506503443,
+      "grad_norm": 2.5758608454690717,
+      "learning_rate": 1.9951885213076224e-05,
+      "loss": 0.772,
+      "step": 789
+    },
+    {
+      "epoch": 0.12088752869166029,
+      "grad_norm": 2.8652593890546627,
+      "learning_rate": 1.9951642098568098e-05,
+      "loss": 0.8747,
+      "step": 790
+    },
+    {
+      "epoch": 0.12104055087987758,
+      "grad_norm": 3.0338057992308727,
+      "learning_rate": 1.9951398372890698e-05,
+      "loss": 0.8661,
+      "step": 791
+    },
+    {
+      "epoch": 0.12119357306809488,
+      "grad_norm": 2.921221055918673,
+      "learning_rate": 1.9951154036058996e-05,
+      "loss": 0.9477,
+      "step": 792
+    },
+    {
+      "epoch": 0.12134659525631217,
+      "grad_norm": 2.9305423440900045,
+      "learning_rate": 1.9950909088087998e-05,
+      "loss": 0.8084,
+      "step": 793
+    },
+    {
+      "epoch": 0.12149961744452946,
+      "grad_norm": 2.8420265459295773,
+      "learning_rate": 1.995066352899274e-05,
+      "loss": 0.7906,
+      "step": 794
+    },
+    {
+      "epoch": 0.12165263963274675,
+      "grad_norm": 2.6560288005388255,
+      "learning_rate": 1.995041735878831e-05,
+      "loss": 0.9672,
+      "step": 795
+    },
+    {
+      "epoch": 0.12180566182096404,
+      "grad_norm": 2.899871208782027,
+      "learning_rate": 1.9950170577489823e-05,
+      "loss": 0.8082,
+      "step": 796
+    },
+    {
+      "epoch": 0.12195868400918133,
+      "grad_norm": 2.426459161276832,
+      "learning_rate": 1.9949923185112437e-05,
+      "loss": 0.7931,
+      "step": 797
+    },
+    {
+      "epoch": 0.12211170619739863,
+      "grad_norm": 2.840994852998752,
+      "learning_rate": 1.9949675181671343e-05,
+      "loss": 0.7934,
+      "step": 798
+    },
+    {
+      "epoch": 0.12226472838561592,
+      "grad_norm": 2.808195404634968,
+      "learning_rate": 1.9949426567181773e-05,
+      "loss": 0.8969,
+      "step": 799
+    },
+    {
+      "epoch": 0.12241775057383321,
+      "grad_norm": 2.534032946860274,
+      "learning_rate": 1.9949177341658995e-05,
+      "loss": 0.8507,
+      "step": 800
+    },
+    {
+      "epoch": 0.1225707727620505,
+      "grad_norm": 2.9255867633682167,
+      "learning_rate": 1.9948927505118312e-05,
+      "loss": 0.9297,
+      "step": 801
+    },
+    {
+      "epoch": 0.12272379495026779,
+      "grad_norm": 2.644371265898017,
+      "learning_rate": 1.9948677057575074e-05,
+      "loss": 0.7543,
+      "step": 802
+    },
+    {
+      "epoch": 0.12287681713848508,
+      "grad_norm": 3.1021955628731375,
+      "learning_rate": 1.9948425999044657e-05,
+      "loss": 0.8462,
+      "step": 803
+    },
+    {
+      "epoch": 0.12302983932670238,
+      "grad_norm": 2.8070466213969336,
+      "learning_rate": 1.9948174329542483e-05,
+      "loss": 0.8161,
+      "step": 804
+    },
+    {
+      "epoch": 0.12318286151491967,
+      "grad_norm": 2.5722394249819294,
+      "learning_rate": 1.9947922049084007e-05,
+      "loss": 0.8785,
+      "step": 805
+    },
+    {
+      "epoch": 0.12333588370313696,
+      "grad_norm": 2.872258390308318,
+      "learning_rate": 1.9947669157684718e-05,
+      "loss": 0.9878,
+      "step": 806
+    },
+    {
+      "epoch": 0.12348890589135425,
+      "grad_norm": 2.8229511128297933,
+      "learning_rate": 1.9947415655360148e-05,
+      "loss": 0.8757,
+      "step": 807
+    },
+    {
+      "epoch": 0.12364192807957154,
+      "grad_norm": 2.974193887323895,
+      "learning_rate": 1.994716154212587e-05,
+      "loss": 0.8089,
+      "step": 808
+    },
+    {
+      "epoch": 0.12379495026778883,
+      "grad_norm": 2.452950207328061,
+      "learning_rate": 1.9946906817997495e-05,
+      "loss": 0.8796,
+      "step": 809
+    },
+    {
+      "epoch": 0.12394797245600613,
+      "grad_norm": 2.6646128965830385,
+      "learning_rate": 1.9946651482990654e-05,
+      "loss": 0.8494,
+      "step": 810
+    },
+    {
+      "epoch": 0.12410099464422342,
+      "grad_norm": 2.974247053047141,
+      "learning_rate": 1.9946395537121033e-05,
+      "loss": 0.8141,
+      "step": 811
+    },
+    {
+      "epoch": 0.12425401683244071,
+      "grad_norm": 2.77324872264311,
+      "learning_rate": 1.9946138980404352e-05,
+      "loss": 0.8281,
+      "step": 812
+    },
+    {
+      "epoch": 0.124407039020658,
+      "grad_norm": 2.8947035176206204,
+      "learning_rate": 1.994588181285637e-05,
+      "loss": 0.865,
+      "step": 813
+    },
+    {
+      "epoch": 0.12456006120887529,
+      "grad_norm": 2.481415993737972,
+      "learning_rate": 1.9945624034492876e-05,
+      "loss": 0.8163,
+      "step": 814
+    },
+    {
+      "epoch": 0.12471308339709258,
+      "grad_norm": 2.812818009991962,
+      "learning_rate": 1.99453656453297e-05,
+      "loss": 0.8919,
+      "step": 815
+    },
+    {
+      "epoch": 0.12486610558530988,
+      "grad_norm": 2.606521546663759,
+      "learning_rate": 1.9945106645382713e-05,
+      "loss": 0.8239,
+      "step": 816
+    },
+    {
+      "epoch": 0.12501912777352717,
+      "grad_norm": 2.7952461518148692,
+      "learning_rate": 1.9944847034667826e-05,
+      "loss": 0.8532,
+      "step": 817
+    },
+    {
+      "epoch": 0.12517214996174444,
+      "grad_norm": 2.7255334999000977,
+      "learning_rate": 1.9944586813200975e-05,
+      "loss": 0.8189,
+      "step": 818
+    },
+    {
+      "epoch": 0.12532517214996175,
+      "grad_norm": 2.3712871371410578,
+      "learning_rate": 1.9944325980998143e-05,
+      "loss": 0.7809,
+      "step": 819
+    },
+    {
+      "epoch": 0.12547819433817903,
+      "grad_norm": 2.7615076995086567,
+      "learning_rate": 1.9944064538075355e-05,
+      "loss": 0.8229,
+      "step": 820
+    },
+    {
+      "epoch": 0.12563121652639633,
+      "grad_norm": 2.932395122434818,
+      "learning_rate": 1.994380248444866e-05,
+      "loss": 0.862,
+      "step": 821
+    },
+    {
+      "epoch": 0.1257842387146136,
+      "grad_norm": 2.6788048670584175,
+      "learning_rate": 1.994353982013415e-05,
+      "loss": 0.8015,
+      "step": 822
+    },
+    {
+      "epoch": 0.12593726090283092,
+      "grad_norm": 3.0278059097501893,
+      "learning_rate": 1.9943276545147966e-05,
+      "loss": 0.9526,
+      "step": 823
+    },
+    {
+      "epoch": 0.1260902830910482,
+      "grad_norm": 2.9252691667903608,
+      "learning_rate": 1.9943012659506268e-05,
+      "loss": 0.8529,
+      "step": 824
+    },
+    {
+      "epoch": 0.1262433052792655,
+      "grad_norm": 2.8583585269123994,
+      "learning_rate": 1.9942748163225264e-05,
+      "loss": 0.8131,
+      "step": 825
+    },
+    {
+      "epoch": 0.12639632746748278,
+      "grad_norm": 3.2704363944279664,
+      "learning_rate": 1.9942483056321204e-05,
+      "loss": 0.8934,
+      "step": 826
+    },
+    {
+      "epoch": 0.12654934965570008,
+      "grad_norm": 2.9576396568597705,
+      "learning_rate": 1.994221733881036e-05,
+      "loss": 0.9039,
+      "step": 827
+    },
+    {
+      "epoch": 0.12670237184391736,
+      "grad_norm": 2.5940180643661406,
+      "learning_rate": 1.9941951010709054e-05,
+      "loss": 0.891,
+      "step": 828
+    },
+    {
+      "epoch": 0.12685539403213467,
+      "grad_norm": 2.7471602663867722,
+      "learning_rate": 1.9941684072033646e-05,
+      "loss": 0.8782,
+      "step": 829
+    },
+    {
+      "epoch": 0.12700841622035194,
+      "grad_norm": 3.0403029906730894,
+      "learning_rate": 1.9941416522800528e-05,
+      "loss": 0.8783,
+      "step": 830
+    },
+    {
+      "epoch": 0.12716143840856925,
+      "grad_norm": 2.967016562837652,
+      "learning_rate": 1.994114836302613e-05,
+      "loss": 0.9094,
+      "step": 831
+    },
+    {
+      "epoch": 0.12731446059678653,
+      "grad_norm": 2.697380123869265,
+      "learning_rate": 1.994087959272692e-05,
+      "loss": 0.8034,
+      "step": 832
+    },
+    {
+      "epoch": 0.12746748278500383,
+      "grad_norm": 2.6505703341369,
+      "learning_rate": 1.9940610211919404e-05,
+      "loss": 0.762,
+      "step": 833
+    },
+    {
+      "epoch": 0.1276205049732211,
+      "grad_norm": 2.9265591451269377,
+      "learning_rate": 1.9940340220620128e-05,
+      "loss": 0.9365,
+      "step": 834
+    },
+    {
+      "epoch": 0.12777352716143842,
+      "grad_norm": 2.6800900011379114,
+      "learning_rate": 1.9940069618845674e-05,
+      "loss": 0.8783,
+      "step": 835
+    },
+    {
+      "epoch": 0.1279265493496557,
+      "grad_norm": 2.7506250769096963,
+      "learning_rate": 1.9939798406612657e-05,
+      "loss": 0.8799,
+      "step": 836
+    },
+    {
+      "epoch": 0.128079571537873,
+      "grad_norm": 2.77472886449925,
+      "learning_rate": 1.9939526583937736e-05,
+      "loss": 0.7991,
+      "step": 837
+    },
+    {
+      "epoch": 0.12823259372609028,
+      "grad_norm": 2.8680315970983177,
+      "learning_rate": 1.9939254150837603e-05,
+      "loss": 0.7995,
+      "step": 838
+    },
+    {
+      "epoch": 0.12838561591430758,
+      "grad_norm": 2.6373678887412253,
+      "learning_rate": 1.993898110732899e-05,
+      "loss": 0.8318,
+      "step": 839
+    },
+    {
+      "epoch": 0.12853863810252486,
+      "grad_norm": 2.912194374140499,
+      "learning_rate": 1.9938707453428665e-05,
+      "loss": 0.8977,
+      "step": 840
+    },
+    {
+      "epoch": 0.12869166029074217,
+      "grad_norm": 2.664614405723492,
+      "learning_rate": 1.9938433189153437e-05,
+      "loss": 0.7767,
+      "step": 841
+    },
+    {
+      "epoch": 0.12884468247895944,
+      "grad_norm": 2.787127868229937,
+      "learning_rate": 1.9938158314520145e-05,
+      "loss": 0.882,
+      "step": 842
+    },
+    {
+      "epoch": 0.12899770466717675,
+      "grad_norm": 3.050383550626497,
+      "learning_rate": 1.9937882829545673e-05,
+      "loss": 0.916,
+      "step": 843
+    },
+    {
+      "epoch": 0.12915072685539403,
+      "grad_norm": 2.749560412191003,
+      "learning_rate": 1.9937606734246943e-05,
+      "loss": 0.8087,
+      "step": 844
+    },
+    {
+      "epoch": 0.12930374904361133,
+      "grad_norm": 3.0722048039396954,
+      "learning_rate": 1.9937330028640903e-05,
+      "loss": 0.8669,
+      "step": 845
+    },
+    {
+      "epoch": 0.1294567712318286,
+      "grad_norm": 2.7330294561178254,
+      "learning_rate": 1.9937052712744552e-05,
+      "loss": 0.9029,
+      "step": 846
+    },
+    {
+      "epoch": 0.12960979342004592,
+      "grad_norm": 2.846524441855076,
+      "learning_rate": 1.993677478657492e-05,
+      "loss": 0.8833,
+      "step": 847
+    },
+    {
+      "epoch": 0.1297628156082632,
+      "grad_norm": 2.7819980747272983,
+      "learning_rate": 1.9936496250149077e-05,
+      "loss": 0.8333,
+      "step": 848
+    },
+    {
+      "epoch": 0.1299158377964805,
+      "grad_norm": 2.8784712404720016,
+      "learning_rate": 1.9936217103484126e-05,
+      "loss": 0.8171,
+      "step": 849
+    },
+    {
+      "epoch": 0.13006885998469778,
+      "grad_norm": 2.6849041458366423,
+      "learning_rate": 1.9935937346597213e-05,
+      "loss": 0.8568,
+      "step": 850
+    },
+    {
+      "epoch": 0.13022188217291508,
+      "grad_norm": 2.7137336876141913,
+      "learning_rate": 1.9935656979505518e-05,
+      "loss": 0.7485,
+      "step": 851
+    },
+    {
+      "epoch": 0.13037490436113236,
+      "grad_norm": 2.679237726202307,
+      "learning_rate": 1.993537600222626e-05,
+      "loss": 0.8848,
+      "step": 852
+    },
+    {
+      "epoch": 0.13052792654934967,
+      "grad_norm": 2.9623536876476013,
+      "learning_rate": 1.993509441477669e-05,
+      "loss": 0.9203,
+      "step": 853
+    },
+    {
+      "epoch": 0.13068094873756694,
+      "grad_norm": 3.0078297852392613,
+      "learning_rate": 1.9934812217174112e-05,
+      "loss": 0.8651,
+      "step": 854
+    },
+    {
+      "epoch": 0.13083397092578425,
+      "grad_norm": 2.924067755236479,
+      "learning_rate": 1.9934529409435845e-05,
+      "loss": 0.7741,
+      "step": 855
+    },
+    {
+      "epoch": 0.13098699311400153,
+      "grad_norm": 2.7098897161661557,
+      "learning_rate": 1.9934245991579265e-05,
+      "loss": 0.8354,
+      "step": 856
+    },
+    {
+      "epoch": 0.13114001530221883,
+      "grad_norm": 2.650605982041534,
+      "learning_rate": 1.9933961963621777e-05,
+      "loss": 0.819,
+      "step": 857
+    },
+    {
+      "epoch": 0.1312930374904361,
+      "grad_norm": 2.9612775731325844,
+      "learning_rate": 1.993367732558082e-05,
+      "loss": 0.9485,
+      "step": 858
+    },
+    {
+      "epoch": 0.13144605967865342,
+      "grad_norm": 2.6114054372319644,
+      "learning_rate": 1.993339207747388e-05,
+      "loss": 0.7404,
+      "step": 859
+    },
+    {
+      "epoch": 0.1315990818668707,
+      "grad_norm": 3.0422415925389,
+      "learning_rate": 1.9933106219318474e-05,
+      "loss": 0.8357,
+      "step": 860
+    },
+    {
+      "epoch": 0.131752104055088,
+      "grad_norm": 2.898403955243632,
+      "learning_rate": 1.9932819751132153e-05,
+      "loss": 0.78,
+      "step": 861
+    },
+    {
+      "epoch": 0.13190512624330528,
+      "grad_norm": 2.9726198703448383,
+      "learning_rate": 1.9932532672932515e-05,
+      "loss": 0.7844,
+      "step": 862
+    },
+    {
+      "epoch": 0.13205814843152258,
+      "grad_norm": 2.961682855155522,
+      "learning_rate": 1.993224498473719e-05,
+      "loss": 0.8929,
+      "step": 863
+    },
+    {
+      "epoch": 0.13221117061973986,
+      "grad_norm": 2.9954828687652064,
+      "learning_rate": 1.9931956686563848e-05,
+      "loss": 0.9105,
+      "step": 864
+    },
+    {
+      "epoch": 0.13236419280795717,
+      "grad_norm": 2.5682006472322163,
+      "learning_rate": 1.9931667778430188e-05,
+      "loss": 0.717,
+      "step": 865
+    },
+    {
+      "epoch": 0.13251721499617444,
+      "grad_norm": 2.771936142857185,
+      "learning_rate": 1.9931378260353957e-05,
+      "loss": 0.8252,
+      "step": 866
+    },
+    {
+      "epoch": 0.13267023718439175,
+      "grad_norm": 2.6416325626387613,
+      "learning_rate": 1.9931088132352933e-05,
+      "loss": 0.7803,
+      "step": 867
+    },
+    {
+      "epoch": 0.13282325937260903,
+      "grad_norm": 2.833863980241223,
+      "learning_rate": 1.993079739444494e-05,
+      "loss": 0.9553,
+      "step": 868
+    },
+    {
+      "epoch": 0.13297628156082633,
+      "grad_norm": 2.6645516366644424,
+      "learning_rate": 1.993050604664783e-05,
+      "loss": 0.7632,
+      "step": 869
+    },
+    {
+      "epoch": 0.1331293037490436,
+      "grad_norm": 2.7887899146599278,
+      "learning_rate": 1.9930214088979492e-05,
+      "loss": 0.7871,
+      "step": 870
+    },
+    {
+      "epoch": 0.13328232593726091,
+      "grad_norm": 2.7303368182841417,
+      "learning_rate": 1.9929921521457865e-05,
+      "loss": 0.8759,
+      "step": 871
+    },
+    {
+      "epoch": 0.1334353481254782,
+      "grad_norm": 2.5203669088161336,
+      "learning_rate": 1.9929628344100907e-05,
+      "loss": 0.8028,
+      "step": 872
+    },
+    {
+      "epoch": 0.1335883703136955,
+      "grad_norm": 2.7611918709365124,
+      "learning_rate": 1.9929334556926628e-05,
+      "loss": 0.8294,
+      "step": 873
+    },
+    {
+      "epoch": 0.13374139250191278,
+      "grad_norm": 2.6936499626342556,
+      "learning_rate": 1.992904015995307e-05,
+      "loss": 0.8079,
+      "step": 874
+    },
+    {
+      "epoch": 0.13389441469013008,
+      "grad_norm": 2.593917756784823,
+      "learning_rate": 1.9928745153198313e-05,
+      "loss": 0.7629,
+      "step": 875
+    },
+    {
+      "epoch": 0.13404743687834736,
+      "grad_norm": 2.578488093197608,
+      "learning_rate": 1.9928449536680476e-05,
+      "loss": 0.7517,
+      "step": 876
+    },
+    {
+      "epoch": 0.13420045906656466,
+      "grad_norm": 2.499522490435344,
+      "learning_rate": 1.9928153310417712e-05,
+      "loss": 0.8801,
+      "step": 877
+    },
+    {
+      "epoch": 0.13435348125478194,
+      "grad_norm": 2.867320358801422,
+      "learning_rate": 1.9927856474428215e-05,
+      "loss": 0.9224,
+      "step": 878
+    },
+    {
+      "epoch": 0.13450650344299925,
+      "grad_norm": 2.5535956136527846,
+      "learning_rate": 1.9927559028730212e-05,
+      "loss": 0.7598,
+      "step": 879
+    },
+    {
+      "epoch": 0.13465952563121653,
+      "grad_norm": 2.779439531223504,
+      "learning_rate": 1.992726097334197e-05,
+      "loss": 0.9068,
+      "step": 880
+    },
+    {
+      "epoch": 0.13481254781943383,
+      "grad_norm": 2.821228950007975,
+      "learning_rate": 1.9926962308281802e-05,
+      "loss": 0.7944,
+      "step": 881
+    },
+    {
+      "epoch": 0.1349655700076511,
+      "grad_norm": 2.6330247037928824,
+      "learning_rate": 1.992666303356804e-05,
+      "loss": 0.7822,
+      "step": 882
+    },
+    {
+      "epoch": 0.13511859219586841,
+      "grad_norm": 2.840038821329292,
+      "learning_rate": 1.992636314921907e-05,
+      "loss": 0.8301,
+      "step": 883
+    },
+    {
+      "epoch": 0.1352716143840857,
+      "grad_norm": 2.6793404976515434,
+      "learning_rate": 1.9926062655253305e-05,
+      "loss": 0.7915,
+      "step": 884
+    },
+    {
+      "epoch": 0.135424636572303,
+      "grad_norm": 2.689530769295994,
+      "learning_rate": 1.9925761551689203e-05,
+      "loss": 0.9103,
+      "step": 885
+    },
+    {
+      "epoch": 0.13557765876052028,
+      "grad_norm": 2.887001810355891,
+      "learning_rate": 1.9925459838545252e-05,
+      "loss": 0.8768,
+      "step": 886
+    },
+    {
+      "epoch": 0.13573068094873755,
+      "grad_norm": 2.877012186365154,
+      "learning_rate": 1.9925157515839984e-05,
+      "loss": 0.8887,
+      "step": 887
+    },
+    {
+      "epoch": 0.13588370313695486,
+      "grad_norm": 2.8588587834893406,
+      "learning_rate": 1.992485458359197e-05,
+      "loss": 0.8792,
+      "step": 888
+    },
+    {
+      "epoch": 0.13603672532517214,
+      "grad_norm": 3.1905984297528964,
+      "learning_rate": 1.9924551041819807e-05,
+      "loss": 0.895,
+      "step": 889
+    },
+    {
+      "epoch": 0.13618974751338944,
+      "grad_norm": 2.7402318928244083,
+      "learning_rate": 1.9924246890542137e-05,
+      "loss": 0.8693,
+      "step": 890
+    },
+    {
+      "epoch": 0.13634276970160672,
+      "grad_norm": 2.818887709481401,
+      "learning_rate": 1.9923942129777644e-05,
+      "loss": 0.7851,
+      "step": 891
+    },
+    {
+      "epoch": 0.13649579188982403,
+      "grad_norm": 2.6259577978312407,
+      "learning_rate": 1.992363675954504e-05,
+      "loss": 0.9576,
+      "step": 892
+    },
+    {
+      "epoch": 0.1366488140780413,
+      "grad_norm": 2.863449512098866,
+      "learning_rate": 1.9923330779863084e-05,
+      "loss": 0.9511,
+      "step": 893
+    },
+    {
+      "epoch": 0.1368018362662586,
+      "grad_norm": 2.562371413427193,
+      "learning_rate": 1.992302419075056e-05,
+      "loss": 0.8163,
+      "step": 894
+    },
+    {
+      "epoch": 0.1369548584544759,
+      "grad_norm": 2.8870573577788727,
+      "learning_rate": 1.992271699222631e-05,
+      "loss": 0.9528,
+      "step": 895
+    },
+    {
+      "epoch": 0.1371078806426932,
+      "grad_norm": 2.6937895637721723,
+      "learning_rate": 1.9922409184309184e-05,
+      "loss": 0.8091,
+      "step": 896
+    },
+    {
+      "epoch": 0.13726090283091047,
+      "grad_norm": 2.6630413615518163,
+      "learning_rate": 1.9922100767018095e-05,
+      "loss": 0.7998,
+      "step": 897
+    },
+    {
+      "epoch": 0.13741392501912778,
+      "grad_norm": 2.625511387398233,
+      "learning_rate": 1.9921791740371982e-05,
+      "loss": 0.9357,
+      "step": 898
+    },
+    {
+      "epoch": 0.13756694720734505,
+      "grad_norm": 2.758458249926434,
+      "learning_rate": 1.9921482104389827e-05,
+      "loss": 0.8647,
+      "step": 899
+    },
+    {
+      "epoch": 0.13771996939556236,
+      "grad_norm": 2.976298176316275,
+      "learning_rate": 1.992117185909064e-05,
+      "loss": 0.7932,
+      "step": 900
+    },
+    {
+      "epoch": 0.13787299158377964,
+      "grad_norm": 2.68384833081395,
+      "learning_rate": 1.9920861004493477e-05,
+      "loss": 0.9454,
+      "step": 901
+    },
+    {
+      "epoch": 0.13802601377199694,
+      "grad_norm": 2.7413100902098653,
+      "learning_rate": 1.992054954061743e-05,
+      "loss": 0.8505,
+      "step": 902
+    },
+    {
+      "epoch": 0.13817903596021422,
+      "grad_norm": 3.218955544647937,
+      "learning_rate": 1.9920237467481628e-05,
+      "loss": 0.9033,
+      "step": 903
+    },
+    {
+      "epoch": 0.13833205814843152,
+      "grad_norm": 2.5946224723970035,
+      "learning_rate": 1.9919924785105235e-05,
+      "loss": 0.7389,
+      "step": 904
+    },
+    {
+      "epoch": 0.1384850803366488,
+      "grad_norm": 2.6176807752494726,
+      "learning_rate": 1.991961149350745e-05,
+      "loss": 0.8513,
+      "step": 905
+    },
+    {
+      "epoch": 0.1386381025248661,
+      "grad_norm": 3.1291931843932335,
+      "learning_rate": 1.991929759270752e-05,
+      "loss": 0.8245,
+      "step": 906
+    },
+    {
+      "epoch": 0.13879112471308339,
+      "grad_norm": 2.8736841925384025,
+      "learning_rate": 1.991898308272472e-05,
+      "loss": 0.9262,
+      "step": 907
+    },
+    {
+      "epoch": 0.1389441469013007,
+      "grad_norm": 2.609901479819786,
+      "learning_rate": 1.9918667963578368e-05,
+      "loss": 0.783,
+      "step": 908
+    },
+    {
+      "epoch": 0.13909716908951797,
+      "grad_norm": 2.6331118876099873,
+      "learning_rate": 1.991835223528781e-05,
+      "loss": 0.7673,
+      "step": 909
+    },
+    {
+      "epoch": 0.13925019127773527,
+      "grad_norm": 2.9613767339162305,
+      "learning_rate": 1.9918035897872445e-05,
+      "loss": 0.8643,
+      "step": 910
+    },
+    {
+      "epoch": 0.13940321346595255,
+      "grad_norm": 3.039493882439516,
+      "learning_rate": 1.9917718951351692e-05,
+      "loss": 0.9873,
+      "step": 911
+    },
+    {
+      "epoch": 0.13955623565416986,
+      "grad_norm": 3.239850256883201,
+      "learning_rate": 1.991740139574502e-05,
+      "loss": 0.884,
+      "step": 912
+    },
+    {
+      "epoch": 0.13970925784238714,
+      "grad_norm": 2.8892605380358294,
+      "learning_rate": 1.9917083231071933e-05,
+      "loss": 0.8356,
+      "step": 913
+    },
+    {
+      "epoch": 0.13986228003060444,
+      "grad_norm": 3.2132002955197034,
+      "learning_rate": 1.991676445735197e-05,
+      "loss": 0.837,
+      "step": 914
+    },
+    {
+      "epoch": 0.14001530221882172,
+      "grad_norm": 3.2006127364915913,
+      "learning_rate": 1.9916445074604705e-05,
+      "loss": 0.8063,
+      "step": 915
+    },
+    {
+      "epoch": 0.14016832440703902,
+      "grad_norm": 2.899427273902653,
+      "learning_rate": 1.9916125082849755e-05,
+      "loss": 0.7584,
+      "step": 916
+    },
+    {
+      "epoch": 0.1403213465952563,
+      "grad_norm": 2.914187996315459,
+      "learning_rate": 1.991580448210677e-05,
+      "loss": 0.8503,
+      "step": 917
+    },
+    {
+      "epoch": 0.1404743687834736,
+      "grad_norm": 2.804034994852326,
+      "learning_rate": 1.9915483272395445e-05,
+      "loss": 0.8292,
+      "step": 918
+    },
+    {
+      "epoch": 0.14062739097169089,
+      "grad_norm": 2.680970492007158,
+      "learning_rate": 1.99151614537355e-05,
+      "loss": 0.9034,
+      "step": 919
+    },
+    {
+      "epoch": 0.1407804131599082,
+      "grad_norm": 2.7084201578294516,
+      "learning_rate": 1.9914839026146702e-05,
+      "loss": 0.8518,
+      "step": 920
+    },
+    {
+      "epoch": 0.14093343534812547,
+      "grad_norm": 2.553189897331636,
+      "learning_rate": 1.9914515989648852e-05,
+      "loss": 0.7878,
+      "step": 921
+    },
+    {
+      "epoch": 0.14108645753634277,
+      "grad_norm": 3.050378968443345,
+      "learning_rate": 1.991419234426179e-05,
+      "loss": 0.784,
+      "step": 922
+    },
+    {
+      "epoch": 0.14123947972456005,
+      "grad_norm": 2.977989619703046,
+      "learning_rate": 1.991386809000539e-05,
+      "loss": 1.0098,
+      "step": 923
+    },
+    {
+      "epoch": 0.14139250191277736,
+      "grad_norm": 2.8242683639181667,
+      "learning_rate": 1.991354322689957e-05,
+      "loss": 0.8582,
+      "step": 924
+    },
+    {
+      "epoch": 0.14154552410099464,
+      "grad_norm": 2.7128793061444596,
+      "learning_rate": 1.991321775496428e-05,
+      "loss": 0.7909,
+      "step": 925
+    },
+    {
+      "epoch": 0.14169854628921194,
+      "grad_norm": 2.9979822954093884,
+      "learning_rate": 1.9912891674219502e-05,
+      "loss": 0.9373,
+      "step": 926
+    },
+    {
+      "epoch": 0.14185156847742922,
+      "grad_norm": 2.9482931261488936,
+      "learning_rate": 1.991256498468527e-05,
+      "loss": 0.6891,
+      "step": 927
+    },
+    {
+      "epoch": 0.14200459066564652,
+      "grad_norm": 3.0299388842655723,
+      "learning_rate": 1.9912237686381643e-05,
+      "loss": 0.8743,
+      "step": 928
+    },
+    {
+      "epoch": 0.1421576128538638,
+      "grad_norm": 2.6801748229278917,
+      "learning_rate": 1.991190977932872e-05,
+      "loss": 0.863,
+      "step": 929
+    },
+    {
+      "epoch": 0.1423106350420811,
+      "grad_norm": 3.085060897266117,
+      "learning_rate": 1.9911581263546643e-05,
+      "loss": 0.8992,
+      "step": 930
+    },
+    {
+      "epoch": 0.14246365723029839,
+      "grad_norm": 2.800811746754135,
+      "learning_rate": 1.991125213905559e-05,
+      "loss": 0.9399,
+      "step": 931
+    },
+    {
+      "epoch": 0.1426166794185157,
+      "grad_norm": 2.830458349120088,
+      "learning_rate": 1.991092240587577e-05,
+      "loss": 0.9702,
+      "step": 932
+    },
+    {
+      "epoch": 0.14276970160673297,
+      "grad_norm": 2.8432982293955016,
+      "learning_rate": 1.991059206402743e-05,
+      "loss": 0.8011,
+      "step": 933
+    },
+    {
+      "epoch": 0.14292272379495027,
+      "grad_norm": 2.739167967766181,
+      "learning_rate": 1.9910261113530863e-05,
+      "loss": 0.8164,
+      "step": 934
+    },
+    {
+      "epoch": 0.14307574598316755,
+      "grad_norm": 2.563160167423824,
+      "learning_rate": 1.9909929554406388e-05,
+      "loss": 0.7779,
+      "step": 935
+    },
+    {
+      "epoch": 0.14322876817138486,
+      "grad_norm": 3.186579436293255,
+      "learning_rate": 1.9909597386674374e-05,
+      "loss": 0.7571,
+      "step": 936
+    },
+    {
+      "epoch": 0.14338179035960213,
+      "grad_norm": 2.708795534392877,
+      "learning_rate": 1.990926461035522e-05,
+      "loss": 0.8351,
+      "step": 937
+    },
+    {
+      "epoch": 0.14353481254781944,
+      "grad_norm": 2.5425722820092065,
+      "learning_rate": 1.990893122546936e-05,
+      "loss": 0.8501,
+      "step": 938
+    },
+    {
+      "epoch": 0.14368783473603672,
+      "grad_norm": 2.670809594852604,
+      "learning_rate": 1.990859723203727e-05,
+      "loss": 0.7912,
+      "step": 939
+    },
+    {
+      "epoch": 0.14384085692425402,
+      "grad_norm": 3.1185793441810232,
+      "learning_rate": 1.9908262630079454e-05,
+      "loss": 0.8205,
+      "step": 940
+    },
+    {
+      "epoch": 0.1439938791124713,
+      "grad_norm": 2.731495965280675,
+      "learning_rate": 1.9907927419616477e-05,
+      "loss": 0.7788,
+      "step": 941
+    },
+    {
+      "epoch": 0.1441469013006886,
+      "grad_norm": 3.275322213863118,
+      "learning_rate": 1.9907591600668916e-05,
+      "loss": 1.0141,
+      "step": 942
+    },
+    {
+      "epoch": 0.14429992348890588,
+      "grad_norm": 2.6128430204439765,
+      "learning_rate": 1.9907255173257393e-05,
+      "loss": 0.9174,
+      "step": 943
+    },
+    {
+      "epoch": 0.1444529456771232,
+      "grad_norm": 2.908802026725082,
+      "learning_rate": 1.9906918137402574e-05,
+      "loss": 0.8608,
+      "step": 944
+    },
+    {
+      "epoch": 0.14460596786534047,
+      "grad_norm": 2.7143909626282055,
+      "learning_rate": 1.990658049312516e-05,
+      "loss": 0.772,
+      "step": 945
+    },
+    {
+      "epoch": 0.14475899005355777,
+      "grad_norm": 2.7624040519127706,
+      "learning_rate": 1.9906242240445878e-05,
+      "loss": 0.9278,
+      "step": 946
+    },
+    {
+      "epoch": 0.14491201224177505,
+      "grad_norm": 2.7166448739759246,
+      "learning_rate": 1.990590337938551e-05,
+      "loss": 0.8404,
+      "step": 947
+    },
+    {
+      "epoch": 0.14506503442999236,
+      "grad_norm": 2.9800310838070847,
+      "learning_rate": 1.990556390996486e-05,
+      "loss": 0.9234,
+      "step": 948
+    },
+    {
+      "epoch": 0.14521805661820963,
+      "grad_norm": 3.0606738732407477,
+      "learning_rate": 1.990522383220478e-05,
+      "loss": 0.8102,
+      "step": 949
+    },
+    {
+      "epoch": 0.14537107880642694,
+      "grad_norm": 3.1187386786744176,
+      "learning_rate": 1.9904883146126157e-05,
+      "loss": 0.7964,
+      "step": 950
+    },
+    {
+      "epoch": 0.14552410099464422,
+      "grad_norm": 2.848220059970956,
+      "learning_rate": 1.990454185174991e-05,
+      "loss": 0.897,
+      "step": 951
+    },
+    {
+      "epoch": 0.14567712318286152,
+      "grad_norm": 2.6548873702733613,
+      "learning_rate": 1.9904199949097e-05,
+      "loss": 0.7779,
+      "step": 952
+    },
+    {
+      "epoch": 0.1458301453710788,
+      "grad_norm": 2.846580038404024,
+      "learning_rate": 1.990385743818843e-05,
+      "loss": 0.8872,
+      "step": 953
+    },
+    {
+      "epoch": 0.1459831675592961,
+      "grad_norm": 3.1386716203357143,
+      "learning_rate": 1.9903514319045224e-05,
+      "loss": 0.981,
+      "step": 954
+    },
+    {
+      "epoch": 0.14613618974751338,
+      "grad_norm": 2.6161336399690107,
+      "learning_rate": 1.990317059168847e-05,
+      "loss": 0.8891,
+      "step": 955
+    },
+    {
+      "epoch": 0.1462892119357307,
+      "grad_norm": 2.9030163643864477,
+      "learning_rate": 1.990282625613926e-05,
+      "loss": 0.9119,
+      "step": 956
+    },
+    {
+      "epoch": 0.14644223412394797,
+      "grad_norm": 2.6876433319201656,
+      "learning_rate": 1.9902481312418754e-05,
+      "loss": 0.9507,
+      "step": 957
+    },
+    {
+      "epoch": 0.14659525631216527,
+      "grad_norm": 2.654743185149196,
+      "learning_rate": 1.990213576054813e-05,
+      "loss": 0.7974,
+      "step": 958
+    },
+    {
+      "epoch": 0.14674827850038255,
+      "grad_norm": 2.444896860388207,
+      "learning_rate": 1.9901789600548612e-05,
+      "loss": 0.7975,
+      "step": 959
+    },
+    {
+      "epoch": 0.14690130068859986,
+      "grad_norm": 2.6102754444080176,
+      "learning_rate": 1.990144283244146e-05,
+      "loss": 0.8526,
+      "step": 960
+    },
+    {
+      "epoch": 0.14705432287681713,
+      "grad_norm": 2.3982164679534983,
+      "learning_rate": 1.990109545624797e-05,
+      "loss": 0.7627,
+      "step": 961
+    },
+    {
+      "epoch": 0.14720734506503444,
+      "grad_norm": 2.816740248073328,
+      "learning_rate": 1.990074747198947e-05,
+      "loss": 0.8622,
+      "step": 962
+    },
+    {
+      "epoch": 0.14736036725325172,
+      "grad_norm": 2.3312900266202554,
+      "learning_rate": 1.9900398879687343e-05,
+      "loss": 0.6515,
+      "step": 963
+    },
+    {
+      "epoch": 0.14751338944146902,
+      "grad_norm": 3.27051923019561,
+      "learning_rate": 1.9900049679362982e-05,
+      "loss": 0.9881,
+      "step": 964
+    },
+    {
+      "epoch": 0.1476664116296863,
+      "grad_norm": 2.515455754322227,
+      "learning_rate": 1.9899699871037847e-05,
+      "loss": 0.8661,
+      "step": 965
+    },
+    {
+      "epoch": 0.1478194338179036,
+      "grad_norm": 2.814087317408005,
+      "learning_rate": 1.989934945473341e-05,
+      "loss": 0.8411,
+      "step": 966
+    },
+    {
+      "epoch": 0.14797245600612088,
+      "grad_norm": 2.8017718511765213,
+      "learning_rate": 1.9898998430471202e-05,
+      "loss": 0.9098,
+      "step": 967
+    },
+    {
+      "epoch": 0.1481254781943382,
+      "grad_norm": 2.629421952387108,
+      "learning_rate": 1.9898646798272773e-05,
+      "loss": 0.8979,
+      "step": 968
+    },
+    {
+      "epoch": 0.14827850038255547,
+      "grad_norm": 2.841954083166271,
+      "learning_rate": 1.9898294558159722e-05,
+      "loss": 0.899,
+      "step": 969
+    },
+    {
+      "epoch": 0.14843152257077277,
+      "grad_norm": 2.8553527981636013,
+      "learning_rate": 1.9897941710153677e-05,
+      "loss": 0.8999,
+      "step": 970
+    },
+    {
+      "epoch": 0.14858454475899005,
+      "grad_norm": 2.813663450639181,
+      "learning_rate": 1.989758825427631e-05,
+      "loss": 0.8266,
+      "step": 971
+    },
+    {
+      "epoch": 0.14873756694720736,
+      "grad_norm": 2.8659970900711063,
+      "learning_rate": 1.989723419054933e-05,
+      "loss": 0.7943,
+      "step": 972
+    },
+    {
+      "epoch": 0.14889058913542463,
+      "grad_norm": 2.7945735429929965,
+      "learning_rate": 1.9896879518994483e-05,
+      "loss": 0.794,
+      "step": 973
+    },
+    {
+      "epoch": 0.14904361132364194,
+      "grad_norm": 2.7096673500989565,
+      "learning_rate": 1.9896524239633543e-05,
+      "loss": 0.7983,
+      "step": 974
+    },
+    {
+      "epoch": 0.14919663351185922,
+      "grad_norm": 3.053720034752578,
+      "learning_rate": 1.9896168352488336e-05,
+      "loss": 0.9787,
+      "step": 975
+    },
+    {
+      "epoch": 0.14934965570007652,
+      "grad_norm": 2.6802833840317244,
+      "learning_rate": 1.9895811857580717e-05,
+      "loss": 0.8805,
+      "step": 976
+    },
+    {
+      "epoch": 0.1495026778882938,
+      "grad_norm": 2.9100073841921876,
+      "learning_rate": 1.989545475493258e-05,
+      "loss": 0.9091,
+      "step": 977
+    },
+    {
+      "epoch": 0.1496557000765111,
+      "grad_norm": 2.8957628677948177,
+      "learning_rate": 1.9895097044565853e-05,
+      "loss": 0.8288,
+      "step": 978
+    },
+    {
+      "epoch": 0.14980872226472838,
+      "grad_norm": 3.038264240319414,
+      "learning_rate": 1.989473872650251e-05,
+      "loss": 0.9798,
+      "step": 979
+    },
+    {
+      "epoch": 0.1499617444529457,
+      "grad_norm": 2.5609975366129483,
+      "learning_rate": 1.9894379800764548e-05,
+      "loss": 0.7845,
+      "step": 980
+    },
+    {
+      "epoch": 0.15011476664116297,
+      "grad_norm": 2.7823981767835066,
+      "learning_rate": 1.9894020267374025e-05,
+      "loss": 0.7783,
+      "step": 981
+    },
+    {
+      "epoch": 0.15026778882938027,
+      "grad_norm": 3.3243553824353214,
+      "learning_rate": 1.9893660126353002e-05,
+      "loss": 1.0154,
+      "step": 982
+    },
+    {
+      "epoch": 0.15042081101759755,
+      "grad_norm": 2.480366976477657,
+      "learning_rate": 1.9893299377723608e-05,
+      "loss": 0.7665,
+      "step": 983
+    },
+    {
+      "epoch": 0.15057383320581486,
+      "grad_norm": 2.571561037126822,
+      "learning_rate": 1.9892938021508e-05,
+      "loss": 0.8231,
+      "step": 984
+    },
+    {
+      "epoch": 0.15072685539403213,
+      "grad_norm": 3.066861871830307,
+      "learning_rate": 1.9892576057728366e-05,
+      "loss": 0.8987,
+      "step": 985
+    },
+    {
+      "epoch": 0.15087987758224944,
+      "grad_norm": 3.0035198245097923,
+      "learning_rate": 1.9892213486406937e-05,
+      "loss": 0.8285,
+      "step": 986
+    },
+    {
+      "epoch": 0.15103289977046672,
+      "grad_norm": 2.4764522581212387,
+      "learning_rate": 1.9891850307565976e-05,
+      "loss": 0.7679,
+      "step": 987
+    },
+    {
+      "epoch": 0.15118592195868402,
+      "grad_norm": 2.7267601838920257,
+      "learning_rate": 1.989148652122779e-05,
+      "loss": 0.8122,
+      "step": 988
+    },
+    {
+      "epoch": 0.1513389441469013,
+      "grad_norm": 2.5889920763864374,
+      "learning_rate": 1.9891122127414725e-05,
+      "loss": 0.7737,
+      "step": 989
+    },
+    {
+      "epoch": 0.1514919663351186,
+      "grad_norm": 2.8995325510612138,
+      "learning_rate": 1.9890757126149154e-05,
+      "loss": 0.9062,
+      "step": 990
+    },
+    {
+      "epoch": 0.15164498852333588,
+      "grad_norm": 2.7144127871621495,
+      "learning_rate": 1.9890391517453495e-05,
+      "loss": 0.8394,
+      "step": 991
+    },
+    {
+      "epoch": 0.1517980107115532,
+      "grad_norm": 2.9785649600008,
+      "learning_rate": 1.9890025301350202e-05,
+      "loss": 0.7811,
+      "step": 992
+    },
+    {
+      "epoch": 0.15195103289977047,
+      "grad_norm": 2.7814497882213547,
+      "learning_rate": 1.9889658477861764e-05,
+      "loss": 0.7973,
+      "step": 993
+    },
+    {
+      "epoch": 0.15210405508798777,
+      "grad_norm": 3.16891375953589,
+      "learning_rate": 1.9889291047010713e-05,
+      "loss": 0.8412,
+      "step": 994
+    },
+    {
+      "epoch": 0.15225707727620505,
+      "grad_norm": 2.5061691308106853,
+      "learning_rate": 1.9888923008819607e-05,
+      "loss": 0.7916,
+      "step": 995
+    },
+    {
+      "epoch": 0.15241009946442236,
+      "grad_norm": 2.863002070757968,
+      "learning_rate": 1.9888554363311058e-05,
+      "loss": 0.8654,
+      "step": 996
+    },
+    {
+      "epoch": 0.15256312165263963,
+      "grad_norm": 2.6516947662903343,
+      "learning_rate": 1.9888185110507702e-05,
+      "loss": 0.8693,
+      "step": 997
+    },
+    {
+      "epoch": 0.1527161438408569,
+      "grad_norm": 2.655655971232325,
+      "learning_rate": 1.988781525043221e-05,
+      "loss": 0.8319,
+      "step": 998
+    },
+    {
+      "epoch": 0.15286916602907422,
+      "grad_norm": 2.450582508006812,
+      "learning_rate": 1.9887444783107302e-05,
+      "loss": 0.8347,
+      "step": 999
+    },
+    {
+      "epoch": 0.1530221882172915,
+      "grad_norm": 2.670839990441853,
+      "learning_rate": 1.9887073708555736e-05,
+      "loss": 0.8599,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1531752104055088,
+      "grad_norm": 2.3507137132115736,
+      "learning_rate": 1.9886702026800295e-05,
+      "loss": 0.7578,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15332823259372608,
+      "grad_norm": 2.7099399060859444,
+      "learning_rate": 1.98863297378638e-05,
+      "loss": 0.8655,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15348125478194338,
+      "grad_norm": 2.5790296307424883,
+      "learning_rate": 1.9885956841769124e-05,
+      "loss": 0.8469,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15363427697016066,
+      "grad_norm": 2.265260853462987,
+      "learning_rate": 1.9885583338539162e-05,
+      "loss": 0.8579,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15378729915837797,
+      "grad_norm": 2.4073495126193305,
+      "learning_rate": 1.9885209228196854e-05,
+      "loss": 0.7811,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15394032134659524,
+      "grad_norm": 2.736856123838533,
+      "learning_rate": 1.988483451076518e-05,
+      "loss": 0.8012,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15409334353481255,
+      "grad_norm": 2.9918761922844714,
+      "learning_rate": 1.988445918626715e-05,
+      "loss": 0.9105,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15424636572302983,
+      "grad_norm": 2.770423326239471,
+      "learning_rate": 1.9884083254725808e-05,
+      "loss": 0.862,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15439938791124713,
+      "grad_norm": 2.596810493782701,
+      "learning_rate": 1.988370671616425e-05,
+      "loss": 0.8444,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1545524100994644,
+      "grad_norm": 2.570741057177668,
+      "learning_rate": 1.9883329570605594e-05,
+      "loss": 0.835,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15470543228768172,
+      "grad_norm": 2.4655694875759706,
+      "learning_rate": 1.9882951818073008e-05,
+      "loss": 0.9392,
+      "step": 1011
+    },
+    {
+      "epoch": 0.154858454475899,
+      "grad_norm": 2.881513696814611,
+      "learning_rate": 1.9882573458589687e-05,
+      "loss": 0.871,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1550114766641163,
+      "grad_norm": 3.1776782245839135,
+      "learning_rate": 1.988219449217887e-05,
+      "loss": 0.7849,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15516449885233358,
+      "grad_norm": 2.911172268228049,
+      "learning_rate": 1.988181491886383e-05,
+      "loss": 0.8837,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15531752104055088,
+      "grad_norm": 2.4686030359884743,
+      "learning_rate": 1.9881434738667877e-05,
+      "loss": 0.8141,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15547054322876816,
+      "grad_norm": 2.897765487407931,
+      "learning_rate": 1.988105395161436e-05,
+      "loss": 0.9853,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15562356541698547,
+      "grad_norm": 2.71964317721296,
+      "learning_rate": 1.9880672557726667e-05,
+      "loss": 0.7614,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15577658760520274,
+      "grad_norm": 2.9302612921371485,
+      "learning_rate": 1.9880290557028215e-05,
+      "loss": 0.8674,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15592960979342005,
+      "grad_norm": 2.7434432525327455,
+      "learning_rate": 1.9879907949542475e-05,
+      "loss": 0.8911,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15608263198163733,
+      "grad_norm": 3.1889761223051107,
+      "learning_rate": 1.987952473529293e-05,
+      "loss": 0.8377,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15623565416985463,
+      "grad_norm": 2.81375672067027,
+      "learning_rate": 1.987914091430313e-05,
+      "loss": 0.8919,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1563886763580719,
+      "grad_norm": 2.8180296209140563,
+      "learning_rate": 1.987875648659663e-05,
+      "loss": 0.9621,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15654169854628922,
+      "grad_norm": 3.2066379325461916,
+      "learning_rate": 1.9878371452197053e-05,
+      "loss": 0.957,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1566947207345065,
+      "grad_norm": 2.6277701752914204,
+      "learning_rate": 1.987798581112804e-05,
+      "loss": 0.8304,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1568477429227238,
+      "grad_norm": 2.6966984629555903,
+      "learning_rate": 1.9877599563413277e-05,
+      "loss": 0.7235,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15700076511094108,
+      "grad_norm": 2.73955117421016,
+      "learning_rate": 1.987721270907648e-05,
+      "loss": 0.7999,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15715378729915838,
+      "grad_norm": 3.725546674908976,
+      "learning_rate": 1.9876825248141413e-05,
+      "loss": 0.8339,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15730680948737566,
+      "grad_norm": 2.578167705290565,
+      "learning_rate": 1.9876437180631873e-05,
+      "loss": 0.7681,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15745983167559297,
+      "grad_norm": 2.7771069205470837,
+      "learning_rate": 1.987604850657168e-05,
+      "loss": 0.7977,
+      "step": 1029
+    },
+    {
+      "epoch": 0.15761285386381024,
+      "grad_norm": 2.8394656456803653,
+      "learning_rate": 1.987565922598472e-05,
+      "loss": 0.7927,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15776587605202755,
+      "grad_norm": 2.6510579492848088,
+      "learning_rate": 1.9875269338894888e-05,
+      "loss": 0.725,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15791889824024483,
+      "grad_norm": 2.837067274020622,
+      "learning_rate": 1.9874878845326137e-05,
+      "loss": 0.9011,
+      "step": 1032
+    },
+    {
+      "epoch": 0.15807192042846213,
+      "grad_norm": 2.8520975345419792,
+      "learning_rate": 1.9874487745302446e-05,
+      "loss": 0.8493,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1582249426166794,
+      "grad_norm": 3.1030948946703525,
+      "learning_rate": 1.9874096038847834e-05,
+      "loss": 1.0212,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15837796480489671,
+      "grad_norm": 2.9449253861471143,
+      "learning_rate": 1.9873703725986353e-05,
+      "loss": 0.9892,
+      "step": 1035
+    },
+    {
+      "epoch": 0.158530986993114,
+      "grad_norm": 2.6094643163438906,
+      "learning_rate": 1.9873310806742103e-05,
+      "loss": 0.8207,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1586840091813313,
+      "grad_norm": 2.8265307957870913,
+      "learning_rate": 1.9872917281139208e-05,
+      "loss": 0.79,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15883703136954858,
+      "grad_norm": 2.900708034355136,
+      "learning_rate": 1.9872523149201844e-05,
+      "loss": 0.9009,
+      "step": 1038
+    },
+    {
+      "epoch": 0.15899005355776588,
+      "grad_norm": 2.8149903877112554,
+      "learning_rate": 1.987212841095421e-05,
+      "loss": 0.9657,
+      "step": 1039
+    },
+    {
+      "epoch": 0.15914307574598316,
+      "grad_norm": 2.8216182032358788,
+      "learning_rate": 1.9871733066420548e-05,
+      "loss": 0.916,
+      "step": 1040
+    },
+    {
+      "epoch": 0.15929609793420046,
+      "grad_norm": 3.0351034690911467,
+      "learning_rate": 1.9871337115625146e-05,
+      "loss": 0.8505,
+      "step": 1041
+    },
+    {
+      "epoch": 0.15944912012241774,
+      "grad_norm": 2.4801658941328246,
+      "learning_rate": 1.9870940558592312e-05,
+      "loss": 0.8227,
+      "step": 1042
+    },
+    {
+      "epoch": 0.15960214231063505,
+      "grad_norm": 2.745895940283436,
+      "learning_rate": 1.98705433953464e-05,
+      "loss": 0.6975,
+      "step": 1043
+    },
+    {
+      "epoch": 0.15975516449885233,
+      "grad_norm": 2.8931887662338664,
+      "learning_rate": 1.9870145625911806e-05,
+      "loss": 0.8135,
+      "step": 1044
+    },
+    {
+      "epoch": 0.15990818668706963,
+      "grad_norm": 3.020183308577327,
+      "learning_rate": 1.9869747250312956e-05,
+      "loss": 0.9374,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1600612088752869,
+      "grad_norm": 2.91238434506869,
+      "learning_rate": 1.986934826857432e-05,
+      "loss": 0.7985,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16021423106350421,
+      "grad_norm": 2.8074374602676264,
+      "learning_rate": 1.9868948680720396e-05,
+      "loss": 0.762,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1603672532517215,
+      "grad_norm": 2.8655970184823967,
+      "learning_rate": 1.9868548486775726e-05,
+      "loss": 0.9527,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1605202754399388,
+      "grad_norm": 3.3510217577059924,
+      "learning_rate": 1.9868147686764887e-05,
+      "loss": 0.9492,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16067329762815608,
+      "grad_norm": 2.628760206614971,
+      "learning_rate": 1.9867746280712494e-05,
+      "loss": 0.7946,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16082631981637338,
+      "grad_norm": 2.70806765868512,
+      "learning_rate": 1.98673442686432e-05,
+      "loss": 0.8205,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16097934200459066,
+      "grad_norm": 2.4609824531759905,
+      "learning_rate": 1.986694165058169e-05,
+      "loss": 0.8578,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16113236419280796,
+      "grad_norm": 3.141891000216467,
+      "learning_rate": 1.9866538426552698e-05,
+      "loss": 0.9828,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16128538638102524,
+      "grad_norm": 3.061783048506876,
+      "learning_rate": 1.986613459658098e-05,
+      "loss": 0.8136,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16143840856924255,
+      "grad_norm": 2.6758943412455745,
+      "learning_rate": 1.9865730160691343e-05,
+      "loss": 0.8794,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16159143075745983,
+      "grad_norm": 2.7201604131125876,
+      "learning_rate": 1.986532511890862e-05,
+      "loss": 0.9024,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16174445294567713,
+      "grad_norm": 3.5299447043757337,
+      "learning_rate": 1.9864919471257685e-05,
+      "loss": 0.8843,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1618974751338944,
+      "grad_norm": 2.801548938363029,
+      "learning_rate": 1.9864513217763458e-05,
+      "loss": 0.8217,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16205049732211171,
+      "grad_norm": 2.5274711526613194,
+      "learning_rate": 1.9864106358450884e-05,
+      "loss": 0.8106,
+      "step": 1059
+    },
+    {
+      "epoch": 0.162203519510329,
+      "grad_norm": 2.7733624193991093,
+      "learning_rate": 1.9863698893344947e-05,
+      "loss": 0.7936,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1623565416985463,
+      "grad_norm": 2.5893301183647273,
+      "learning_rate": 1.9863290822470675e-05,
+      "loss": 0.792,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16250956388676358,
+      "grad_norm": 2.8266067518782934,
+      "learning_rate": 1.9862882145853127e-05,
+      "loss": 0.9559,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16266258607498088,
+      "grad_norm": 2.951855197304437,
+      "learning_rate": 1.986247286351741e-05,
+      "loss": 0.8523,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16281560826319816,
+      "grad_norm": 2.378517248975909,
+      "learning_rate": 1.9862062975488645e-05,
+      "loss": 0.8576,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16296863045141546,
+      "grad_norm": 2.795790531253483,
+      "learning_rate": 1.986165248179201e-05,
+      "loss": 0.8383,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16312165263963274,
+      "grad_norm": 2.618907367926037,
+      "learning_rate": 1.9861241382452724e-05,
+      "loss": 0.8413,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16327467482785005,
+      "grad_norm": 2.8494588393913336,
+      "learning_rate": 1.9860829677496024e-05,
+      "loss": 0.8867,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16342769701606732,
+      "grad_norm": 2.5805689918609773,
+      "learning_rate": 1.98604173669472e-05,
+      "loss": 0.8482,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16358071920428463,
+      "grad_norm": 2.711715960933929,
+      "learning_rate": 1.9860004450831566e-05,
+      "loss": 0.935,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1637337413925019,
+      "grad_norm": 2.7090531671259797,
+      "learning_rate": 1.9859590929174487e-05,
+      "loss": 0.7718,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1638867635807192,
+      "grad_norm": 2.6128893829365705,
+      "learning_rate": 1.985917680200136e-05,
+      "loss": 0.9498,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1640397857689365,
+      "grad_norm": 2.8396674469625998,
+      "learning_rate": 1.985876206933762e-05,
+      "loss": 0.9958,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1641928079571538,
+      "grad_norm": 2.8043208696375586,
+      "learning_rate": 1.9858346731208732e-05,
+      "loss": 0.8735,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16434583014537107,
+      "grad_norm": 2.684593665840827,
+      "learning_rate": 1.98579307876402e-05,
+      "loss": 0.8892,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16449885233358838,
+      "grad_norm": 2.596969620943833,
+      "learning_rate": 1.9857514238657576e-05,
+      "loss": 0.8741,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16465187452180566,
+      "grad_norm": 2.9338466904139917,
+      "learning_rate": 1.9857097084286445e-05,
+      "loss": 0.9696,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16480489671002296,
+      "grad_norm": 2.6540784341987087,
+      "learning_rate": 1.9856679324552415e-05,
+      "loss": 0.8988,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16495791889824024,
+      "grad_norm": 2.7019156426629034,
+      "learning_rate": 1.985626095948115e-05,
+      "loss": 0.8065,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16511094108645755,
+      "grad_norm": 2.615891811054439,
+      "learning_rate": 1.9855841989098343e-05,
+      "loss": 0.7563,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16526396327467482,
+      "grad_norm": 2.409266564450927,
+      "learning_rate": 1.985542241342972e-05,
+      "loss": 0.8721,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16541698546289213,
+      "grad_norm": 2.5645546399602397,
+      "learning_rate": 1.985500223250105e-05,
+      "loss": 0.7746,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1655700076511094,
+      "grad_norm": 2.606535862888543,
+      "learning_rate": 1.985458144633815e-05,
+      "loss": 0.9245,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1657230298393267,
+      "grad_norm": 3.102399821161681,
+      "learning_rate": 1.9854160054966845e-05,
+      "loss": 0.8806,
+      "step": 1083
+    },
+    {
+      "epoch": 0.165876052027544,
+      "grad_norm": 2.5601279913055994,
+      "learning_rate": 1.985373805841302e-05,
+      "loss": 0.8015,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1660290742157613,
+      "grad_norm": 2.61724976207117,
+      "learning_rate": 1.9853315456702594e-05,
+      "loss": 0.7555,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16618209640397857,
+      "grad_norm": 2.725383208135442,
+      "learning_rate": 1.9852892249861522e-05,
+      "loss": 0.8368,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16633511859219588,
+      "grad_norm": 2.765961715816743,
+      "learning_rate": 1.985246843791579e-05,
+      "loss": 0.8246,
+      "step": 1087
+    },
+    {
+      "epoch": 0.16648814078041316,
+      "grad_norm": 2.8936965697100505,
+      "learning_rate": 1.985204402089143e-05,
+      "loss": 0.8021,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16664116296863046,
+      "grad_norm": 2.713935819782728,
+      "learning_rate": 1.9851618998814503e-05,
+      "loss": 0.7641,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16679418515684774,
+      "grad_norm": 2.7965825996455957,
+      "learning_rate": 1.9851193371711113e-05,
+      "loss": 0.8254,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16694720734506505,
+      "grad_norm": 2.695814350476115,
+      "learning_rate": 1.9850767139607404e-05,
+      "loss": 0.8409,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16710022953328232,
+      "grad_norm": 2.8587205196165515,
+      "learning_rate": 1.9850340302529545e-05,
+      "loss": 0.8999,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16725325172149963,
+      "grad_norm": 2.616711026159911,
+      "learning_rate": 1.9849912860503753e-05,
+      "loss": 0.8085,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1674062739097169,
+      "grad_norm": 2.942385943591848,
+      "learning_rate": 1.984948481355628e-05,
+      "loss": 0.9305,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1675592960979342,
+      "grad_norm": 2.8217129762291595,
+      "learning_rate": 1.9849056161713415e-05,
+      "loss": 0.8427,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1677123182861515,
+      "grad_norm": 2.92693636303928,
+      "learning_rate": 1.984862690500148e-05,
+      "loss": 0.8569,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1678653404743688,
+      "grad_norm": 2.558373593775746,
+      "learning_rate": 1.984819704344684e-05,
+      "loss": 0.9141,
+      "step": 1097
+    },
+    {
+      "epoch": 0.16801836266258607,
+      "grad_norm": 2.6662816349876004,
+      "learning_rate": 1.984776657707589e-05,
+      "loss": 0.8173,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16817138485080338,
+      "grad_norm": 2.914507887638554,
+      "learning_rate": 1.984733550591507e-05,
+      "loss": 0.8787,
+      "step": 1099
+    },
+    {
+      "epoch": 0.16832440703902066,
+      "grad_norm": 2.909457834414278,
+      "learning_rate": 1.9846903829990857e-05,
+      "loss": 0.9803,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16847742922723796,
+      "grad_norm": 2.639150497743505,
+      "learning_rate": 1.9846471549329758e-05,
+      "loss": 0.8015,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16863045141545524,
+      "grad_norm": 2.594125634192865,
+      "learning_rate": 1.984603866395832e-05,
+      "loss": 0.7903,
+      "step": 1102
+    },
+    {
+      "epoch": 0.16878347360367255,
+      "grad_norm": 2.369582246149647,
+      "learning_rate": 1.984560517390313e-05,
+      "loss": 0.7992,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16893649579188982,
+      "grad_norm": 2.555457979137369,
+      "learning_rate": 1.984517107919081e-05,
+      "loss": 0.8961,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16908951798010713,
+      "grad_norm": 2.615173559437897,
+      "learning_rate": 1.984473637984802e-05,
+      "loss": 0.9241,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1692425401683244,
+      "grad_norm": 2.6096821694922188,
+      "learning_rate": 1.9844301075901455e-05,
+      "loss": 0.8641,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1693955623565417,
+      "grad_norm": 2.559217882335341,
+      "learning_rate": 1.984386516737785e-05,
+      "loss": 0.927,
+      "step": 1107
+    },
+    {
+      "epoch": 0.169548584544759,
+      "grad_norm": 2.6360804443043073,
+      "learning_rate": 1.9843428654303977e-05,
+      "loss": 0.8263,
+      "step": 1108
+    },
+    {
+      "epoch": 0.16970160673297627,
+      "grad_norm": 2.7158983215786465,
+      "learning_rate": 1.984299153670664e-05,
+      "loss": 0.813,
+      "step": 1109
+    },
+    {
+      "epoch": 0.16985462892119357,
+      "grad_norm": 2.850752676076143,
+      "learning_rate": 1.984255381461269e-05,
+      "loss": 0.7828,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17000765110941085,
+      "grad_norm": 2.849267604288,
+      "learning_rate": 1.9842115488049006e-05,
+      "loss": 0.8832,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17016067329762816,
+      "grad_norm": 2.8725800405214286,
+      "learning_rate": 1.9841676557042505e-05,
+      "loss": 0.8527,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17031369548584543,
+      "grad_norm": 2.7923030948061034,
+      "learning_rate": 1.9841237021620147e-05,
+      "loss": 0.9163,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17046671767406274,
+      "grad_norm": 2.6630219303474822,
+      "learning_rate": 1.9840796881808922e-05,
+      "loss": 0.774,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17061973986228002,
+      "grad_norm": 2.798714223571827,
+      "learning_rate": 1.9840356137635865e-05,
+      "loss": 0.8822,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17077276205049732,
+      "grad_norm": 2.8105521810924237,
+      "learning_rate": 1.983991478912804e-05,
+      "loss": 0.8292,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1709257842387146,
+      "grad_norm": 2.9679950044260988,
+      "learning_rate": 1.9839472836312558e-05,
+      "loss": 0.8392,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1710788064269319,
+      "grad_norm": 2.5546239972059923,
+      "learning_rate": 1.9839030279216557e-05,
+      "loss": 0.8779,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17123182861514918,
+      "grad_norm": 2.6350772004841367,
+      "learning_rate": 1.983858711786721e-05,
+      "loss": 0.9282,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1713848508033665,
+      "grad_norm": 2.40493852138416,
+      "learning_rate": 1.9838143352291747e-05,
+      "loss": 0.7634,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17153787299158377,
+      "grad_norm": 2.904701675083358,
+      "learning_rate": 1.9837698982517408e-05,
+      "loss": 0.8338,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17169089517980107,
+      "grad_norm": 2.6070853971468466,
+      "learning_rate": 1.983725400857149e-05,
+      "loss": 0.9001,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17184391736801835,
+      "grad_norm": 2.6858795073590023,
+      "learning_rate": 1.983680843048132e-05,
+      "loss": 0.9843,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17199693955623566,
+      "grad_norm": 2.44821873068937,
+      "learning_rate": 1.9836362248274262e-05,
+      "loss": 0.8795,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17214996174445293,
+      "grad_norm": 2.699038981515868,
+      "learning_rate": 1.983591546197772e-05,
+      "loss": 0.8473,
+      "step": 1125
+    },
+    {
+      "epoch": 0.17230298393267024,
+      "grad_norm": 2.4135541032139654,
+      "learning_rate": 1.9835468071619132e-05,
+      "loss": 0.8204,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17245600612088752,
+      "grad_norm": 2.545206003492443,
+      "learning_rate": 1.983502007722597e-05,
+      "loss": 0.8033,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17260902830910482,
+      "grad_norm": 2.5812248093359975,
+      "learning_rate": 1.9834571478825752e-05,
+      "loss": 0.7242,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1727620504973221,
+      "grad_norm": 3.048017837391477,
+      "learning_rate": 1.983412227644603e-05,
+      "loss": 0.9028,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1729150726855394,
+      "grad_norm": 3.0440038836635597,
+      "learning_rate": 1.983367247011438e-05,
+      "loss": 0.7642,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17306809487375668,
+      "grad_norm": 2.4646195094640513,
+      "learning_rate": 1.9833222059858438e-05,
+      "loss": 0.737,
+      "step": 1131
+    },
+    {
+      "epoch": 0.173221117061974,
+      "grad_norm": 2.9135955129234916,
+      "learning_rate": 1.9832771045705862e-05,
+      "loss": 0.8295,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17337413925019127,
+      "grad_norm": 2.517525658493596,
+      "learning_rate": 1.9832319427684352e-05,
+      "loss": 0.9521,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17352716143840857,
+      "grad_norm": 2.563415773261328,
+      "learning_rate": 1.983186720582164e-05,
+      "loss": 0.6452,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17368018362662585,
+      "grad_norm": 2.551066647247192,
+      "learning_rate": 1.98314143801455e-05,
+      "loss": 0.8596,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17383320581484316,
+      "grad_norm": 2.309987774636032,
+      "learning_rate": 1.983096095068374e-05,
+      "loss": 0.7836,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17398622800306043,
+      "grad_norm": 2.968328488790905,
+      "learning_rate": 1.983050691746421e-05,
+      "loss": 0.967,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17413925019127774,
+      "grad_norm": 2.70579171197833,
+      "learning_rate": 1.9830052280514795e-05,
+      "loss": 0.9944,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17429227237949502,
+      "grad_norm": 2.461411813599091,
+      "learning_rate": 1.9829597039863416e-05,
+      "loss": 0.8744,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17444529456771232,
+      "grad_norm": 2.907849288421829,
+      "learning_rate": 1.9829141195538025e-05,
+      "loss": 0.8599,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1745983167559296,
+      "grad_norm": 2.8382037466956875,
+      "learning_rate": 1.9828684747566625e-05,
+      "loss": 0.8959,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1747513389441469,
+      "grad_norm": 2.748550273869275,
+      "learning_rate": 1.982822769597724e-05,
+      "loss": 0.8289,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17490436113236418,
+      "grad_norm": 2.8067275425318052,
+      "learning_rate": 1.9827770040797946e-05,
+      "loss": 0.9106,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1750573833205815,
+      "grad_norm": 2.846345213122419,
+      "learning_rate": 1.982731178205685e-05,
+      "loss": 0.7729,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17521040550879877,
+      "grad_norm": 3.2293890063364796,
+      "learning_rate": 1.982685291978209e-05,
+      "loss": 0.9251,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17536342769701607,
+      "grad_norm": 2.858144968984894,
+      "learning_rate": 1.9826393454001848e-05,
+      "loss": 0.7992,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17551644988523335,
+      "grad_norm": 2.8071648371182345,
+      "learning_rate": 1.9825933384744343e-05,
+      "loss": 0.8398,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17566947207345066,
+      "grad_norm": 2.875344520665991,
+      "learning_rate": 1.9825472712037836e-05,
+      "loss": 0.8488,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17582249426166793,
+      "grad_norm": 2.6795215539666732,
+      "learning_rate": 1.9825011435910606e-05,
+      "loss": 0.8287,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17597551644988524,
+      "grad_norm": 2.6519672698029524,
+      "learning_rate": 1.982454955639099e-05,
+      "loss": 0.9665,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17612853863810252,
+      "grad_norm": 2.9398416836616263,
+      "learning_rate": 1.982408707350735e-05,
+      "loss": 0.9661,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17628156082631982,
+      "grad_norm": 2.5372992843733444,
+      "learning_rate": 1.9823623987288096e-05,
+      "loss": 0.7973,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1764345830145371,
+      "grad_norm": 2.5779136243237804,
+      "learning_rate": 1.9823160297761657e-05,
+      "loss": 0.8823,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1765876052027544,
+      "grad_norm": 2.6819788898485943,
+      "learning_rate": 1.9822696004956516e-05,
+      "loss": 0.757,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17674062739097168,
+      "grad_norm": 2.7476885475300827,
+      "learning_rate": 1.982223110890119e-05,
+      "loss": 0.9533,
+      "step": 1155
+    },
+    {
+      "epoch": 0.176893649579189,
+      "grad_norm": 2.5992700129932493,
+      "learning_rate": 1.9821765609624223e-05,
+      "loss": 0.7441,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17704667176740627,
+      "grad_norm": 2.8631617649345626,
+      "learning_rate": 1.982129950715421e-05,
+      "loss": 0.8553,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17719969395562357,
+      "grad_norm": 2.8390171782673996,
+      "learning_rate": 1.9820832801519772e-05,
+      "loss": 0.8184,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17735271614384085,
+      "grad_norm": 2.712066208521167,
+      "learning_rate": 1.9820365492749577e-05,
+      "loss": 0.8139,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17750573833205815,
+      "grad_norm": 2.613425607348815,
+      "learning_rate": 1.9819897580872313e-05,
+      "loss": 0.8787,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17765876052027543,
+      "grad_norm": 2.7109084220530533,
+      "learning_rate": 1.9819429065916725e-05,
+      "loss": 0.9128,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17781178270849274,
+      "grad_norm": 2.9223737813686057,
+      "learning_rate": 1.9818959947911585e-05,
+      "loss": 0.8176,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17796480489671002,
+      "grad_norm": 2.778954100850369,
+      "learning_rate": 1.9818490226885703e-05,
+      "loss": 0.8663,
+      "step": 1163
+    },
+    {
+      "epoch": 0.17811782708492732,
+      "grad_norm": 2.6877635305323593,
+      "learning_rate": 1.9818019902867924e-05,
+      "loss": 0.9036,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1782708492731446,
+      "grad_norm": 3.094006409195956,
+      "learning_rate": 1.9817548975887136e-05,
+      "loss": 0.822,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1784238714613619,
+      "grad_norm": 2.487275168235667,
+      "learning_rate": 1.981707744597226e-05,
+      "loss": 0.8939,
+      "step": 1166
+    },
+    {
+      "epoch": 0.17857689364957918,
+      "grad_norm": 2.328184674285479,
+      "learning_rate": 1.981660531315225e-05,
+      "loss": 0.8309,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1787299158377965,
+      "grad_norm": 2.3136962546636464,
+      "learning_rate": 1.9816132577456105e-05,
+      "loss": 0.7128,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17888293802601377,
+      "grad_norm": 2.644074314289146,
+      "learning_rate": 1.981565923891286e-05,
+      "loss": 0.8859,
+      "step": 1169
+    },
+    {
+      "epoch": 0.17903596021423107,
+      "grad_norm": 2.5503162195908544,
+      "learning_rate": 1.9815185297551584e-05,
+      "loss": 0.8524,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17918898240244835,
+      "grad_norm": 2.702749830814934,
+      "learning_rate": 1.981471075340138e-05,
+      "loss": 0.8143,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17934200459066565,
+      "grad_norm": 2.6253033708655713,
+      "learning_rate": 1.981423560649139e-05,
+      "loss": 0.7908,
+      "step": 1172
+    },
+    {
+      "epoch": 0.17949502677888293,
+      "grad_norm": 2.5121491386593924,
+      "learning_rate": 1.98137598568508e-05,
+      "loss": 0.8573,
+      "step": 1173
+    },
+    {
+      "epoch": 0.17964804896710024,
+      "grad_norm": 3.5925654232893134,
+      "learning_rate": 1.9813283504508828e-05,
+      "loss": 0.8668,
+      "step": 1174
+    },
+    {
+      "epoch": 0.17980107115531752,
+      "grad_norm": 3.0099094303601195,
+      "learning_rate": 1.9812806549494723e-05,
+      "loss": 0.7845,
+      "step": 1175
+    },
+    {
+      "epoch": 0.17995409334353482,
+      "grad_norm": 7.263338996571074,
+      "learning_rate": 1.981232899183778e-05,
+      "loss": 0.9048,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1801071155317521,
+      "grad_norm": 4.591653363432459,
+      "learning_rate": 1.9811850831567327e-05,
+      "loss": 0.7974,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1802601377199694,
+      "grad_norm": 3.825510779009282,
+      "learning_rate": 1.9811372068712734e-05,
+      "loss": 0.8295,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18041315990818668,
+      "grad_norm": 2.9921320333176706,
+      "learning_rate": 1.9810892703303398e-05,
+      "loss": 0.8919,
+      "step": 1179
+    },
+    {
+      "epoch": 0.180566182096404,
+      "grad_norm": 2.7588519917332617,
+      "learning_rate": 1.981041273536876e-05,
+      "loss": 0.7638,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18071920428462127,
+      "grad_norm": 3.083353753553559,
+      "learning_rate": 1.9809932164938297e-05,
+      "loss": 0.942,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18087222647283857,
+      "grad_norm": 2.728694292712886,
+      "learning_rate": 1.9809450992041522e-05,
+      "loss": 0.9411,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18102524866105585,
+      "grad_norm": 2.7319529906411835,
+      "learning_rate": 1.980896921670799e-05,
+      "loss": 0.874,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18117827084927315,
+      "grad_norm": 2.7865710261296917,
+      "learning_rate": 1.9808486838967286e-05,
+      "loss": 0.9014,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18133129303749043,
+      "grad_norm": 2.5143812899440268,
+      "learning_rate": 1.9808003858849032e-05,
+      "loss": 0.8666,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18148431522570774,
+      "grad_norm": 2.4430869214626516,
+      "learning_rate": 1.980752027638289e-05,
+      "loss": 0.7535,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18163733741392502,
+      "grad_norm": 2.2877893367824726,
+      "learning_rate": 1.9807036091598563e-05,
+      "loss": 0.6914,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18179035960214232,
+      "grad_norm": 3.0901473909433124,
+      "learning_rate": 1.9806551304525784e-05,
+      "loss": 0.795,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1819433817903596,
+      "grad_norm": 2.7091237157952612,
+      "learning_rate": 1.9806065915194326e-05,
+      "loss": 0.8558,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1820964039785769,
+      "grad_norm": 2.7289003459871743,
+      "learning_rate": 1.9805579923633997e-05,
+      "loss": 0.9272,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18224942616679418,
+      "grad_norm": 2.9092943582139377,
+      "learning_rate": 1.980509332987465e-05,
+      "loss": 0.7949,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1824024483550115,
+      "grad_norm": 2.8490336831731002,
+      "learning_rate": 1.980460613394616e-05,
+      "loss": 0.9731,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18255547054322876,
+      "grad_norm": 2.5321760664034225,
+      "learning_rate": 1.9804118335878452e-05,
+      "loss": 0.8599,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18270849273144607,
+      "grad_norm": 2.582180096249362,
+      "learning_rate": 1.980362993570148e-05,
+      "loss": 0.7725,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18286151491966335,
+      "grad_norm": 3.050407598760349,
+      "learning_rate": 1.9803140933445246e-05,
+      "loss": 0.9098,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18301453710788065,
+      "grad_norm": 2.9433028267353025,
+      "learning_rate": 1.980265132913978e-05,
+      "loss": 0.9321,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18316755929609793,
+      "grad_norm": 2.439342164588578,
+      "learning_rate": 1.980216112281514e-05,
+      "loss": 0.745,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18332058148431524,
+      "grad_norm": 2.8658445163454886,
+      "learning_rate": 1.9801670314501445e-05,
+      "loss": 0.9419,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18347360367253251,
+      "grad_norm": 2.3996411264820825,
+      "learning_rate": 1.980117890422883e-05,
+      "loss": 0.7613,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18362662586074982,
+      "grad_norm": 2.634152697556565,
+      "learning_rate": 1.9800686892027476e-05,
+      "loss": 0.838,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1837796480489671,
+      "grad_norm": 2.507128419724863,
+      "learning_rate": 1.9800194277927598e-05,
+      "loss": 0.7374,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1839326702371844,
+      "grad_norm": 2.5150477885289453,
+      "learning_rate": 1.9799701061959453e-05,
+      "loss": 0.8103,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18408569242540168,
+      "grad_norm": 2.7743946382832214,
+      "learning_rate": 1.9799207244153328e-05,
+      "loss": 0.7684,
+      "step": 1203
+    },
+    {
+      "epoch": 0.184238714613619,
+      "grad_norm": 2.7916364471729262,
+      "learning_rate": 1.979871282453955e-05,
+      "loss": 0.9085,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18439173680183626,
+      "grad_norm": 2.6761395292421652,
+      "learning_rate": 1.9798217803148488e-05,
+      "loss": 0.8737,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18454475899005357,
+      "grad_norm": 2.7327345821631353,
+      "learning_rate": 1.9797722180010536e-05,
+      "loss": 0.8907,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18469778117827085,
+      "grad_norm": 2.926716209435097,
+      "learning_rate": 1.979722595515614e-05,
+      "loss": 0.9471,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18485080336648815,
+      "grad_norm": 2.4288423001399724,
+      "learning_rate": 1.979672912861577e-05,
+      "loss": 0.8499,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18500382555470543,
+      "grad_norm": 2.8360903216784865,
+      "learning_rate": 1.9796231700419937e-05,
+      "loss": 0.9216,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18515684774292274,
+      "grad_norm": 2.9513922158366466,
+      "learning_rate": 1.9795733670599194e-05,
+      "loss": 0.8389,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18530986993114001,
+      "grad_norm": 2.480276836129265,
+      "learning_rate": 1.9795235039184122e-05,
+      "loss": 0.8037,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18546289211935732,
+      "grad_norm": 2.5886986020209632,
+      "learning_rate": 1.979473580620535e-05,
+      "loss": 0.837,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1856159143075746,
+      "grad_norm": 2.673110598202028,
+      "learning_rate": 1.9794235971693537e-05,
+      "loss": 0.8506,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1857689364957919,
+      "grad_norm": 2.8120113685946833,
+      "learning_rate": 1.9793735535679373e-05,
+      "loss": 0.8873,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18592195868400918,
+      "grad_norm": 2.4841528470474525,
+      "learning_rate": 1.97932344981936e-05,
+      "loss": 0.8227,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1860749808722265,
+      "grad_norm": 2.6262957390756423,
+      "learning_rate": 1.9792732859266985e-05,
+      "loss": 0.831,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18622800306044376,
+      "grad_norm": 2.7867995827118905,
+      "learning_rate": 1.979223061893033e-05,
+      "loss": 0.8887,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18638102524866107,
+      "grad_norm": 2.4630014832920133,
+      "learning_rate": 1.9791727777214494e-05,
+      "loss": 0.7623,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18653404743687835,
+      "grad_norm": 2.8608171398956714,
+      "learning_rate": 1.9791224334150344e-05,
+      "loss": 0.8973,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18668706962509563,
+      "grad_norm": 2.625067202467613,
+      "learning_rate": 1.9790720289768807e-05,
+      "loss": 0.88,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18684009181331293,
+      "grad_norm": 2.6002917464889475,
+      "learning_rate": 1.9790215644100834e-05,
+      "loss": 0.8389,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1869931140015302,
+      "grad_norm": 2.9459681926716046,
+      "learning_rate": 1.978971039717742e-05,
+      "loss": 0.8296,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18714613618974751,
+      "grad_norm": 2.713734464464466,
+      "learning_rate": 1.978920454902959e-05,
+      "loss": 0.8426,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1872991583779648,
+      "grad_norm": 2.7203914331532415,
+      "learning_rate": 1.9788698099688416e-05,
+      "loss": 0.9216,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1874521805661821,
+      "grad_norm": 2.410057875984283,
+      "learning_rate": 1.9788191049184998e-05,
+      "loss": 0.8333,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18760520275439937,
+      "grad_norm": 2.7449351455709827,
+      "learning_rate": 1.9787683397550476e-05,
+      "loss": 0.8998,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18775822494261668,
+      "grad_norm": 2.852865788819563,
+      "learning_rate": 1.9787175144816024e-05,
+      "loss": 0.7877,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18791124713083396,
+      "grad_norm": 2.50764849784216,
+      "learning_rate": 1.9786666291012865e-05,
+      "loss": 0.8286,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18806426931905126,
+      "grad_norm": 2.572027875501263,
+      "learning_rate": 1.9786156836172237e-05,
+      "loss": 0.8445,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18821729150726854,
+      "grad_norm": 2.484590343748145,
+      "learning_rate": 1.9785646780325435e-05,
+      "loss": 0.8945,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18837031369548585,
+      "grad_norm": 2.5100868462046817,
+      "learning_rate": 1.9785136123503787e-05,
+      "loss": 0.8939,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18852333588370312,
+      "grad_norm": 2.455177735428078,
+      "learning_rate": 1.9784624865738643e-05,
+      "loss": 0.7709,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18867635807192043,
+      "grad_norm": 2.4916566077574567,
+      "learning_rate": 1.9784113007061414e-05,
+      "loss": 0.8872,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1888293802601377,
+      "grad_norm": 2.430966096318744,
+      "learning_rate": 1.9783600547503528e-05,
+      "loss": 0.9165,
+      "step": 1234
+    },
+    {
+      "epoch": 0.188982402448355,
+      "grad_norm": 2.4779511899944264,
+      "learning_rate": 1.978308748709646e-05,
+      "loss": 0.7428,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1891354246365723,
+      "grad_norm": 2.7661213468146175,
+      "learning_rate": 1.9782573825871715e-05,
+      "loss": 0.7765,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1892884468247896,
+      "grad_norm": 2.5469813018884793,
+      "learning_rate": 1.9782059563860844e-05,
+      "loss": 0.7575,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18944146901300687,
+      "grad_norm": 2.6609455008160396,
+      "learning_rate": 1.9781544701095426e-05,
+      "loss": 0.9117,
+      "step": 1238
+    },
+    {
+      "epoch": 0.18959449120122418,
+      "grad_norm": 2.8769593063146948,
+      "learning_rate": 1.9781029237607082e-05,
+      "loss": 0.8893,
+      "step": 1239
+    },
+    {
+      "epoch": 0.18974751338944146,
+      "grad_norm": 2.910752462671504,
+      "learning_rate": 1.9780513173427472e-05,
+      "loss": 0.862,
+      "step": 1240
+    },
+    {
+      "epoch": 0.18990053557765876,
+      "grad_norm": 2.8258652202291135,
+      "learning_rate": 1.977999650858828e-05,
+      "loss": 0.9601,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19005355776587604,
+      "grad_norm": 2.3361137453572756,
+      "learning_rate": 1.977947924312125e-05,
+      "loss": 0.6773,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19020657995409335,
+      "grad_norm": 2.732147684563504,
+      "learning_rate": 1.9778961377058138e-05,
+      "loss": 0.9071,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19035960214231062,
+      "grad_norm": 2.6873683102739014,
+      "learning_rate": 1.9778442910430753e-05,
+      "loss": 0.7924,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19051262433052793,
+      "grad_norm": 2.8385621841540196,
+      "learning_rate": 1.9777923843270937e-05,
+      "loss": 0.9363,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1906656465187452,
+      "grad_norm": 2.6345781764376106,
+      "learning_rate": 1.9777404175610563e-05,
+      "loss": 0.8486,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1908186687069625,
+      "grad_norm": 2.7282412099331967,
+      "learning_rate": 1.9776883907481556e-05,
+      "loss": 0.8968,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1909716908951798,
+      "grad_norm": 3.1198583665695385,
+      "learning_rate": 1.9776363038915853e-05,
+      "loss": 0.8591,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1911247130833971,
+      "grad_norm": 2.9650963500849605,
+      "learning_rate": 1.9775841569945455e-05,
+      "loss": 0.895,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19127773527161437,
+      "grad_norm": 3.005956067515491,
+      "learning_rate": 1.9775319500602383e-05,
+      "loss": 0.8774,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19143075745983168,
+      "grad_norm": 2.785558603450888,
+      "learning_rate": 1.9774796830918696e-05,
+      "loss": 0.7999,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19158377964804896,
+      "grad_norm": 2.754290803774593,
+      "learning_rate": 1.97742735609265e-05,
+      "loss": 0.8619,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19173680183626626,
+      "grad_norm": 2.59123133889139,
+      "learning_rate": 1.9773749690657923e-05,
+      "loss": 0.8181,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19188982402448354,
+      "grad_norm": 2.423499848751921,
+      "learning_rate": 1.9773225220145144e-05,
+      "loss": 0.858,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19204284621270085,
+      "grad_norm": 2.651915413290144,
+      "learning_rate": 1.9772700149420374e-05,
+      "loss": 0.9273,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19219586840091812,
+      "grad_norm": 2.4797552410990154,
+      "learning_rate": 1.9772174478515853e-05,
+      "loss": 0.8235,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19234889058913543,
+      "grad_norm": 2.5741982732357824,
+      "learning_rate": 1.977164820746387e-05,
+      "loss": 0.77,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1925019127773527,
+      "grad_norm": 2.610715685337163,
+      "learning_rate": 1.9771121336296744e-05,
+      "loss": 0.8574,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19265493496557,
+      "grad_norm": 2.4847035418283254,
+      "learning_rate": 1.9770593865046832e-05,
+      "loss": 0.8437,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1928079571537873,
+      "grad_norm": 2.4929585213596672,
+      "learning_rate": 1.9770065793746528e-05,
+      "loss": 0.8486,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1929609793420046,
+      "grad_norm": 3.044420651513117,
+      "learning_rate": 1.9769537122428264e-05,
+      "loss": 0.9013,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19311400153022187,
+      "grad_norm": 2.6620986896543237,
+      "learning_rate": 1.9769007851124505e-05,
+      "loss": 0.9234,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19326702371843918,
+      "grad_norm": 2.8157426489061135,
+      "learning_rate": 1.9768477979867757e-05,
+      "loss": 0.9035,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19342004590665646,
+      "grad_norm": 2.529021689411045,
+      "learning_rate": 1.9767947508690562e-05,
+      "loss": 0.8663,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19357306809487376,
+      "grad_norm": 2.5843925171763535,
+      "learning_rate": 1.97674164376255e-05,
+      "loss": 0.8378,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19372609028309104,
+      "grad_norm": 3.0335901445285525,
+      "learning_rate": 1.9766884766705182e-05,
+      "loss": 0.8908,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19387911247130835,
+      "grad_norm": 3.145047505720264,
+      "learning_rate": 1.9766352495962263e-05,
+      "loss": 1.0083,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19403213465952562,
+      "grad_norm": 2.4039128378461543,
+      "learning_rate": 1.976581962542943e-05,
+      "loss": 0.7459,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19418515684774293,
+      "grad_norm": 2.2548402455018133,
+      "learning_rate": 1.976528615513941e-05,
+      "loss": 0.8713,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1943381790359602,
+      "grad_norm": 2.5574308945342876,
+      "learning_rate": 1.9764752085124968e-05,
+      "loss": 0.8306,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1944912012241775,
+      "grad_norm": 2.73069294916681,
+      "learning_rate": 1.97642174154189e-05,
+      "loss": 0.8369,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1946442234123948,
+      "grad_norm": 2.48980869464948,
+      "learning_rate": 1.9763682146054036e-05,
+      "loss": 0.7243,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1947972456006121,
+      "grad_norm": 2.736038468896169,
+      "learning_rate": 1.9763146277063264e-05,
+      "loss": 0.8022,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19495026778882937,
+      "grad_norm": 2.698636507396173,
+      "learning_rate": 1.976260980847948e-05,
+      "loss": 0.8665,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19510328997704668,
+      "grad_norm": 2.728859710561004,
+      "learning_rate": 1.976207274033564e-05,
+      "loss": 0.7747,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19525631216526396,
+      "grad_norm": 2.4838979015966607,
+      "learning_rate": 1.976153507266472e-05,
+      "loss": 0.8504,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19540933435348126,
+      "grad_norm": 2.659727034754188,
+      "learning_rate": 1.976099680549975e-05,
+      "loss": 0.9027,
+      "step": 1277
+    },
+    {
+      "epoch": 0.19556235654169854,
+      "grad_norm": 2.7146676990146257,
+      "learning_rate": 1.9760457938873773e-05,
+      "loss": 0.8086,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19571537872991585,
+      "grad_norm": 2.476438101864568,
+      "learning_rate": 1.9759918472819896e-05,
+      "loss": 0.8024,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19586840091813312,
+      "grad_norm": 2.494365497726393,
+      "learning_rate": 1.9759378407371242e-05,
+      "loss": 0.8212,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19602142310635043,
+      "grad_norm": 2.7005260217584444,
+      "learning_rate": 1.9758837742560984e-05,
+      "loss": 0.8976,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1961744452945677,
+      "grad_norm": 2.590370938150912,
+      "learning_rate": 1.975829647842232e-05,
+      "loss": 0.7617,
+      "step": 1282
+    },
+    {
+      "epoch": 0.196327467482785,
+      "grad_norm": 2.5882464926232824,
+      "learning_rate": 1.9757754614988495e-05,
+      "loss": 0.7785,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1964804896710023,
+      "grad_norm": 2.502547989366547,
+      "learning_rate": 1.975721215229279e-05,
+      "loss": 0.9019,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1966335118592196,
+      "grad_norm": 2.5390514701365268,
+      "learning_rate": 1.9756669090368514e-05,
+      "loss": 0.821,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19678653404743687,
+      "grad_norm": 2.639839580822911,
+      "learning_rate": 1.975612542924902e-05,
+      "loss": 0.8262,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19693955623565418,
+      "grad_norm": 2.4825562925435443,
+      "learning_rate": 1.97555811689677e-05,
+      "loss": 0.79,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19709257842387146,
+      "grad_norm": 2.7116933120169704,
+      "learning_rate": 1.9755036309557974e-05,
+      "loss": 0.9368,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19724560061208876,
+      "grad_norm": 2.663602543382191,
+      "learning_rate": 1.9754490851053306e-05,
+      "loss": 0.8395,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19739862280030604,
+      "grad_norm": 2.642048603050421,
+      "learning_rate": 1.9753944793487195e-05,
+      "loss": 0.9173,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19755164498852334,
+      "grad_norm": 2.619082257934699,
+      "learning_rate": 1.975339813689318e-05,
+      "loss": 0.8949,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19770466717674062,
+      "grad_norm": 2.469477303471826,
+      "learning_rate": 1.9752850881304827e-05,
+      "loss": 0.8011,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19785768936495793,
+      "grad_norm": 2.332170742162521,
+      "learning_rate": 1.9752303026755747e-05,
+      "loss": 0.7746,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1980107115531752,
+      "grad_norm": 2.465506481746153,
+      "learning_rate": 1.975175457327959e-05,
+      "loss": 0.7254,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1981637337413925,
+      "grad_norm": 2.504242954456863,
+      "learning_rate": 1.975120552091003e-05,
+      "loss": 0.8851,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1983167559296098,
+      "grad_norm": 2.7474656960473345,
+      "learning_rate": 1.97506558696808e-05,
+      "loss": 0.8854,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1984697781178271,
+      "grad_norm": 2.55032537694633,
+      "learning_rate": 1.9750105619625644e-05,
+      "loss": 0.8987,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19862280030604437,
+      "grad_norm": 3.147485983482909,
+      "learning_rate": 1.9749554770778358e-05,
+      "loss": 0.8879,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19877582249426168,
+      "grad_norm": 2.774617380445243,
+      "learning_rate": 1.974900332317277e-05,
+      "loss": 0.9347,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19892884468247896,
+      "grad_norm": 2.5337569604258623,
+      "learning_rate": 1.974845127684276e-05,
+      "loss": 0.7679,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19908186687069626,
+      "grad_norm": 2.432147574123346,
+      "learning_rate": 1.9747898631822213e-05,
+      "loss": 0.8188,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19923488905891354,
+      "grad_norm": 2.6080426910470775,
+      "learning_rate": 1.9747345388145082e-05,
+      "loss": 0.8023,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19938791124713084,
+      "grad_norm": 2.4662419988741497,
+      "learning_rate": 1.9746791545845335e-05,
+      "loss": 0.7576,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19954093343534812,
+      "grad_norm": 2.6788843876154353,
+      "learning_rate": 1.974623710495699e-05,
+      "loss": 0.807,
+      "step": 1304
+    },
+    {
+      "epoch": 0.19969395562356543,
+      "grad_norm": 2.575256113628228,
+      "learning_rate": 1.9745682065514096e-05,
+      "loss": 0.8279,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1998469778117827,
+      "grad_norm": 2.714741610802897,
+      "learning_rate": 1.9745126427550742e-05,
+      "loss": 0.8878,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8040762303011126,
+      "learning_rate": 1.9744570191101053e-05,
+      "loss": 0.8519,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2001530221882173,
+      "grad_norm": 2.824683717072596,
+      "learning_rate": 1.9744013356199186e-05,
+      "loss": 0.8524,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2003060443764346,
+      "grad_norm": 2.6712623088554306,
+      "learning_rate": 1.974345592287934e-05,
+      "loss": 0.9416,
+      "step": 1309
+    },
+    {
+      "epoch": 0.20045906656465187,
+      "grad_norm": 2.6566874873753794,
+      "learning_rate": 1.9742897891175746e-05,
+      "loss": 0.8073,
+      "step": 1310
+    },
+    {
+      "epoch": 0.20061208875286918,
+      "grad_norm": 2.8105095398771174,
+      "learning_rate": 1.9742339261122682e-05,
+      "loss": 0.8246,
+      "step": 1311
+    },
+    {
+      "epoch": 0.20076511094108646,
+      "grad_norm": 2.4516251972652334,
+      "learning_rate": 1.9741780032754452e-05,
+      "loss": 0.8449,
+      "step": 1312
+    },
+    {
+      "epoch": 0.20091813312930376,
+      "grad_norm": 2.7270021374282036,
+      "learning_rate": 1.9741220206105398e-05,
+      "loss": 0.9152,
+      "step": 1313
+    },
+    {
+      "epoch": 0.20107115531752104,
+      "grad_norm": 2.6346059254428815,
+      "learning_rate": 1.9740659781209905e-05,
+      "loss": 0.9087,
+      "step": 1314
+    },
+    {
+      "epoch": 0.20122417750573834,
+      "grad_norm": 2.6401950334296433,
+      "learning_rate": 1.9740098758102388e-05,
+      "loss": 0.8442,
+      "step": 1315
+    },
+    {
+      "epoch": 0.20137719969395562,
+      "grad_norm": 2.590632699986551,
+      "learning_rate": 1.9739537136817303e-05,
+      "loss": 0.7427,
+      "step": 1316
+    },
+    {
+      "epoch": 0.20153022188217293,
+      "grad_norm": 2.515449790487725,
+      "learning_rate": 1.973897491738914e-05,
+      "loss": 0.852,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2016832440703902,
+      "grad_norm": 2.5728999188481927,
+      "learning_rate": 1.973841209985243e-05,
+      "loss": 0.8264,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2018362662586075,
+      "grad_norm": 2.762783647383601,
+      "learning_rate": 1.973784868424174e-05,
+      "loss": 0.9527,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2019892884468248,
+      "grad_norm": 2.5039467780176,
+      "learning_rate": 1.9737284670591662e-05,
+      "loss": 0.8145,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2021423106350421,
+      "grad_norm": 2.741462373035885,
+      "learning_rate": 1.973672005893684e-05,
+      "loss": 0.8645,
+      "step": 1321
+    },
+    {
+      "epoch": 0.20229533282325937,
+      "grad_norm": 2.589649068698681,
+      "learning_rate": 1.973615484931195e-05,
+      "loss": 0.9089,
+      "step": 1322
+    },
+    {
+      "epoch": 0.20244835501147668,
+      "grad_norm": 2.2420383393474625,
+      "learning_rate": 1.9735589041751702e-05,
+      "loss": 0.7429,
+      "step": 1323
+    },
+    {
+      "epoch": 0.20260137719969395,
+      "grad_norm": 2.8834726302997473,
+      "learning_rate": 1.9735022636290845e-05,
+      "loss": 0.8063,
+      "step": 1324
+    },
+    {
+      "epoch": 0.20275439938791126,
+      "grad_norm": 2.8281724474513608,
+      "learning_rate": 1.9734455632964166e-05,
+      "loss": 0.9345,
+      "step": 1325
+    },
+    {
+      "epoch": 0.20290742157612854,
+      "grad_norm": 2.6418928860290967,
+      "learning_rate": 1.9733888031806485e-05,
+      "loss": 0.8527,
+      "step": 1326
+    },
+    {
+      "epoch": 0.20306044376434584,
+      "grad_norm": 2.801742930828167,
+      "learning_rate": 1.973331983285266e-05,
+      "loss": 0.902,
+      "step": 1327
+    },
+    {
+      "epoch": 0.20321346595256312,
+      "grad_norm": 2.724387569739316,
+      "learning_rate": 1.9732751036137588e-05,
+      "loss": 0.9353,
+      "step": 1328
+    },
+    {
+      "epoch": 0.20336648814078043,
+      "grad_norm": 2.852911846115286,
+      "learning_rate": 1.97321816416962e-05,
+      "loss": 0.9431,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2035195103289977,
+      "grad_norm": 2.9385006447379203,
+      "learning_rate": 1.973161164956346e-05,
+      "loss": 0.7805,
+      "step": 1330
+    },
+    {
+      "epoch": 0.20367253251721498,
+      "grad_norm": 2.6617267844097,
+      "learning_rate": 1.9731041059774387e-05,
+      "loss": 0.9011,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2038255547054323,
+      "grad_norm": 2.535089745451929,
+      "learning_rate": 1.973046987236401e-05,
+      "loss": 0.8131,
+      "step": 1332
+    },
+    {
+      "epoch": 0.20397857689364957,
+      "grad_norm": 2.5575312901886074,
+      "learning_rate": 1.972989808736741e-05,
+      "loss": 0.9255,
+      "step": 1333
+    },
+    {
+      "epoch": 0.20413159908186687,
+      "grad_norm": 2.9489071986972246,
+      "learning_rate": 1.972932570481971e-05,
+      "loss": 0.8734,
+      "step": 1334
+    },
+    {
+      "epoch": 0.20428462127008415,
+      "grad_norm": 2.6401501138319814,
+      "learning_rate": 1.9728752724756052e-05,
+      "loss": 0.8659,
+      "step": 1335
+    },
+    {
+      "epoch": 0.20443764345830145,
+      "grad_norm": 2.57743017743404,
+      "learning_rate": 1.9728179147211634e-05,
+      "loss": 0.6302,
+      "step": 1336
+    },
+    {
+      "epoch": 0.20459066564651873,
+      "grad_norm": 2.696725935953648,
+      "learning_rate": 1.9727604972221674e-05,
+      "loss": 0.9445,
+      "step": 1337
+    },
+    {
+      "epoch": 0.20474368783473604,
+      "grad_norm": 2.627476393799533,
+      "learning_rate": 1.9727030199821443e-05,
+      "loss": 0.7543,
+      "step": 1338
+    },
+    {
+      "epoch": 0.20489671002295332,
+      "grad_norm": 2.3880575508880604,
+      "learning_rate": 1.9726454830046233e-05,
+      "loss": 0.7506,
+      "step": 1339
+    },
+    {
+      "epoch": 0.20504973221117062,
+      "grad_norm": 2.7835239139366177,
+      "learning_rate": 1.9725878862931376e-05,
+      "loss": 0.8102,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2052027543993879,
+      "grad_norm": 2.7321404831940614,
+      "learning_rate": 1.9725302298512257e-05,
+      "loss": 0.7956,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2053557765876052,
+      "grad_norm": 2.457638556107576,
+      "learning_rate": 1.9724725136824277e-05,
+      "loss": 0.8221,
+      "step": 1342
+    },
+    {
+      "epoch": 0.20550879877582248,
+      "grad_norm": 2.4536782205822516,
+      "learning_rate": 1.9724147377902884e-05,
+      "loss": 0.71,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2056618209640398,
+      "grad_norm": 2.7635778874710564,
+      "learning_rate": 1.9723569021783557e-05,
+      "loss": 0.8422,
+      "step": 1344
+    },
+    {
+      "epoch": 0.20581484315225707,
+      "grad_norm": 2.600934468297566,
+      "learning_rate": 1.9722990068501818e-05,
+      "loss": 0.8543,
+      "step": 1345
+    },
+    {
+      "epoch": 0.20596786534047437,
+      "grad_norm": 2.5400207546187388,
+      "learning_rate": 1.972241051809322e-05,
+      "loss": 0.8007,
+      "step": 1346
+    },
+    {
+      "epoch": 0.20612088752869165,
+      "grad_norm": 2.709332583626073,
+      "learning_rate": 1.9721830370593364e-05,
+      "loss": 0.8138,
+      "step": 1347
+    },
+    {
+      "epoch": 0.20627390971690895,
+      "grad_norm": 2.724189789443129,
+      "learning_rate": 1.972124962603787e-05,
+      "loss": 0.9304,
+      "step": 1348
+    },
+    {
+      "epoch": 0.20642693190512623,
+      "grad_norm": 2.4858669918490834,
+      "learning_rate": 1.9720668284462407e-05,
+      "loss": 0.7141,
+      "step": 1349
+    },
+    {
+      "epoch": 0.20657995409334354,
+      "grad_norm": 2.4065464378747254,
+      "learning_rate": 1.9720086345902675e-05,
+      "loss": 0.7101,
+      "step": 1350
+    },
+    {
+      "epoch": 0.20673297628156082,
+      "grad_norm": 2.524649685819583,
+      "learning_rate": 1.9719503810394417e-05,
+      "loss": 0.9611,
+      "step": 1351
+    },
+    {
+      "epoch": 0.20688599846977812,
+      "grad_norm": 2.7392602306838083,
+      "learning_rate": 1.9718920677973407e-05,
+      "loss": 0.8701,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2070390206579954,
+      "grad_norm": 2.8684316541922716,
+      "learning_rate": 1.9718336948675457e-05,
+      "loss": 0.8392,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2071920428462127,
+      "grad_norm": 2.797057123846008,
+      "learning_rate": 1.9717752622536417e-05,
+      "loss": 0.8712,
+      "step": 1354
+    },
+    {
+      "epoch": 0.20734506503442998,
+      "grad_norm": 2.7197161819835194,
+      "learning_rate": 1.9717167699592173e-05,
+      "loss": 0.8252,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2074980872226473,
+      "grad_norm": 2.3072808087513605,
+      "learning_rate": 1.9716582179878645e-05,
+      "loss": 0.803,
+      "step": 1356
+    },
+    {
+      "epoch": 0.20765110941086456,
+      "grad_norm": 2.923572676762194,
+      "learning_rate": 1.97159960634318e-05,
+      "loss": 0.9197,
+      "step": 1357
+    },
+    {
+      "epoch": 0.20780413159908187,
+      "grad_norm": 2.792292441459427,
+      "learning_rate": 1.9715409350287618e-05,
+      "loss": 0.8846,
+      "step": 1358
+    },
+    {
+      "epoch": 0.20795715378729915,
+      "grad_norm": 2.793577571557601,
+      "learning_rate": 1.9714822040482144e-05,
+      "loss": 0.9456,
+      "step": 1359
+    },
+    {
+      "epoch": 0.20811017597551645,
+      "grad_norm": 2.5197653839780094,
+      "learning_rate": 1.9714234134051446e-05,
+      "loss": 0.8231,
+      "step": 1360
+    },
+    {
+      "epoch": 0.20826319816373373,
+      "grad_norm": 2.5856070289772943,
+      "learning_rate": 1.9713645631031628e-05,
+      "loss": 0.8052,
+      "step": 1361
+    },
+    {
+      "epoch": 0.20841622035195104,
+      "grad_norm": 2.4578741682910694,
+      "learning_rate": 1.971305653145883e-05,
+      "loss": 0.7517,
+      "step": 1362
+    },
+    {
+      "epoch": 0.20856924254016831,
+      "grad_norm": 2.519889808986514,
+      "learning_rate": 1.9712466835369234e-05,
+      "loss": 0.8433,
+      "step": 1363
+    },
+    {
+      "epoch": 0.20872226472838562,
+      "grad_norm": 2.377119912314756,
+      "learning_rate": 1.9711876542799056e-05,
+      "loss": 0.8387,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2088752869166029,
+      "grad_norm": 2.577350155100122,
+      "learning_rate": 1.9711285653784543e-05,
+      "loss": 0.8693,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2090283091048202,
+      "grad_norm": 2.3743002683845544,
+      "learning_rate": 1.971069416836199e-05,
+      "loss": 0.7715,
+      "step": 1366
+    },
+    {
+      "epoch": 0.20918133129303748,
+      "grad_norm": 2.5970257954430167,
+      "learning_rate": 1.9710102086567714e-05,
+      "loss": 0.8364,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2093343534812548,
+      "grad_norm": 2.463807934762222,
+      "learning_rate": 1.9709509408438087e-05,
+      "loss": 0.9367,
+      "step": 1368
+    },
+    {
+      "epoch": 0.20948737566947206,
+      "grad_norm": 2.5215756250590173,
+      "learning_rate": 1.9708916134009507e-05,
+      "loss": 0.9188,
+      "step": 1369
+    },
+    {
+      "epoch": 0.20964039785768937,
+      "grad_norm": 2.5332861323220026,
+      "learning_rate": 1.97083222633184e-05,
+      "loss": 0.8427,
+      "step": 1370
+    },
+    {
+      "epoch": 0.20979342004590665,
+      "grad_norm": 3.1280702666688738,
+      "learning_rate": 1.9707727796401243e-05,
+      "loss": 0.8749,
+      "step": 1371
+    },
+    {
+      "epoch": 0.20994644223412395,
+      "grad_norm": 2.7501921842495447,
+      "learning_rate": 1.970713273329455e-05,
+      "loss": 0.929,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21009946442234123,
+      "grad_norm": 2.489037234979162,
+      "learning_rate": 1.970653707403486e-05,
+      "loss": 0.8092,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21025248661055854,
+      "grad_norm": 2.6729264987086436,
+      "learning_rate": 1.9705940818658753e-05,
+      "loss": 0.8513,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21040550879877581,
+      "grad_norm": 2.7135658037185815,
+      "learning_rate": 1.9705343967202853e-05,
+      "loss": 0.8504,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21055853098699312,
+      "grad_norm": 2.748488375492164,
+      "learning_rate": 1.970474651970381e-05,
+      "loss": 0.8899,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2107115531752104,
+      "grad_norm": 2.851236345307254,
+      "learning_rate": 1.9704148476198323e-05,
+      "loss": 0.8719,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2108645753634277,
+      "grad_norm": 2.7526677950267637,
+      "learning_rate": 1.9703549836723112e-05,
+      "loss": 0.9131,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21101759755164498,
+      "grad_norm": 2.7251933214674784,
+      "learning_rate": 1.9702950601314948e-05,
+      "loss": 0.7765,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2111706197398623,
+      "grad_norm": 2.422786440415208,
+      "learning_rate": 1.9702350770010625e-05,
+      "loss": 0.816,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21132364192807956,
+      "grad_norm": 2.8510678470175823,
+      "learning_rate": 1.9701750342846985e-05,
+      "loss": 0.9267,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21147666411629687,
+      "grad_norm": 2.445685396692264,
+      "learning_rate": 1.970114931986091e-05,
+      "loss": 0.7711,
+      "step": 1382
+    },
+    {
+      "epoch": 0.21162968630451415,
+      "grad_norm": 2.5729642145381395,
+      "learning_rate": 1.9700547701089297e-05,
+      "loss": 0.8154,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21178270849273145,
+      "grad_norm": 2.3053407957709378,
+      "learning_rate": 1.9699945486569102e-05,
+      "loss": 0.8914,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21193573068094873,
+      "grad_norm": 2.3823797011223373,
+      "learning_rate": 1.969934267633731e-05,
+      "loss": 0.7488,
+      "step": 1385
+    },
+    {
+      "epoch": 0.21208875286916604,
+      "grad_norm": 2.482422953417084,
+      "learning_rate": 1.969873927043094e-05,
+      "loss": 0.8749,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21224177505738331,
+      "grad_norm": 2.779771599316904,
+      "learning_rate": 1.969813526888705e-05,
+      "loss": 0.8083,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21239479724560062,
+      "grad_norm": 2.651475617855006,
+      "learning_rate": 1.9697530671742733e-05,
+      "loss": 0.8822,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2125478194338179,
+      "grad_norm": 2.4441971083865113,
+      "learning_rate": 1.969692547903512e-05,
+      "loss": 0.9099,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2127008416220352,
+      "grad_norm": 2.901432036147306,
+      "learning_rate": 1.969631969080138e-05,
+      "loss": 0.8652,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21285386381025248,
+      "grad_norm": 2.7019808196456974,
+      "learning_rate": 1.9695713307078718e-05,
+      "loss": 0.8447,
+      "step": 1391
+    },
+    {
+      "epoch": 0.21300688599846979,
+      "grad_norm": 2.524291797108313,
+      "learning_rate": 1.9695106327904367e-05,
+      "loss": 0.9771,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21315990818668706,
+      "grad_norm": 2.432632551140893,
+      "learning_rate": 1.9694498753315613e-05,
+      "loss": 0.816,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21331293037490437,
+      "grad_norm": 3.0185482439437012,
+      "learning_rate": 1.9693890583349762e-05,
+      "loss": 1.0286,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21346595256312165,
+      "grad_norm": 2.473777410059679,
+      "learning_rate": 1.9693281818044168e-05,
+      "loss": 0.8313,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21361897475133895,
+      "grad_norm": 2.9398419262263973,
+      "learning_rate": 1.969267245743622e-05,
+      "loss": 0.9696,
+      "step": 1396
+    },
+    {
+      "epoch": 0.21377199693955623,
+      "grad_norm": 2.5598888551944197,
+      "learning_rate": 1.9692062501563333e-05,
+      "loss": 0.7388,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21392501912777354,
+      "grad_norm": 2.6365253523922365,
+      "learning_rate": 1.9691451950462977e-05,
+      "loss": 0.8378,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2140780413159908,
+      "grad_norm": 2.6272557509945313,
+      "learning_rate": 1.9690840804172644e-05,
+      "loss": 0.7908,
+      "step": 1399
+    },
+    {
+      "epoch": 0.21423106350420812,
+      "grad_norm": 2.457956626306656,
+      "learning_rate": 1.9690229062729863e-05,
+      "loss": 0.9102,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2143840856924254,
+      "grad_norm": 2.7856149622579003,
+      "learning_rate": 1.968961672617221e-05,
+      "loss": 0.7951,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2145371078806427,
+      "grad_norm": 2.5890160262046766,
+      "learning_rate": 1.9689003794537286e-05,
+      "loss": 0.8368,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21469013006885998,
+      "grad_norm": 2.5397138932158865,
+      "learning_rate": 1.9688390267862737e-05,
+      "loss": 0.8831,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21484315225707729,
+      "grad_norm": 2.5658431838731195,
+      "learning_rate": 1.968777614618624e-05,
+      "loss": 0.8381,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21499617444529456,
+      "grad_norm": 2.551108162079543,
+      "learning_rate": 1.9687161429545512e-05,
+      "loss": 0.9045,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21514919663351187,
+      "grad_norm": 2.4869151629476436,
+      "learning_rate": 1.96865461179783e-05,
+      "loss": 0.7808,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21530221882172915,
+      "grad_norm": 2.733337979301424,
+      "learning_rate": 1.96859302115224e-05,
+      "loss": 0.9888,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21545524100994645,
+      "grad_norm": 2.6573788126722486,
+      "learning_rate": 1.9685313710215634e-05,
+      "loss": 0.8592,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21560826319816373,
+      "grad_norm": 2.4780481991806367,
+      "learning_rate": 1.9684696614095865e-05,
+      "loss": 0.7615,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21576128538638104,
+      "grad_norm": 2.8468670949990713,
+      "learning_rate": 1.968407892320099e-05,
+      "loss": 0.8195,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2159143075745983,
+      "grad_norm": 2.542501474269545,
+      "learning_rate": 1.968346063756894e-05,
+      "loss": 0.7449,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21606732976281562,
+      "grad_norm": 2.5323321906232006,
+      "learning_rate": 1.9682841757237693e-05,
+      "loss": 0.8259,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2162203519510329,
+      "grad_norm": 2.6822987672510616,
+      "learning_rate": 1.9682222282245255e-05,
+      "loss": 0.9143,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2163733741392502,
+      "grad_norm": 2.661368983707966,
+      "learning_rate": 1.9681602212629668e-05,
+      "loss": 0.8521,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21652639632746748,
+      "grad_norm": 2.4421559076683925,
+      "learning_rate": 1.9680981548429017e-05,
+      "loss": 0.7393,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21667941851568479,
+      "grad_norm": 2.7182110574776375,
+      "learning_rate": 1.9680360289681415e-05,
+      "loss": 0.8152,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21683244070390206,
+      "grad_norm": 2.7587215745106426,
+      "learning_rate": 1.967973843642502e-05,
+      "loss": 0.9543,
+      "step": 1417
+    },
+    {
+      "epoch": 0.21698546289211937,
+      "grad_norm": 2.443920441762142,
+      "learning_rate": 1.967911598869802e-05,
+      "loss": 0.8022,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21713848508033665,
+      "grad_norm": 2.770489472508194,
+      "learning_rate": 1.967849294653864e-05,
+      "loss": 0.8629,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21729150726855395,
+      "grad_norm": 2.6168180326566186,
+      "learning_rate": 1.9677869309985146e-05,
+      "loss": 0.8642,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21744452945677123,
+      "grad_norm": 2.752417489975351,
+      "learning_rate": 1.9677245079075837e-05,
+      "loss": 0.789,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21759755164498853,
+      "grad_norm": 2.683755581101207,
+      "learning_rate": 1.967662025384905e-05,
+      "loss": 0.9186,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2177505738332058,
+      "grad_norm": 2.804693835725841,
+      "learning_rate": 1.967599483434316e-05,
+      "loss": 0.9001,
+      "step": 1423
+    },
+    {
+      "epoch": 0.21790359602142312,
+      "grad_norm": 3.135183658906024,
+      "learning_rate": 1.9675368820596575e-05,
+      "loss": 0.9598,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2180566182096404,
+      "grad_norm": 3.0719430556931067,
+      "learning_rate": 1.9674742212647738e-05,
+      "loss": 0.8038,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2182096403978577,
+      "grad_norm": 2.60160695139502,
+      "learning_rate": 1.9674115010535135e-05,
+      "loss": 0.8428,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21836266258607498,
+      "grad_norm": 2.6151319424341897,
+      "learning_rate": 1.9673487214297284e-05,
+      "loss": 0.8505,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21851568477429228,
+      "grad_norm": 2.575352044936147,
+      "learning_rate": 1.967285882397274e-05,
+      "loss": 0.826,
+      "step": 1428
+    },
+    {
+      "epoch": 0.21866870696250956,
+      "grad_norm": 2.628612019649863,
+      "learning_rate": 1.9672229839600098e-05,
+      "loss": 0.7893,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21882172915072687,
+      "grad_norm": 2.512796903702364,
+      "learning_rate": 1.9671600261217978e-05,
+      "loss": 0.7888,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21897475133894415,
+      "grad_norm": 2.6086882195208525,
+      "learning_rate": 1.9670970088865052e-05,
+      "loss": 0.794,
+      "step": 1431
+    },
+    {
+      "epoch": 0.21912777352716145,
+      "grad_norm": 2.9600265005459225,
+      "learning_rate": 1.9670339322580023e-05,
+      "loss": 0.932,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21928079571537873,
+      "grad_norm": 2.366110521394664,
+      "learning_rate": 1.966970796240162e-05,
+      "loss": 0.885,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21943381790359603,
+      "grad_norm": 2.406125948360677,
+      "learning_rate": 1.966907600836863e-05,
+      "loss": 0.8062,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2195868400918133,
+      "grad_norm": 2.735448209351719,
+      "learning_rate": 1.9668443460519854e-05,
+      "loss": 0.9436,
+      "step": 1435
+    },
+    {
+      "epoch": 0.21973986228003062,
+      "grad_norm": 2.468087419582429,
+      "learning_rate": 1.966781031889414e-05,
+      "loss": 0.8398,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2198928844682479,
+      "grad_norm": 2.4874153722433316,
+      "learning_rate": 1.9667176583530377e-05,
+      "loss": 0.8523,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2200459066564652,
+      "grad_norm": 2.6445890152061917,
+      "learning_rate": 1.9666542254467478e-05,
+      "loss": 1.0136,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22019892884468248,
+      "grad_norm": 2.607525745722285,
+      "learning_rate": 1.9665907331744404e-05,
+      "loss": 0.8658,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22035195103289976,
+      "grad_norm": 2.7498225059483348,
+      "learning_rate": 1.966527181540015e-05,
+      "loss": 0.8795,
+      "step": 1440
+    },
+    {
+      "epoch": 0.22050497322111706,
+      "grad_norm": 2.449442723272099,
+      "learning_rate": 1.9664635705473745e-05,
+      "loss": 0.7556,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22065799540933434,
+      "grad_norm": 2.5182627516208265,
+      "learning_rate": 1.9663999002004247e-05,
+      "loss": 0.9188,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22081101759755165,
+      "grad_norm": 3.0815320780904214,
+      "learning_rate": 1.966336170503077e-05,
+      "loss": 0.8873,
+      "step": 1443
+    },
+    {
+      "epoch": 0.22096403978576892,
+      "grad_norm": 2.1521962301015205,
+      "learning_rate": 1.9662723814592443e-05,
+      "loss": 0.6666,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22111706197398623,
+      "grad_norm": 2.617498957904519,
+      "learning_rate": 1.9662085330728448e-05,
+      "loss": 0.8849,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2212700841622035,
+      "grad_norm": 2.6878494492762717,
+      "learning_rate": 1.9661446253477995e-05,
+      "loss": 0.8731,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2214231063504208,
+      "grad_norm": 2.4562980436730086,
+      "learning_rate": 1.966080658288033e-05,
+      "loss": 0.7922,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2215761285386381,
+      "grad_norm": 2.4564471373976278,
+      "learning_rate": 1.966016631897474e-05,
+      "loss": 0.8254,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2217291507268554,
+      "grad_norm": 2.649776130638235,
+      "learning_rate": 1.9659525461800546e-05,
+      "loss": 0.9063,
+      "step": 1449
+    },
+    {
+      "epoch": 0.22188217291507267,
+      "grad_norm": 2.2959740420216237,
+      "learning_rate": 1.9658884011397105e-05,
+      "loss": 0.7877,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22203519510328998,
+      "grad_norm": 2.5813619059244166,
+      "learning_rate": 1.965824196780381e-05,
+      "loss": 0.8595,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22218821729150726,
+      "grad_norm": 2.584172187703942,
+      "learning_rate": 1.9657599331060097e-05,
+      "loss": 0.7845,
+      "step": 1452
+    },
+    {
+      "epoch": 0.22234123947972456,
+      "grad_norm": 2.7062610966861564,
+      "learning_rate": 1.9656956101205426e-05,
+      "loss": 0.8721,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22249426166794184,
+      "grad_norm": 2.374713272758218,
+      "learning_rate": 1.9656312278279298e-05,
+      "loss": 0.6398,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22264728385615914,
+      "grad_norm": 2.562393686891158,
+      "learning_rate": 1.965566786232126e-05,
+      "loss": 0.8596,
+      "step": 1455
+    },
+    {
+      "epoch": 0.22280030604437642,
+      "grad_norm": 2.2857421843137518,
+      "learning_rate": 1.9655022853370886e-05,
+      "loss": 0.7233,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22295332823259373,
+      "grad_norm": 2.364375523326793,
+      "learning_rate": 1.965437725146779e-05,
+      "loss": 0.777,
+      "step": 1457
+    },
+    {
+      "epoch": 0.223106350420811,
+      "grad_norm": 2.326524319190945,
+      "learning_rate": 1.9653731056651616e-05,
+      "loss": 0.7536,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2232593726090283,
+      "grad_norm": 2.417153715891703,
+      "learning_rate": 1.9653084268962054e-05,
+      "loss": 0.7981,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2234123947972456,
+      "grad_norm": 2.5776287610942306,
+      "learning_rate": 1.965243688843882e-05,
+      "loss": 0.7941,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2235654169854629,
+      "grad_norm": 2.392772961147377,
+      "learning_rate": 1.965178891512168e-05,
+      "loss": 0.9104,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22371843917368017,
+      "grad_norm": 2.487273411363196,
+      "learning_rate": 1.9651140349050428e-05,
+      "loss": 0.7555,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22387146136189748,
+      "grad_norm": 2.7143668921668245,
+      "learning_rate": 1.9650491190264887e-05,
+      "loss": 0.8837,
+      "step": 1463
+    },
+    {
+      "epoch": 0.22402448355011476,
+      "grad_norm": 2.734873235580024,
+      "learning_rate": 1.964984143880493e-05,
+      "loss": 0.815,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22417750573833206,
+      "grad_norm": 2.928990160124733,
+      "learning_rate": 1.9649191094710457e-05,
+      "loss": 0.8404,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22433052792654934,
+      "grad_norm": 2.5593116581170583,
+      "learning_rate": 1.9648540158021416e-05,
+      "loss": 0.7353,
+      "step": 1466
+    },
+    {
+      "epoch": 0.22448355011476664,
+      "grad_norm": 2.6266572490060804,
+      "learning_rate": 1.9647888628777777e-05,
+      "loss": 0.9329,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22463657230298392,
+      "grad_norm": 2.8704971084818878,
+      "learning_rate": 1.9647236507019552e-05,
+      "loss": 0.9432,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22478959449120123,
+      "grad_norm": 2.590453922700811,
+      "learning_rate": 1.9646583792786795e-05,
+      "loss": 0.7691,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2249426166794185,
+      "grad_norm": 3.1415820506912624,
+      "learning_rate": 1.9645930486119588e-05,
+      "loss": 1.0224,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2250956388676358,
+      "grad_norm": 2.5309118552802836,
+      "learning_rate": 1.964527658705806e-05,
+      "loss": 0.886,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2252486610558531,
+      "grad_norm": 2.5972158241040657,
+      "learning_rate": 1.9644622095642358e-05,
+      "loss": 0.6973,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2254016832440704,
+      "grad_norm": 2.5546268806279455,
+      "learning_rate": 1.9643967011912685e-05,
+      "loss": 0.8449,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22555470543228767,
+      "grad_norm": 2.3479011681317314,
+      "learning_rate": 1.964331133590927e-05,
+      "loss": 0.8669,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22570772762050498,
+      "grad_norm": 2.515862803842189,
+      "learning_rate": 1.9642655067672384e-05,
+      "loss": 0.8085,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22586074980872226,
+      "grad_norm": 2.3668911776660755,
+      "learning_rate": 1.9641998207242325e-05,
+      "loss": 0.7926,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22601377199693956,
+      "grad_norm": 2.280772898900187,
+      "learning_rate": 1.9641340754659438e-05,
+      "loss": 0.5904,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22616679418515684,
+      "grad_norm": 2.465742578131,
+      "learning_rate": 1.9640682709964098e-05,
+      "loss": 0.7349,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22631981637337414,
+      "grad_norm": 2.530176243255086,
+      "learning_rate": 1.964002407319672e-05,
+      "loss": 0.8513,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22647283856159142,
+      "grad_norm": 2.5405734648853873,
+      "learning_rate": 1.9639364844397745e-05,
+      "loss": 0.9168,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22662586074980873,
+      "grad_norm": 2.8804862950288337,
+      "learning_rate": 1.9638705023607668e-05,
+      "loss": 0.8588,
+      "step": 1481
+    },
+    {
+      "epoch": 0.226778882938026,
+      "grad_norm": 2.260546018171997,
+      "learning_rate": 1.9638044610867013e-05,
+      "loss": 0.7765,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2269319051262433,
+      "grad_norm": 2.536491527405855,
+      "learning_rate": 1.963738360621633e-05,
+      "loss": 0.764,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2270849273144606,
+      "grad_norm": 2.0029390016703763,
+      "learning_rate": 1.963672200969622e-05,
+      "loss": 0.7933,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2272379495026779,
+      "grad_norm": 2.681926900943292,
+      "learning_rate": 1.9636059821347315e-05,
+      "loss": 0.8794,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22739097169089517,
+      "grad_norm": 2.460030338333229,
+      "learning_rate": 1.9635397041210274e-05,
+      "loss": 0.8535,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22754399387911248,
+      "grad_norm": 2.666549304410844,
+      "learning_rate": 1.963473366932581e-05,
+      "loss": 0.7369,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22769701606732975,
+      "grad_norm": 2.5057662138797676,
+      "learning_rate": 1.9634069705734662e-05,
+      "loss": 0.8288,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22785003825554706,
+      "grad_norm": 2.5683484789985256,
+      "learning_rate": 1.9633405150477603e-05,
+      "loss": 0.7962,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22800306044376434,
+      "grad_norm": 3.0952874862064816,
+      "learning_rate": 1.9632740003595447e-05,
+      "loss": 0.879,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22815608263198164,
+      "grad_norm": 2.8884222769603065,
+      "learning_rate": 1.9632074265129044e-05,
+      "loss": 0.7994,
+      "step": 1491
+    },
+    {
+      "epoch": 0.22830910482019892,
+      "grad_norm": 2.735873639194007,
+      "learning_rate": 1.963140793511928e-05,
+      "loss": 0.8619,
+      "step": 1492
+    },
+    {
+      "epoch": 0.22846212700841623,
+      "grad_norm": 2.631822280768736,
+      "learning_rate": 1.9630741013607077e-05,
+      "loss": 0.8749,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2286151491966335,
+      "grad_norm": 2.577737303029769,
+      "learning_rate": 1.9630073500633392e-05,
+      "loss": 0.8527,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2287681713848508,
+      "grad_norm": 2.852657029630802,
+      "learning_rate": 1.962940539623922e-05,
+      "loss": 0.9213,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2289211935730681,
+      "grad_norm": 2.5092520712765447,
+      "learning_rate": 1.962873670046559e-05,
+      "loss": 0.7863,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2290742157612854,
+      "grad_norm": 3.0271335789574065,
+      "learning_rate": 1.962806741335357e-05,
+      "loss": 0.8167,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22922723794950267,
+      "grad_norm": 2.671356145739248,
+      "learning_rate": 1.962739753494427e-05,
+      "loss": 0.9197,
+      "step": 1498
+    },
+    {
+      "epoch": 0.22938026013771998,
+      "grad_norm": 2.6843150913405354,
+      "learning_rate": 1.9626727065278827e-05,
+      "loss": 0.9332,
+      "step": 1499
+    },
+    {
+      "epoch": 0.22953328232593725,
+      "grad_norm": 2.5279817214021385,
+      "learning_rate": 1.962605600439841e-05,
+      "loss": 0.8936,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22968630451415456,
+      "grad_norm": 2.825279808824693,
+      "learning_rate": 1.9625384352344235e-05,
+      "loss": 0.8902,
+      "step": 1501
+    },
+    {
+      "epoch": 0.22983932670237184,
+      "grad_norm": 2.592876328071922,
+      "learning_rate": 1.9624712109157554e-05,
+      "loss": 0.9093,
+      "step": 1502
+    },
+    {
+      "epoch": 0.22999234889058914,
+      "grad_norm": 2.328063927339844,
+      "learning_rate": 1.9624039274879654e-05,
+      "loss": 0.7551,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23014537107880642,
+      "grad_norm": 2.597131304137858,
+      "learning_rate": 1.962336584955185e-05,
+      "loss": 0.8054,
+      "step": 1504
+    },
+    {
+      "epoch": 0.23029839326702373,
+      "grad_norm": 2.561933334091904,
+      "learning_rate": 1.9622691833215502e-05,
+      "loss": 0.8699,
+      "step": 1505
+    },
+    {
+      "epoch": 0.230451415455241,
+      "grad_norm": 2.6185390256998904,
+      "learning_rate": 1.9622017225912e-05,
+      "loss": 0.882,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2306044376434583,
+      "grad_norm": 2.4996637324030844,
+      "learning_rate": 1.9621342027682785e-05,
+      "loss": 0.7593,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2307574598316756,
+      "grad_norm": 2.337279426626155,
+      "learning_rate": 1.9620666238569317e-05,
+      "loss": 0.7016,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2309104820198929,
+      "grad_norm": 2.644195150141663,
+      "learning_rate": 1.96199898586131e-05,
+      "loss": 0.8604,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23106350420811017,
+      "grad_norm": 2.6142675375463624,
+      "learning_rate": 1.9619312887855666e-05,
+      "loss": 0.8094,
+      "step": 1510
+    },
+    {
+      "epoch": 0.23121652639632748,
+      "grad_norm": 2.331606054900535,
+      "learning_rate": 1.96186353263386e-05,
+      "loss": 0.7867,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23136954858454475,
+      "grad_norm": 2.8201130356617057,
+      "learning_rate": 1.9617957174103512e-05,
+      "loss": 0.9134,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23152257077276206,
+      "grad_norm": 3.166099027860909,
+      "learning_rate": 1.9617278431192045e-05,
+      "loss": 0.7511,
+      "step": 1513
+    },
+    {
+      "epoch": 0.23167559296097934,
+      "grad_norm": 2.5832948593251595,
+      "learning_rate": 1.961659909764589e-05,
+      "loss": 0.6853,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23182861514919664,
+      "grad_norm": 3.0072774469631534,
+      "learning_rate": 1.9615919173506763e-05,
+      "loss": 0.8164,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23198163733741392,
+      "grad_norm": 2.5885075319239084,
+      "learning_rate": 1.961523865881642e-05,
+      "loss": 0.8215,
+      "step": 1516
+    },
+    {
+      "epoch": 0.23213465952563123,
+      "grad_norm": 2.5422510418493127,
+      "learning_rate": 1.9614557553616658e-05,
+      "loss": 0.8625,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2322876817138485,
+      "grad_norm": 2.994853277911608,
+      "learning_rate": 1.96138758579493e-05,
+      "loss": 1.0072,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2324407039020658,
+      "grad_norm": 3.8292304341773473,
+      "learning_rate": 1.961319357185622e-05,
+      "loss": 0.8986,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2325937260902831,
+      "grad_norm": 2.4955902381832273,
+      "learning_rate": 1.9612510695379318e-05,
+      "loss": 0.8552,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2327467482785004,
+      "grad_norm": 2.5365745905944976,
+      "learning_rate": 1.9611827228560526e-05,
+      "loss": 0.833,
+      "step": 1521
+    },
+    {
+      "epoch": 0.23289977046671767,
+      "grad_norm": 2.5287596898930675,
+      "learning_rate": 1.9611143171441824e-05,
+      "loss": 0.7321,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23305279265493498,
+      "grad_norm": 2.5755887252276604,
+      "learning_rate": 1.961045852406522e-05,
+      "loss": 0.8743,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23320581484315225,
+      "grad_norm": 2.6042826275761564,
+      "learning_rate": 1.960977328647276e-05,
+      "loss": 0.9544,
+      "step": 1524
+    },
+    {
+      "epoch": 0.23335883703136956,
+      "grad_norm": 2.622514539233016,
+      "learning_rate": 1.9609087458706528e-05,
+      "loss": 0.9218,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23351185921958684,
+      "grad_norm": 2.4768249685418735,
+      "learning_rate": 1.9608401040808648e-05,
+      "loss": 0.8509,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23366488140780414,
+      "grad_norm": 2.5099475586849036,
+      "learning_rate": 1.960771403282127e-05,
+      "loss": 0.7543,
+      "step": 1527
+    },
+    {
+      "epoch": 0.23381790359602142,
+      "grad_norm": 2.4341180953680035,
+      "learning_rate": 1.9607026434786587e-05,
+      "loss": 0.7317,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23397092578423873,
+      "grad_norm": 2.722094841689053,
+      "learning_rate": 1.960633824674683e-05,
+      "loss": 0.8697,
+      "step": 1529
+    },
+    {
+      "epoch": 0.234123947972456,
+      "grad_norm": 2.8260615483151588,
+      "learning_rate": 1.960564946874426e-05,
+      "loss": 0.8851,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2342769701606733,
+      "grad_norm": 2.5417326953835824,
+      "learning_rate": 1.960496010082118e-05,
+      "loss": 0.8586,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2344299923488906,
+      "grad_norm": 2.5826334101465944,
+      "learning_rate": 1.9604270143019923e-05,
+      "loss": 0.8883,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2345830145371079,
+      "grad_norm": 2.678630361021353,
+      "learning_rate": 1.9603579595382866e-05,
+      "loss": 0.8858,
+      "step": 1533
+    },
+    {
+      "epoch": 0.23473603672532517,
+      "grad_norm": 2.5681223877610346,
+      "learning_rate": 1.9602888457952414e-05,
+      "loss": 0.8861,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23488905891354248,
+      "grad_norm": 2.556048444081866,
+      "learning_rate": 1.9602196730771017e-05,
+      "loss": 0.9105,
+      "step": 1535
+    },
+    {
+      "epoch": 0.23504208110175975,
+      "grad_norm": 2.2656890021267966,
+      "learning_rate": 1.9601504413881155e-05,
+      "loss": 0.8354,
+      "step": 1536
+    },
+    {
+      "epoch": 0.23519510328997706,
+      "grad_norm": 2.6891425941651375,
+      "learning_rate": 1.9600811507325346e-05,
+      "loss": 0.7968,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23534812547819434,
+      "grad_norm": 2.6911301647078805,
+      "learning_rate": 1.9600118011146142e-05,
+      "loss": 0.8348,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23550114766641164,
+      "grad_norm": 2.9627416505056092,
+      "learning_rate": 1.9599423925386135e-05,
+      "loss": 0.8664,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23565416985462892,
+      "grad_norm": 2.5888079738464387,
+      "learning_rate": 1.9598729250087953e-05,
+      "loss": 0.834,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23580719204284623,
+      "grad_norm": 2.8789926115688007,
+      "learning_rate": 1.9598033985294262e-05,
+      "loss": 0.9464,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2359602142310635,
+      "grad_norm": 2.575145545960676,
+      "learning_rate": 1.9597338131047747e-05,
+      "loss": 0.8041,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2361132364192808,
+      "grad_norm": 2.4750916407006516,
+      "learning_rate": 1.9596641687391157e-05,
+      "loss": 0.938,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2362662586074981,
+      "grad_norm": 2.3274446744899477,
+      "learning_rate": 1.9595944654367257e-05,
+      "loss": 0.8365,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2364192807957154,
+      "grad_norm": 2.474251844393818,
+      "learning_rate": 1.959524703201886e-05,
+      "loss": 0.8646,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23657230298393267,
+      "grad_norm": 2.5147567170404153,
+      "learning_rate": 1.9594548820388802e-05,
+      "loss": 0.8311,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23672532517214998,
+      "grad_norm": 2.5166123534386813,
+      "learning_rate": 1.959385001951997e-05,
+      "loss": 0.8027,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23687834736036725,
+      "grad_norm": 2.740847576540663,
+      "learning_rate": 1.9593150629455276e-05,
+      "loss": 0.953,
+      "step": 1548
+    },
+    {
+      "epoch": 0.23703136954858456,
+      "grad_norm": 2.3306374527547735,
+      "learning_rate": 1.9592450650237672e-05,
+      "loss": 0.8136,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23718439173680184,
+      "grad_norm": 2.777170941620961,
+      "learning_rate": 1.9591750081910147e-05,
+      "loss": 0.8697,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23733741392501911,
+      "grad_norm": 2.5805389495767574,
+      "learning_rate": 1.9591048924515727e-05,
+      "loss": 0.8805,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23749043611323642,
+      "grad_norm": 2.514446714606269,
+      "learning_rate": 1.9590347178097472e-05,
+      "loss": 0.763,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2376434583014537,
+      "grad_norm": 2.8092464729269424,
+      "learning_rate": 1.9589644842698483e-05,
+      "loss": 0.8816,
+      "step": 1553
+    },
+    {
+      "epoch": 0.237796480489671,
+      "grad_norm": 2.483884600727865,
+      "learning_rate": 1.9588941918361887e-05,
+      "loss": 0.8546,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23794950267788828,
+      "grad_norm": 2.474736699712295,
+      "learning_rate": 1.958823840513085e-05,
+      "loss": 0.8574,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23810252486610559,
+      "grad_norm": 2.4571375008471383,
+      "learning_rate": 1.9587534303048593e-05,
+      "loss": 0.8287,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23825554705432286,
+      "grad_norm": 2.3407274090716546,
+      "learning_rate": 1.9586829612158344e-05,
+      "loss": 0.6982,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23840856924254017,
+      "grad_norm": 2.377616391628943,
+      "learning_rate": 1.958612433250338e-05,
+      "loss": 0.8394,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23856159143075745,
+      "grad_norm": 2.5486874556181407,
+      "learning_rate": 1.9585418464127024e-05,
+      "loss": 0.761,
+      "step": 1559
+    },
+    {
+      "epoch": 0.23871461361897475,
+      "grad_norm": 2.766035827712233,
+      "learning_rate": 1.958471200707262e-05,
+      "loss": 0.9256,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23886763580719203,
+      "grad_norm": 2.7270798479698035,
+      "learning_rate": 1.958400496138356e-05,
+      "loss": 0.8441,
+      "step": 1561
+    },
+    {
+      "epoch": 0.23902065799540934,
+      "grad_norm": 2.5294552861595556,
+      "learning_rate": 1.9583297327103258e-05,
+      "loss": 0.8277,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2391736801836266,
+      "grad_norm": 2.5286458879300002,
+      "learning_rate": 1.9582589104275178e-05,
+      "loss": 0.9408,
+      "step": 1563
+    },
+    {
+      "epoch": 0.23932670237184392,
+      "grad_norm": 2.290446335806122,
+      "learning_rate": 1.9581880292942813e-05,
+      "loss": 0.7325,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2394797245600612,
+      "grad_norm": 2.542319946851815,
+      "learning_rate": 1.9581170893149696e-05,
+      "loss": 0.8568,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2396327467482785,
+      "grad_norm": 2.6996924947494594,
+      "learning_rate": 1.9580460904939393e-05,
+      "loss": 0.8699,
+      "step": 1566
+    },
+    {
+      "epoch": 0.23978576893649578,
+      "grad_norm": 2.7752847934882894,
+      "learning_rate": 1.9579750328355505e-05,
+      "loss": 0.8267,
+      "step": 1567
+    },
+    {
+      "epoch": 0.23993879112471309,
+      "grad_norm": 2.564510429754208,
+      "learning_rate": 1.957903916344168e-05,
+      "loss": 0.8308,
+      "step": 1568
+    },
+    {
+      "epoch": 0.24009181331293036,
+      "grad_norm": 2.425102215801101,
+      "learning_rate": 1.957832741024158e-05,
+      "loss": 0.806,
+      "step": 1569
+    },
+    {
+      "epoch": 0.24024483550114767,
+      "grad_norm": 2.6676736135165364,
+      "learning_rate": 1.957761506879892e-05,
+      "loss": 0.8601,
+      "step": 1570
+    },
+    {
+      "epoch": 0.24039785768936495,
+      "grad_norm": 2.4198971288915163,
+      "learning_rate": 1.9576902139157455e-05,
+      "loss": 0.77,
+      "step": 1571
+    },
+    {
+      "epoch": 0.24055087987758225,
+      "grad_norm": 2.2966349523292307,
+      "learning_rate": 1.9576188621360966e-05,
+      "loss": 0.7461,
+      "step": 1572
+    },
+    {
+      "epoch": 0.24070390206579953,
+      "grad_norm": 2.6079230848347956,
+      "learning_rate": 1.957547451545327e-05,
+      "loss": 0.8364,
+      "step": 1573
+    },
+    {
+      "epoch": 0.24085692425401684,
+      "grad_norm": 2.758369012562051,
+      "learning_rate": 1.9574759821478223e-05,
+      "loss": 0.9007,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2410099464422341,
+      "grad_norm": 2.3613131551646673,
+      "learning_rate": 1.9574044539479722e-05,
+      "loss": 0.8054,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24116296863045142,
+      "grad_norm": 2.464175192909486,
+      "learning_rate": 1.9573328669501693e-05,
+      "loss": 0.7988,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2413159908186687,
+      "grad_norm": 2.8590820543841837,
+      "learning_rate": 1.9572612211588094e-05,
+      "loss": 0.7991,
+      "step": 1577
+    },
+    {
+      "epoch": 0.241469013006886,
+      "grad_norm": 2.6513303722544386,
+      "learning_rate": 1.9571895165782933e-05,
+      "loss": 0.8747,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24162203519510328,
+      "grad_norm": 2.6797342921700555,
+      "learning_rate": 1.9571177532130245e-05,
+      "loss": 0.9601,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24177505738332059,
+      "grad_norm": 2.656625618019867,
+      "learning_rate": 1.9570459310674104e-05,
+      "loss": 0.9296,
+      "step": 1580
+    },
+    {
+      "epoch": 0.24192807957153786,
+      "grad_norm": 2.476973262298971,
+      "learning_rate": 1.9569740501458615e-05,
+      "loss": 0.8351,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24208110175975517,
+      "grad_norm": 2.6908872499133283,
+      "learning_rate": 1.9569021104527924e-05,
+      "loss": 0.9745,
+      "step": 1582
+    },
+    {
+      "epoch": 0.24223412394797245,
+      "grad_norm": 2.400852771610583,
+      "learning_rate": 1.956830111992621e-05,
+      "loss": 0.8625,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24238714613618975,
+      "grad_norm": 2.7311721684322574,
+      "learning_rate": 1.95675805476977e-05,
+      "loss": 0.8294,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24254016832440703,
+      "grad_norm": 3.031322195765501,
+      "learning_rate": 1.9566859387886638e-05,
+      "loss": 0.8948,
+      "step": 1585
+    },
+    {
+      "epoch": 0.24269319051262433,
+      "grad_norm": 2.386831353760463,
+      "learning_rate": 1.956613764053731e-05,
+      "loss": 0.8555,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2428462127008416,
+      "grad_norm": 2.5958286967053223,
+      "learning_rate": 1.956541530569405e-05,
+      "loss": 0.8871,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24299923488905892,
+      "grad_norm": 2.351972886279123,
+      "learning_rate": 1.9564692383401218e-05,
+      "loss": 0.7892,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2431522570772762,
+      "grad_norm": 2.563518227284029,
+      "learning_rate": 1.9563968873703206e-05,
+      "loss": 0.8589,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2433052792654935,
+      "grad_norm": 2.7469234929921242,
+      "learning_rate": 1.956324477664445e-05,
+      "loss": 0.9476,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24345830145371078,
+      "grad_norm": 2.5793234210805354,
+      "learning_rate": 1.9562520092269426e-05,
+      "loss": 0.9077,
+      "step": 1591
+    },
+    {
+      "epoch": 0.24361132364192808,
+      "grad_norm": 2.487999582227551,
+      "learning_rate": 1.956179482062263e-05,
+      "loss": 0.8971,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24376434583014536,
+      "grad_norm": 2.5411762234948734,
+      "learning_rate": 1.9561068961748612e-05,
+      "loss": 0.7937,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24391736801836267,
+      "grad_norm": 2.7545988237977275,
+      "learning_rate": 1.9560342515691942e-05,
+      "loss": 0.9105,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24407039020657995,
+      "grad_norm": 2.338750092973846,
+      "learning_rate": 1.9559615482497237e-05,
+      "loss": 0.8024,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24422341239479725,
+      "grad_norm": 2.5892173696364145,
+      "learning_rate": 1.9558887862209153e-05,
+      "loss": 0.9056,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24437643458301453,
+      "grad_norm": 2.779375109330424,
+      "learning_rate": 1.955815965487237e-05,
+      "loss": 0.8351,
+      "step": 1597
+    },
+    {
+      "epoch": 0.24452945677123183,
+      "grad_norm": 2.85621778641634,
+      "learning_rate": 1.9557430860531605e-05,
+      "loss": 0.8458,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2446824789594491,
+      "grad_norm": 3.0230061074447794,
+      "learning_rate": 1.955670147923163e-05,
+      "loss": 0.8472,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24483550114766642,
+      "grad_norm": 2.81465165939412,
+      "learning_rate": 1.9555971511017225e-05,
+      "loss": 0.8206,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2449885233358837,
+      "grad_norm": 2.4138314641623984,
+      "learning_rate": 1.9555240955933227e-05,
+      "loss": 0.8138,
+      "step": 1601
+    },
+    {
+      "epoch": 0.245141545524101,
+      "grad_norm": 2.8178514221242126,
+      "learning_rate": 1.9554509814024502e-05,
+      "loss": 0.9016,
+      "step": 1602
+    },
+    {
+      "epoch": 0.24529456771231828,
+      "grad_norm": 2.5742380441275796,
+      "learning_rate": 1.9553778085335953e-05,
+      "loss": 0.8801,
+      "step": 1603
+    },
+    {
+      "epoch": 0.24544758990053558,
+      "grad_norm": 2.2493911411311296,
+      "learning_rate": 1.955304576991252e-05,
+      "loss": 0.7488,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24560061208875286,
+      "grad_norm": 2.618681068118831,
+      "learning_rate": 1.9552312867799168e-05,
+      "loss": 0.8048,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24575363427697017,
+      "grad_norm": 2.882616971639018,
+      "learning_rate": 1.955157937904092e-05,
+      "loss": 0.9131,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24590665646518745,
+      "grad_norm": 2.5178465278639792,
+      "learning_rate": 1.955084530368281e-05,
+      "loss": 0.8579,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24605967865340475,
+      "grad_norm": 2.760529263401625,
+      "learning_rate": 1.955011064176993e-05,
+      "loss": 0.8543,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24621270084162203,
+      "grad_norm": 2.6012155141285143,
+      "learning_rate": 1.9549375393347397e-05,
+      "loss": 1.1095,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24636572302983933,
+      "grad_norm": 2.379103223108997,
+      "learning_rate": 1.9548639558460363e-05,
+      "loss": 0.7103,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2465187452180566,
+      "grad_norm": 2.4002591433736105,
+      "learning_rate": 1.954790313715402e-05,
+      "loss": 0.8428,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24667176740627392,
+      "grad_norm": 2.4612350390198134,
+      "learning_rate": 1.9547166129473592e-05,
+      "loss": 0.8282,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2468247895944912,
+      "grad_norm": 2.7307204781232044,
+      "learning_rate": 1.9546428535464342e-05,
+      "loss": 0.8591,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2469778117827085,
+      "grad_norm": 2.4852927020406375,
+      "learning_rate": 1.954569035517157e-05,
+      "loss": 0.8924,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24713083397092578,
+      "grad_norm": 2.4986642056279673,
+      "learning_rate": 1.9544951588640613e-05,
+      "loss": 0.8122,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24728385615914308,
+      "grad_norm": 2.4127278623822526,
+      "learning_rate": 1.954421223591684e-05,
+      "loss": 0.6704,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24743687834736036,
+      "grad_norm": 2.4815806633333413,
+      "learning_rate": 1.9543472297045652e-05,
+      "loss": 0.9251,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24758990053557767,
+      "grad_norm": 2.5451951843300957,
+      "learning_rate": 1.95427317720725e-05,
+      "loss": 0.77,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24774292272379494,
+      "grad_norm": 2.6155792651075465,
+      "learning_rate": 1.9541990661042855e-05,
+      "loss": 0.7563,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24789594491201225,
+      "grad_norm": 2.6364533209981116,
+      "learning_rate": 1.9541248964002237e-05,
+      "loss": 0.7738,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24804896710022953,
+      "grad_norm": 2.4555048705871365,
+      "learning_rate": 1.9540506680996194e-05,
+      "loss": 0.844,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24820198928844683,
+      "grad_norm": 2.5501946462827707,
+      "learning_rate": 1.9539763812070315e-05,
+      "loss": 0.8376,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2483550114766641,
+      "grad_norm": 2.393788755143763,
+      "learning_rate": 1.953902035727022e-05,
+      "loss": 0.8484,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24850803366488142,
+      "grad_norm": 2.7383748703121813,
+      "learning_rate": 1.9538276316641563e-05,
+      "loss": 0.8219,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2486610558530987,
+      "grad_norm": 2.7572133382018844,
+      "learning_rate": 1.9537531690230047e-05,
+      "loss": 0.8975,
+      "step": 1625
+    },
+    {
+      "epoch": 0.248814078041316,
+      "grad_norm": 2.515487334192462,
+      "learning_rate": 1.95367864780814e-05,
+      "loss": 0.8214,
+      "step": 1626
+    },
+    {
+      "epoch": 0.24896710022953328,
+      "grad_norm": 2.48662376579883,
+      "learning_rate": 1.9536040680241386e-05,
+      "loss": 0.8922,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24912012241775058,
+      "grad_norm": 2.9232465851956424,
+      "learning_rate": 1.953529429675581e-05,
+      "loss": 0.9557,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24927314460596786,
+      "grad_norm": 2.6749171268509278,
+      "learning_rate": 1.9534547327670507e-05,
+      "loss": 0.8247,
+      "step": 1629
+    },
+    {
+      "epoch": 0.24942616679418517,
+      "grad_norm": 2.4601486076921932,
+      "learning_rate": 1.9533799773031355e-05,
+      "loss": 0.8779,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24957918898240244,
+      "grad_norm": 2.24378715386169,
+      "learning_rate": 1.9533051632884262e-05,
+      "loss": 0.7781,
+      "step": 1631
+    },
+    {
+      "epoch": 0.24973221117061975,
+      "grad_norm": 2.2968548749052977,
+      "learning_rate": 1.9532302907275172e-05,
+      "loss": 0.8105,
+      "step": 1632
+    },
+    {
+      "epoch": 0.24988523335883703,
+      "grad_norm": 2.5467470529707246,
+      "learning_rate": 1.9531553596250076e-05,
+      "loss": 0.8232,
+      "step": 1633
+    },
+    {
+      "epoch": 0.25003825554705433,
+      "grad_norm": 2.7549851558941967,
+      "learning_rate": 1.953080369985498e-05,
+      "loss": 0.866,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2501912777352716,
+      "grad_norm": 2.5660121185422793,
+      "learning_rate": 1.9530053218135947e-05,
+      "loss": 0.8282,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2503442999234889,
+      "grad_norm": 2.8719794869613264,
+      "learning_rate": 1.9529302151139062e-05,
+      "loss": 0.9236,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2504973221117062,
+      "grad_norm": 2.708214993200167,
+      "learning_rate": 1.9528550498910454e-05,
+      "loss": 0.8374,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2506503442999235,
+      "grad_norm": 2.604499565659554,
+      "learning_rate": 1.9527798261496283e-05,
+      "loss": 0.9509,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2508033664881408,
+      "grad_norm": 2.4786457903057473,
+      "learning_rate": 1.952704543894275e-05,
+      "loss": 0.7756,
+      "step": 1639
+    },
+    {
+      "epoch": 0.25095638867635806,
+      "grad_norm": 2.6065249288861603,
+      "learning_rate": 1.9526292031296086e-05,
+      "loss": 0.8183,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2511094108645754,
+      "grad_norm": 2.644774458166758,
+      "learning_rate": 1.9525538038602563e-05,
+      "loss": 0.8495,
+      "step": 1641
+    },
+    {
+      "epoch": 0.25126243305279267,
+      "grad_norm": 2.7492322691250735,
+      "learning_rate": 1.9524783460908482e-05,
+      "loss": 0.8283,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25141545524100994,
+      "grad_norm": 2.3011963876888757,
+      "learning_rate": 1.952402829826019e-05,
+      "loss": 0.7487,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2515684774292272,
+      "grad_norm": 2.885621875830341,
+      "learning_rate": 1.952327255070406e-05,
+      "loss": 0.86,
+      "step": 1644
+    },
+    {
+      "epoch": 0.25172149961744456,
+      "grad_norm": 2.4666631770259895,
+      "learning_rate": 1.9522516218286508e-05,
+      "loss": 0.7159,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25187452180566183,
+      "grad_norm": 2.3549607097582608,
+      "learning_rate": 1.9521759301053983e-05,
+      "loss": 0.8202,
+      "step": 1646
+    },
+    {
+      "epoch": 0.2520275439938791,
+      "grad_norm": 2.4507380115808357,
+      "learning_rate": 1.9521001799052967e-05,
+      "loss": 0.8397,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2521805661820964,
+      "grad_norm": 2.5911841336571433,
+      "learning_rate": 1.952024371232999e-05,
+      "loss": 0.9088,
+      "step": 1648
+    },
+    {
+      "epoch": 0.2523335883703137,
+      "grad_norm": 2.6750521928124256,
+      "learning_rate": 1.9519485040931597e-05,
+      "loss": 0.793,
+      "step": 1649
+    },
+    {
+      "epoch": 0.252486610558531,
+      "grad_norm": 2.7834755569688503,
+      "learning_rate": 1.951872578490439e-05,
+      "loss": 0.8295,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2526396327467483,
+      "grad_norm": 2.237899404312587,
+      "learning_rate": 1.9517965944295e-05,
+      "loss": 0.8427,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25279265493496555,
+      "grad_norm": 2.4852949812861906,
+      "learning_rate": 1.951720551915008e-05,
+      "loss": 0.8564,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2529456771231829,
+      "grad_norm": 2.632829816927367,
+      "learning_rate": 1.951644450951634e-05,
+      "loss": 0.7319,
+      "step": 1653
+    },
+    {
+      "epoch": 0.25309869931140017,
+      "grad_norm": 2.40826639657792,
+      "learning_rate": 1.951568291544051e-05,
+      "loss": 0.8,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25325172149961744,
+      "grad_norm": 2.3396314856198694,
+      "learning_rate": 1.9514920736969374e-05,
+      "loss": 0.8018,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2534047436878347,
+      "grad_norm": 2.4346994936885356,
+      "learning_rate": 1.951415797414973e-05,
+      "loss": 0.8905,
+      "step": 1656
+    },
+    {
+      "epoch": 0.25355776587605205,
+      "grad_norm": 2.644073064070106,
+      "learning_rate": 1.9513394627028423e-05,
+      "loss": 0.9219,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25371078806426933,
+      "grad_norm": 2.4464354186477433,
+      "learning_rate": 1.9512630695652336e-05,
+      "loss": 0.7972,
+      "step": 1658
+    },
+    {
+      "epoch": 0.2538638102524866,
+      "grad_norm": 2.590677816544767,
+      "learning_rate": 1.9511866180068385e-05,
+      "loss": 0.7883,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2540168324407039,
+      "grad_norm": 2.361634237448809,
+      "learning_rate": 1.9511101080323524e-05,
+      "loss": 0.7745,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2541698546289212,
+      "grad_norm": 2.577411232973206,
+      "learning_rate": 1.9510335396464736e-05,
+      "loss": 0.867,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2543228768171385,
+      "grad_norm": 2.9471633774424477,
+      "learning_rate": 1.9509569128539048e-05,
+      "loss": 0.902,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2544758990053558,
+      "grad_norm": 2.3233960734786345,
+      "learning_rate": 1.9508802276593514e-05,
+      "loss": 0.7731,
+      "step": 1663
+    },
+    {
+      "epoch": 0.25462892119357305,
+      "grad_norm": 3.0664701594842296,
+      "learning_rate": 1.9508034840675236e-05,
+      "loss": 0.8958,
+      "step": 1664
+    },
+    {
+      "epoch": 0.25478194338179033,
+      "grad_norm": 2.7694989075868914,
+      "learning_rate": 1.9507266820831344e-05,
+      "loss": 0.8821,
+      "step": 1665
+    },
+    {
+      "epoch": 0.25493496557000767,
+      "grad_norm": 2.4964849392361006,
+      "learning_rate": 1.9506498217109003e-05,
+      "loss": 0.8569,
+      "step": 1666
+    },
+    {
+      "epoch": 0.25508798775822494,
+      "grad_norm": 2.5239295777251534,
+      "learning_rate": 1.9505729029555417e-05,
+      "loss": 0.9162,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2552410099464422,
+      "grad_norm": 2.5052913112386426,
+      "learning_rate": 1.9504959258217823e-05,
+      "loss": 0.922,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2553940321346595,
+      "grad_norm": 2.36279363557302,
+      "learning_rate": 1.95041889031435e-05,
+      "loss": 0.8044,
+      "step": 1669
+    },
+    {
+      "epoch": 0.25554705432287683,
+      "grad_norm": 2.349428630092639,
+      "learning_rate": 1.9503417964379754e-05,
+      "loss": 0.8111,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2557000765110941,
+      "grad_norm": 2.4173485389188247,
+      "learning_rate": 1.9502646441973933e-05,
+      "loss": 0.8887,
+      "step": 1671
+    },
+    {
+      "epoch": 0.2558530986993114,
+      "grad_norm": 2.3382379947820344,
+      "learning_rate": 1.9501874335973422e-05,
+      "loss": 0.848,
+      "step": 1672
+    },
+    {
+      "epoch": 0.25600612088752867,
+      "grad_norm": 2.610146018950607,
+      "learning_rate": 1.9501101646425633e-05,
+      "loss": 0.8671,
+      "step": 1673
+    },
+    {
+      "epoch": 0.256159143075746,
+      "grad_norm": 2.4976105687857166,
+      "learning_rate": 1.9500328373378026e-05,
+      "loss": 0.79,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2563121652639633,
+      "grad_norm": 2.4734090750179103,
+      "learning_rate": 1.9499554516878088e-05,
+      "loss": 0.8003,
+      "step": 1675
+    },
+    {
+      "epoch": 0.25646518745218055,
+      "grad_norm": 2.481384562214497,
+      "learning_rate": 1.949878007697334e-05,
+      "loss": 0.8418,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25661820964039783,
+      "grad_norm": 2.0977882078690824,
+      "learning_rate": 1.9498005053711354e-05,
+      "loss": 0.819,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25677123182861517,
+      "grad_norm": 2.421363761561868,
+      "learning_rate": 1.9497229447139717e-05,
+      "loss": 0.9224,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25692425401683244,
+      "grad_norm": 2.5633269767032894,
+      "learning_rate": 1.949645325730607e-05,
+      "loss": 0.849,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2570772762050497,
+      "grad_norm": 2.8582041097741344,
+      "learning_rate": 1.949567648425808e-05,
+      "loss": 0.8003,
+      "step": 1680
+    },
+    {
+      "epoch": 0.257230298393267,
+      "grad_norm": 2.6758874575609957,
+      "learning_rate": 1.9494899128043442e-05,
+      "loss": 0.8408,
+      "step": 1681
+    },
+    {
+      "epoch": 0.25738332058148433,
+      "grad_norm": 2.5888294107853147,
+      "learning_rate": 1.949412118870991e-05,
+      "loss": 0.7712,
+      "step": 1682
+    },
+    {
+      "epoch": 0.2575363427697016,
+      "grad_norm": 2.485349688173273,
+      "learning_rate": 1.9493342666305254e-05,
+      "loss": 0.7898,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2576893649579189,
+      "grad_norm": 2.4074664249571707,
+      "learning_rate": 1.9492563560877285e-05,
+      "loss": 0.9074,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25784238714613616,
+      "grad_norm": 3.006519080206942,
+      "learning_rate": 1.949178387247385e-05,
+      "loss": 0.9606,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2579954093343535,
+      "grad_norm": 2.9344864762687353,
+      "learning_rate": 1.9491003601142842e-05,
+      "loss": 0.7411,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2581484315225708,
+      "grad_norm": 2.160570396212931,
+      "learning_rate": 1.949022274693217e-05,
+      "loss": 0.6894,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25830145371078805,
+      "grad_norm": 2.438882169334196,
+      "learning_rate": 1.9489441309889794e-05,
+      "loss": 0.7715,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25845447589900533,
+      "grad_norm": 2.436285424122042,
+      "learning_rate": 1.9488659290063702e-05,
+      "loss": 0.7852,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25860749808722266,
+      "grad_norm": 2.854857000650002,
+      "learning_rate": 1.9487876687501926e-05,
+      "loss": 0.8264,
+      "step": 1690
+    },
+    {
+      "epoch": 0.25876052027543994,
+      "grad_norm": 2.271191622502827,
+      "learning_rate": 1.9487093502252526e-05,
+      "loss": 0.7694,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2589135424636572,
+      "grad_norm": 2.489235162164264,
+      "learning_rate": 1.94863097343636e-05,
+      "loss": 0.7871,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2590665646518745,
+      "grad_norm": 3.0714855530371015,
+      "learning_rate": 1.948552538388328e-05,
+      "loss": 0.8953,
+      "step": 1693
+    },
+    {
+      "epoch": 0.25921958684009183,
+      "grad_norm": 2.5084220890499007,
+      "learning_rate": 1.9484740450859743e-05,
+      "loss": 0.8391,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2593726090283091,
+      "grad_norm": 2.5910117597918116,
+      "learning_rate": 1.9483954935341186e-05,
+      "loss": 0.8472,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2595256312165264,
+      "grad_norm": 2.4874414300183982,
+      "learning_rate": 1.9483168837375856e-05,
+      "loss": 0.8618,
+      "step": 1696
+    },
+    {
+      "epoch": 0.25967865340474366,
+      "grad_norm": 2.4390882665604536,
+      "learning_rate": 1.9482382157012033e-05,
+      "loss": 0.8633,
+      "step": 1697
+    },
+    {
+      "epoch": 0.259831675592961,
+      "grad_norm": 2.7537562928055355,
+      "learning_rate": 1.9481594894298023e-05,
+      "loss": 0.8594,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2599846977811783,
+      "grad_norm": 2.3613754614350078,
+      "learning_rate": 1.9480807049282177e-05,
+      "loss": 0.8136,
+      "step": 1699
+    },
+    {
+      "epoch": 0.26013771996939555,
+      "grad_norm": 2.611634039266382,
+      "learning_rate": 1.9480018622012884e-05,
+      "loss": 0.9738,
+      "step": 1700
+    },
+    {
+      "epoch": 0.26029074215761283,
+      "grad_norm": 2.4645726484909,
+      "learning_rate": 1.9479229612538558e-05,
+      "loss": 0.8931,
+      "step": 1701
+    },
+    {
+      "epoch": 0.26044376434583016,
+      "grad_norm": 2.4373875784863284,
+      "learning_rate": 1.9478440020907662e-05,
+      "loss": 0.8618,
+      "step": 1702
+    },
+    {
+      "epoch": 0.26059678653404744,
+      "grad_norm": 2.642912343584658,
+      "learning_rate": 1.9477649847168685e-05,
+      "loss": 0.8661,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2607498087222647,
+      "grad_norm": 2.433110527065469,
+      "learning_rate": 1.9476859091370153e-05,
+      "loss": 0.7962,
+      "step": 1704
+    },
+    {
+      "epoch": 0.260902830910482,
+      "grad_norm": 2.5788838749611966,
+      "learning_rate": 1.947606775356063e-05,
+      "loss": 0.9376,
+      "step": 1705
+    },
+    {
+      "epoch": 0.26105585309869933,
+      "grad_norm": 2.3973604286815458,
+      "learning_rate": 1.9475275833788714e-05,
+      "loss": 0.8217,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2612088752869166,
+      "grad_norm": 2.669068275728912,
+      "learning_rate": 1.9474483332103043e-05,
+      "loss": 0.8883,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2613618974751339,
+      "grad_norm": 2.5071451612106883,
+      "learning_rate": 1.947369024855229e-05,
+      "loss": 0.8878,
+      "step": 1708
+    },
+    {
+      "epoch": 0.26151491966335116,
+      "grad_norm": 2.2570325349738556,
+      "learning_rate": 1.947289658318515e-05,
+      "loss": 0.8426,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2616679418515685,
+      "grad_norm": 2.5071883489100015,
+      "learning_rate": 1.947210233605038e-05,
+      "loss": 0.8524,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2618209640397858,
+      "grad_norm": 2.4286658527705343,
+      "learning_rate": 1.9471307507196746e-05,
+      "loss": 0.9167,
+      "step": 1711
+    },
+    {
+      "epoch": 0.26197398622800305,
+      "grad_norm": 2.5819637026712683,
+      "learning_rate": 1.9470512096673065e-05,
+      "loss": 0.7922,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26212700841622033,
+      "grad_norm": 2.4051224205801027,
+      "learning_rate": 1.946971610452819e-05,
+      "loss": 0.7694,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26228003060443766,
+      "grad_norm": 2.5521899267184667,
+      "learning_rate": 1.9468919530811002e-05,
+      "loss": 0.847,
+      "step": 1714
+    },
+    {
+      "epoch": 0.26243305279265494,
+      "grad_norm": 2.4020801572092516,
+      "learning_rate": 1.9468122375570425e-05,
+      "loss": 0.8737,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2625860749808722,
+      "grad_norm": 2.3615195314704582,
+      "learning_rate": 1.946732463885541e-05,
+      "loss": 0.7862,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2627390971690895,
+      "grad_norm": 2.5988427829601415,
+      "learning_rate": 1.9466526320714956e-05,
+      "loss": 0.8121,
+      "step": 1717
+    },
+    {
+      "epoch": 0.26289211935730683,
+      "grad_norm": 2.720083299441531,
+      "learning_rate": 1.9465727421198086e-05,
+      "loss": 0.9416,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2630451415455241,
+      "grad_norm": 2.3491570168515534,
+      "learning_rate": 1.9464927940353865e-05,
+      "loss": 0.8364,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2631981637337414,
+      "grad_norm": 2.3320233151056846,
+      "learning_rate": 1.9464127878231393e-05,
+      "loss": 0.9151,
+      "step": 1720
+    },
+    {
+      "epoch": 0.26335118592195866,
+      "grad_norm": 2.4609771161448064,
+      "learning_rate": 1.9463327234879805e-05,
+      "loss": 0.7901,
+      "step": 1721
+    },
+    {
+      "epoch": 0.263504208110176,
+      "grad_norm": 2.576087752470663,
+      "learning_rate": 1.946252601034827e-05,
+      "loss": 0.9416,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2636572302983933,
+      "grad_norm": 2.746683862451466,
+      "learning_rate": 1.946172420468599e-05,
+      "loss": 0.9327,
+      "step": 1723
+    },
+    {
+      "epoch": 0.26381025248661055,
+      "grad_norm": 2.544995404935606,
+      "learning_rate": 1.9460921817942217e-05,
+      "loss": 0.7567,
+      "step": 1724
+    },
+    {
+      "epoch": 0.26396327467482783,
+      "grad_norm": 2.2425437074057832,
+      "learning_rate": 1.9460118850166223e-05,
+      "loss": 0.6886,
+      "step": 1725
+    },
+    {
+      "epoch": 0.26411629686304516,
+      "grad_norm": 2.4009536341104036,
+      "learning_rate": 1.9459315301407323e-05,
+      "loss": 0.9018,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26426931905126244,
+      "grad_norm": 2.610099232671235,
+      "learning_rate": 1.9458511171714863e-05,
+      "loss": 0.8207,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2644223412394797,
+      "grad_norm": 2.4320257408549635,
+      "learning_rate": 1.9457706461138233e-05,
+      "loss": 0.7446,
+      "step": 1728
+    },
+    {
+      "epoch": 0.264575363427697,
+      "grad_norm": 2.5185379316749548,
+      "learning_rate": 1.945690116972685e-05,
+      "loss": 0.7479,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26472838561591433,
+      "grad_norm": 2.895728792841351,
+      "learning_rate": 1.945609529753017e-05,
+      "loss": 0.8769,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2648814078041316,
+      "grad_norm": 2.730992822149975,
+      "learning_rate": 1.945528884459768e-05,
+      "loss": 0.916,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2650344299923489,
+      "grad_norm": 2.693860601738405,
+      "learning_rate": 1.945448181097892e-05,
+      "loss": 0.8514,
+      "step": 1732
+    },
+    {
+      "epoch": 0.26518745218056616,
+      "grad_norm": 2.6405405825364494,
+      "learning_rate": 1.9453674196723445e-05,
+      "loss": 0.9356,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2653404743687835,
+      "grad_norm": 2.5688295938831693,
+      "learning_rate": 1.9452866001880852e-05,
+      "loss": 0.9191,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2654934965570008,
+      "grad_norm": 2.499877747605039,
+      "learning_rate": 1.945205722650078e-05,
+      "loss": 0.888,
+      "step": 1735
+    },
+    {
+      "epoch": 0.26564651874521805,
+      "grad_norm": 2.922492979194366,
+      "learning_rate": 1.9451247870632898e-05,
+      "loss": 0.8968,
+      "step": 1736
+    },
+    {
+      "epoch": 0.26579954093343533,
+      "grad_norm": 2.5466573461677076,
+      "learning_rate": 1.9450437934326906e-05,
+      "loss": 0.7677,
+      "step": 1737
+    },
+    {
+      "epoch": 0.26595256312165266,
+      "grad_norm": 2.5820858780258753,
+      "learning_rate": 1.9449627417632554e-05,
+      "loss": 0.8154,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26610558530986994,
+      "grad_norm": 2.2852352856657903,
+      "learning_rate": 1.9448816320599615e-05,
+      "loss": 0.8817,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2662586074980872,
+      "grad_norm": 2.473054849169737,
+      "learning_rate": 1.94480046432779e-05,
+      "loss": 0.8555,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2664116296863045,
+      "grad_norm": 2.761146652490966,
+      "learning_rate": 1.944719238571726e-05,
+      "loss": 0.8623,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26656465187452183,
+      "grad_norm": 2.687345946730486,
+      "learning_rate": 1.944637954796758e-05,
+      "loss": 0.7884,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2667176740627391,
+      "grad_norm": 2.6170414834070415,
+      "learning_rate": 1.9445566130078774e-05,
+      "loss": 0.823,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2668706962509564,
+      "grad_norm": 2.5065210386666315,
+      "learning_rate": 1.94447521321008e-05,
+      "loss": 0.8013,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26702371843917366,
+      "grad_norm": 2.569502254268044,
+      "learning_rate": 1.9443937554083655e-05,
+      "loss": 0.9016,
+      "step": 1745
+    },
+    {
+      "epoch": 0.267176740627391,
+      "grad_norm": 2.318313240796993,
+      "learning_rate": 1.9443122396077357e-05,
+      "loss": 0.8254,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2673297628156083,
+      "grad_norm": 2.3525786375234063,
+      "learning_rate": 1.9442306658131967e-05,
+      "loss": 0.7866,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26748278500382555,
+      "grad_norm": 2.558329696352076,
+      "learning_rate": 1.944149034029759e-05,
+      "loss": 0.8594,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26763580719204283,
+      "grad_norm": 2.285331779879835,
+      "learning_rate": 1.944067344262436e-05,
+      "loss": 0.7182,
+      "step": 1749
+    },
+    {
+      "epoch": 0.26778882938026016,
+      "grad_norm": 2.607640802866723,
+      "learning_rate": 1.9439855965162436e-05,
+      "loss": 0.7751,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26794185156847744,
+      "grad_norm": 2.412052455367587,
+      "learning_rate": 1.9439037907962032e-05,
+      "loss": 0.8451,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2680948737566947,
+      "grad_norm": 2.6293034239554456,
+      "learning_rate": 1.943821927107338e-05,
+      "loss": 0.9424,
+      "step": 1752
+    },
+    {
+      "epoch": 0.268247895944912,
+      "grad_norm": 2.6107845855461695,
+      "learning_rate": 1.9437400054546765e-05,
+      "loss": 0.8322,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26840091813312933,
+      "grad_norm": 2.548893203671856,
+      "learning_rate": 1.9436580258432488e-05,
+      "loss": 0.7892,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2685539403213466,
+      "grad_norm": 3.0202000676385783,
+      "learning_rate": 1.9435759882780905e-05,
+      "loss": 0.9641,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2687069625095639,
+      "grad_norm": 2.5840683936343467,
+      "learning_rate": 1.9434938927642393e-05,
+      "loss": 0.8314,
+      "step": 1756
+    },
+    {
+      "epoch": 0.26885998469778116,
+      "grad_norm": 2.754920380157699,
+      "learning_rate": 1.9434117393067375e-05,
+      "loss": 0.9488,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2690130068859985,
+      "grad_norm": 2.48864745158179,
+      "learning_rate": 1.9433295279106296e-05,
+      "loss": 0.8753,
+      "step": 1758
+    },
+    {
+      "epoch": 0.2691660290742158,
+      "grad_norm": 2.3149579588067573,
+      "learning_rate": 1.9432472585809657e-05,
+      "loss": 0.8767,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26931905126243305,
+      "grad_norm": 2.6859581863399087,
+      "learning_rate": 1.9431649313227972e-05,
+      "loss": 0.8624,
+      "step": 1760
+    },
+    {
+      "epoch": 0.26947207345065033,
+      "grad_norm": 2.2212850321280593,
+      "learning_rate": 1.9430825461411805e-05,
+      "loss": 0.761,
+      "step": 1761
+    },
+    {
+      "epoch": 0.26962509563886766,
+      "grad_norm": 2.4027565272940654,
+      "learning_rate": 1.9430001030411757e-05,
+      "loss": 0.8739,
+      "step": 1762
+    },
+    {
+      "epoch": 0.26977811782708494,
+      "grad_norm": 2.660172589499753,
+      "learning_rate": 1.942917602027845e-05,
+      "loss": 0.9251,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2699311400153022,
+      "grad_norm": 2.743206963419242,
+      "learning_rate": 1.942835043106256e-05,
+      "loss": 0.9153,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2700841622035195,
+      "grad_norm": 2.6989866804235145,
+      "learning_rate": 1.9427524262814786e-05,
+      "loss": 0.9009,
+      "step": 1765
+    },
+    {
+      "epoch": 0.27023718439173683,
+      "grad_norm": 2.487307962781265,
+      "learning_rate": 1.9426697515585865e-05,
+      "loss": 0.7067,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2703902065799541,
+      "grad_norm": 2.4830826951639353,
+      "learning_rate": 1.9425870189426573e-05,
+      "loss": 0.8055,
+      "step": 1767
+    },
+    {
+      "epoch": 0.2705432287681714,
+      "grad_norm": 2.5304574169611134,
+      "learning_rate": 1.942504228438772e-05,
+      "loss": 0.766,
+      "step": 1768
+    },
+    {
+      "epoch": 0.27069625095638866,
+      "grad_norm": 2.328713565548941,
+      "learning_rate": 1.9424213800520147e-05,
+      "loss": 0.7621,
+      "step": 1769
+    },
+    {
+      "epoch": 0.270849273144606,
+      "grad_norm": 2.6578397488472723,
+      "learning_rate": 1.9423384737874738e-05,
+      "loss": 0.8997,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2710022953328233,
+      "grad_norm": 2.406319383474901,
+      "learning_rate": 1.9422555096502406e-05,
+      "loss": 0.6949,
+      "step": 1771
+    },
+    {
+      "epoch": 0.27115531752104055,
+      "grad_norm": 2.320190474995434,
+      "learning_rate": 1.9421724876454108e-05,
+      "loss": 0.7765,
+      "step": 1772
+    },
+    {
+      "epoch": 0.27130833970925783,
+      "grad_norm": 2.255508796586933,
+      "learning_rate": 1.9420894077780826e-05,
+      "loss": 0.7989,
+      "step": 1773
+    },
+    {
+      "epoch": 0.2714613618974751,
+      "grad_norm": 2.4808465555872967,
+      "learning_rate": 1.942006270053358e-05,
+      "loss": 0.7873,
+      "step": 1774
+    },
+    {
+      "epoch": 0.27161438408569244,
+      "grad_norm": 2.6957590465125505,
+      "learning_rate": 1.9419230744763437e-05,
+      "loss": 0.9488,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2717674062739097,
+      "grad_norm": 2.6825223214394205,
+      "learning_rate": 1.9418398210521486e-05,
+      "loss": 0.8551,
+      "step": 1776
+    },
+    {
+      "epoch": 0.271920428462127,
+      "grad_norm": 2.319812276713523,
+      "learning_rate": 1.941756509785885e-05,
+      "loss": 0.8255,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2720734506503443,
+      "grad_norm": 2.736308455119988,
+      "learning_rate": 1.9416731406826704e-05,
+      "loss": 0.784,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2722264728385616,
+      "grad_norm": 2.623518511491986,
+      "learning_rate": 1.9415897137476243e-05,
+      "loss": 0.8191,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2723794950267789,
+      "grad_norm": 2.3011853266463924,
+      "learning_rate": 1.9415062289858702e-05,
+      "loss": 0.7144,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27253251721499616,
+      "grad_norm": 2.820252509123847,
+      "learning_rate": 1.941422686402536e-05,
+      "loss": 0.8803,
+      "step": 1781
+    },
+    {
+      "epoch": 0.27268553940321344,
+      "grad_norm": 2.3683267612707195,
+      "learning_rate": 1.9413390860027512e-05,
+      "loss": 0.7924,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2728385615914308,
+      "grad_norm": 2.489890268313438,
+      "learning_rate": 1.9412554277916506e-05,
+      "loss": 0.8228,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27299158377964805,
+      "grad_norm": 2.5950610129814966,
+      "learning_rate": 1.941171711774372e-05,
+      "loss": 0.8059,
+      "step": 1784
+    },
+    {
+      "epoch": 0.27314460596786533,
+      "grad_norm": 2.4309315036464603,
+      "learning_rate": 1.941087937956057e-05,
+      "loss": 0.8338,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2732976281560826,
+      "grad_norm": 2.4619004282248484,
+      "learning_rate": 1.94100410634185e-05,
+      "loss": 0.775,
+      "step": 1786
+    },
+    {
+      "epoch": 0.27345065034429994,
+      "grad_norm": 2.553665078049513,
+      "learning_rate": 1.9409202169368994e-05,
+      "loss": 0.7981,
+      "step": 1787
+    },
+    {
+      "epoch": 0.2736036725325172,
+      "grad_norm": 2.7012115848050375,
+      "learning_rate": 1.9408362697463576e-05,
+      "loss": 0.7708,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2737566947207345,
+      "grad_norm": 2.548331990200666,
+      "learning_rate": 1.94075226477538e-05,
+      "loss": 0.8804,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2739097169089518,
+      "grad_norm": 2.348960462028775,
+      "learning_rate": 1.9406682020291253e-05,
+      "loss": 0.7993,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2740627390971691,
+      "grad_norm": 2.4769272628241517,
+      "learning_rate": 1.9405840815127567e-05,
+      "loss": 0.8696,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2742157612853864,
+      "grad_norm": 2.659296349245691,
+      "learning_rate": 1.9404999032314397e-05,
+      "loss": 0.8705,
+      "step": 1792
+    },
+    {
+      "epoch": 0.27436878347360366,
+      "grad_norm": 2.5496822689710177,
+      "learning_rate": 1.9404156671903443e-05,
+      "loss": 0.7476,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27452180566182094,
+      "grad_norm": 2.1802122613902726,
+      "learning_rate": 1.9403313733946442e-05,
+      "loss": 0.7877,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2746748278500383,
+      "grad_norm": 2.636214632414026,
+      "learning_rate": 1.9402470218495158e-05,
+      "loss": 0.8912,
+      "step": 1795
+    },
+    {
+      "epoch": 0.27482785003825555,
+      "grad_norm": 2.5169431177436996,
+      "learning_rate": 1.9401626125601395e-05,
+      "loss": 0.7936,
+      "step": 1796
+    },
+    {
+      "epoch": 0.27498087222647283,
+      "grad_norm": 2.6274651789471486,
+      "learning_rate": 1.940078145531699e-05,
+      "loss": 0.8681,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2751338944146901,
+      "grad_norm": 2.6433366981390223,
+      "learning_rate": 1.9399936207693826e-05,
+      "loss": 0.7657,
+      "step": 1798
+    },
+    {
+      "epoch": 0.27528691660290744,
+      "grad_norm": 2.3588536871322914,
+      "learning_rate": 1.9399090382783802e-05,
+      "loss": 0.708,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2754399387911247,
+      "grad_norm": 2.6128210317047076,
+      "learning_rate": 1.9398243980638867e-05,
+      "loss": 0.8084,
+      "step": 1800
+    },
+    {
+      "epoch": 0.275592960979342,
+      "grad_norm": 2.384807803529249,
+      "learning_rate": 1.9397397001311007e-05,
+      "loss": 0.8958,
+      "step": 1801
+    },
+    {
+      "epoch": 0.27574598316755927,
+      "grad_norm": 2.561729566518759,
+      "learning_rate": 1.939654944485223e-05,
+      "loss": 0.8511,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2758990053557766,
+      "grad_norm": 2.409093911820784,
+      "learning_rate": 1.9395701311314594e-05,
+      "loss": 0.7611,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2760520275439939,
+      "grad_norm": 2.369205457750817,
+      "learning_rate": 1.9394852600750184e-05,
+      "loss": 0.7975,
+      "step": 1804
+    },
+    {
+      "epoch": 0.27620504973221116,
+      "grad_norm": 2.501377346696639,
+      "learning_rate": 1.9394003313211126e-05,
+      "loss": 0.7295,
+      "step": 1805
+    },
+    {
+      "epoch": 0.27635807192042844,
+      "grad_norm": 2.419062632867766,
+      "learning_rate": 1.9393153448749572e-05,
+      "loss": 0.8551,
+      "step": 1806
+    },
+    {
+      "epoch": 0.27651109410864577,
+      "grad_norm": 2.4102900990912928,
+      "learning_rate": 1.9392303007417717e-05,
+      "loss": 0.9076,
+      "step": 1807
+    },
+    {
+      "epoch": 0.27666411629686305,
+      "grad_norm": 2.2795357293501457,
+      "learning_rate": 1.9391451989267795e-05,
+      "loss": 0.7942,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2768171384850803,
+      "grad_norm": 2.650801351754482,
+      "learning_rate": 1.9390600394352066e-05,
+      "loss": 0.8162,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2769701606732976,
+      "grad_norm": 2.389261821947559,
+      "learning_rate": 1.9389748222722827e-05,
+      "loss": 0.7481,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27712318286151494,
+      "grad_norm": 2.1568992212804363,
+      "learning_rate": 1.938889547443242e-05,
+      "loss": 0.7289,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2772762050497322,
+      "grad_norm": 2.5354031323568242,
+      "learning_rate": 1.9388042149533214e-05,
+      "loss": 0.8771,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2774292272379495,
+      "grad_norm": 2.7834820672844067,
+      "learning_rate": 1.938718824807761e-05,
+      "loss": 0.8314,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27758224942616677,
+      "grad_norm": 2.342286831367304,
+      "learning_rate": 1.9386333770118054e-05,
+      "loss": 0.7752,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2777352716143841,
+      "grad_norm": 2.431265873833426,
+      "learning_rate": 1.9385478715707024e-05,
+      "loss": 0.8154,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2778882938026014,
+      "grad_norm": 3.472642793672234,
+      "learning_rate": 1.9384623084897025e-05,
+      "loss": 0.8133,
+      "step": 1816
+    },
+    {
+      "epoch": 0.27804131599081866,
+      "grad_norm": 2.4898768514747287,
+      "learning_rate": 1.938376687774061e-05,
+      "loss": 0.9039,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27819433817903594,
+      "grad_norm": 2.453143034799608,
+      "learning_rate": 1.9382910094290367e-05,
+      "loss": 0.7147,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27834736036725327,
+      "grad_norm": 2.3760737629972275,
+      "learning_rate": 1.9382052734598902e-05,
+      "loss": 0.8291,
+      "step": 1819
+    },
+    {
+      "epoch": 0.27850038255547055,
+      "grad_norm": 2.9747112957484365,
+      "learning_rate": 1.938119479871888e-05,
+      "loss": 0.8505,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2786534047436878,
+      "grad_norm": 2.6897445660961323,
+      "learning_rate": 1.9380336286702987e-05,
+      "loss": 0.8376,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2788064269319051,
+      "grad_norm": 2.9716563210946103,
+      "learning_rate": 1.9379477198603944e-05,
+      "loss": 0.9802,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27895944912012244,
+      "grad_norm": 2.508289472632622,
+      "learning_rate": 1.9378617534474514e-05,
+      "loss": 0.9134,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2791124713083397,
+      "grad_norm": 2.458112494711284,
+      "learning_rate": 1.937775729436749e-05,
+      "loss": 0.775,
+      "step": 1824
+    },
+    {
+      "epoch": 0.279265493496557,
+      "grad_norm": 2.334406138363872,
+      "learning_rate": 1.937689647833571e-05,
+      "loss": 0.82,
+      "step": 1825
+    },
+    {
+      "epoch": 0.27941851568477427,
+      "grad_norm": 2.4747860463869005,
+      "learning_rate": 1.937603508643203e-05,
+      "loss": 0.8692,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2795715378729916,
+      "grad_norm": 2.477096190126597,
+      "learning_rate": 1.9375173118709357e-05,
+      "loss": 0.8041,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2797245600612089,
+      "grad_norm": 2.2376490321825253,
+      "learning_rate": 1.937431057522063e-05,
+      "loss": 0.8329,
+      "step": 1828
+    },
+    {
+      "epoch": 0.27987758224942616,
+      "grad_norm": 2.7371937042935217,
+      "learning_rate": 1.9373447456018814e-05,
+      "loss": 0.8114,
+      "step": 1829
+    },
+    {
+      "epoch": 0.28003060443764344,
+      "grad_norm": 2.4625318159596223,
+      "learning_rate": 1.9372583761156924e-05,
+      "loss": 0.8698,
+      "step": 1830
+    },
+    {
+      "epoch": 0.28018362662586077,
+      "grad_norm": 2.572144753716217,
+      "learning_rate": 1.9371719490687994e-05,
+      "loss": 0.7726,
+      "step": 1831
+    },
+    {
+      "epoch": 0.28033664881407805,
+      "grad_norm": 3.1187471865773766,
+      "learning_rate": 1.9370854644665113e-05,
+      "loss": 0.907,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2804896710022953,
+      "grad_norm": 2.8014006276479577,
+      "learning_rate": 1.9369989223141386e-05,
+      "loss": 0.879,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2806426931905126,
+      "grad_norm": 2.3574224893672655,
+      "learning_rate": 1.9369123226169967e-05,
+      "loss": 0.7876,
+      "step": 1834
+    },
+    {
+      "epoch": 0.28079571537872994,
+      "grad_norm": 2.4041536999686706,
+      "learning_rate": 1.936825665380404e-05,
+      "loss": 0.7086,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2809487375669472,
+      "grad_norm": 2.629825929620783,
+      "learning_rate": 1.936738950609682e-05,
+      "loss": 0.8817,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2811017597551645,
+      "grad_norm": 2.501772945163813,
+      "learning_rate": 1.9366521783101566e-05,
+      "loss": 0.8242,
+      "step": 1837
+    },
+    {
+      "epoch": 0.28125478194338177,
+      "grad_norm": 2.5759467678625065,
+      "learning_rate": 1.9365653484871567e-05,
+      "loss": 0.8058,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2814078041315991,
+      "grad_norm": 2.506404915094637,
+      "learning_rate": 1.936478461146015e-05,
+      "loss": 0.7182,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2815608263198164,
+      "grad_norm": 2.654103256153006,
+      "learning_rate": 1.9363915162920676e-05,
+      "loss": 0.8209,
+      "step": 1840
+    },
+    {
+      "epoch": 0.28171384850803366,
+      "grad_norm": 2.352377734931323,
+      "learning_rate": 1.9363045139306536e-05,
+      "loss": 0.7696,
+      "step": 1841
+    },
+    {
+      "epoch": 0.28186687069625094,
+      "grad_norm": 2.4722586829533966,
+      "learning_rate": 1.9362174540671167e-05,
+      "loss": 0.7529,
+      "step": 1842
+    },
+    {
+      "epoch": 0.28201989288446827,
+      "grad_norm": 2.573994784803902,
+      "learning_rate": 1.9361303367068035e-05,
+      "loss": 0.7462,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28217291507268555,
+      "grad_norm": 2.2498326088485974,
+      "learning_rate": 1.9360431618550645e-05,
+      "loss": 0.7454,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2823259372609028,
+      "grad_norm": 2.289597736923765,
+      "learning_rate": 1.9359559295172525e-05,
+      "loss": 0.8085,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2824789594491201,
+      "grad_norm": 2.7762561379472954,
+      "learning_rate": 1.9358686396987256e-05,
+      "loss": 0.7927,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28263198163733744,
+      "grad_norm": 2.6661352318965044,
+      "learning_rate": 1.9357812924048445e-05,
+      "loss": 0.841,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2827850038255547,
+      "grad_norm": 2.5975812171329964,
+      "learning_rate": 1.9356938876409735e-05,
+      "loss": 0.8676,
+      "step": 1848
+    },
+    {
+      "epoch": 0.282938026013772,
+      "grad_norm": 2.125735235990031,
+      "learning_rate": 1.9356064254124803e-05,
+      "loss": 0.7212,
+      "step": 1849
+    },
+    {
+      "epoch": 0.28309104820198927,
+      "grad_norm": 2.7059435749550556,
+      "learning_rate": 1.9355189057247363e-05,
+      "loss": 0.7979,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2832440703902066,
+      "grad_norm": 2.4770619838430448,
+      "learning_rate": 1.9354313285831167e-05,
+      "loss": 0.8406,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2833970925784239,
+      "grad_norm": 2.695527673356984,
+      "learning_rate": 1.9353436939929997e-05,
+      "loss": 0.7533,
+      "step": 1852
+    },
+    {
+      "epoch": 0.28355011476664116,
+      "grad_norm": 3.0149128657035362,
+      "learning_rate": 1.9352560019597675e-05,
+      "loss": 0.8231,
+      "step": 1853
+    },
+    {
+      "epoch": 0.28370313695485844,
+      "grad_norm": 2.6130169564784893,
+      "learning_rate": 1.9351682524888052e-05,
+      "loss": 0.7279,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28385615914307577,
+      "grad_norm": 2.2246128982518347,
+      "learning_rate": 1.935080445585502e-05,
+      "loss": 0.8827,
+      "step": 1855
+    },
+    {
+      "epoch": 0.28400918133129305,
+      "grad_norm": 2.21223342955157,
+      "learning_rate": 1.934992581255251e-05,
+      "loss": 0.727,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2841622035195103,
+      "grad_norm": 2.579774600746488,
+      "learning_rate": 1.934904659503448e-05,
+      "loss": 0.8197,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2843152257077276,
+      "grad_norm": 2.485921849818407,
+      "learning_rate": 1.9348166803354923e-05,
+      "loss": 0.77,
+      "step": 1858
+    },
+    {
+      "epoch": 0.28446824789594494,
+      "grad_norm": 2.407776330565242,
+      "learning_rate": 1.9347286437567868e-05,
+      "loss": 0.8485,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2846212700841622,
+      "grad_norm": 2.462948009991273,
+      "learning_rate": 1.934640549772739e-05,
+      "loss": 0.748,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2847742922723795,
+      "grad_norm": 2.5536317903542,
+      "learning_rate": 1.9345523983887585e-05,
+      "loss": 0.8609,
+      "step": 1861
+    },
+    {
+      "epoch": 0.28492731446059677,
+      "grad_norm": 2.4342429583754637,
+      "learning_rate": 1.9344641896102596e-05,
+      "loss": 0.8274,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2850803366488141,
+      "grad_norm": 2.629328839248284,
+      "learning_rate": 1.934375923442659e-05,
+      "loss": 0.7806,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2852333588370314,
+      "grad_norm": 2.825115788194663,
+      "learning_rate": 1.9342875998913774e-05,
+      "loss": 0.843,
+      "step": 1864
+    },
+    {
+      "epoch": 0.28538638102524866,
+      "grad_norm": 2.432079122950358,
+      "learning_rate": 1.93419921896184e-05,
+      "loss": 0.6456,
+      "step": 1865
+    },
+    {
+      "epoch": 0.28553940321346594,
+      "grad_norm": 2.437685754761619,
+      "learning_rate": 1.9341107806594733e-05,
+      "loss": 0.8309,
+      "step": 1866
+    },
+    {
+      "epoch": 0.28569242540168327,
+      "grad_norm": 2.551675385381315,
+      "learning_rate": 1.9340222849897096e-05,
+      "loss": 0.8865,
+      "step": 1867
+    },
+    {
+      "epoch": 0.28584544758990055,
+      "grad_norm": 2.508467686617371,
+      "learning_rate": 1.9339337319579833e-05,
+      "loss": 0.7803,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2859984697781178,
+      "grad_norm": 2.5126960561564413,
+      "learning_rate": 1.933845121569733e-05,
+      "loss": 0.7849,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2861514919663351,
+      "grad_norm": 2.613958666791803,
+      "learning_rate": 1.9337564538304004e-05,
+      "loss": 0.8853,
+      "step": 1870
+    },
+    {
+      "epoch": 0.28630451415455244,
+      "grad_norm": 2.5244937795138376,
+      "learning_rate": 1.9336677287454316e-05,
+      "loss": 0.8044,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2864575363427697,
+      "grad_norm": 2.565342492613703,
+      "learning_rate": 1.9335789463202744e-05,
+      "loss": 0.7959,
+      "step": 1872
+    },
+    {
+      "epoch": 0.286610558530987,
+      "grad_norm": 2.5403185668043773,
+      "learning_rate": 1.9334901065603823e-05,
+      "loss": 0.7555,
+      "step": 1873
+    },
+    {
+      "epoch": 0.28676358071920427,
+      "grad_norm": 2.4550949680199,
+      "learning_rate": 1.9334012094712108e-05,
+      "loss": 0.7674,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2869166029074216,
+      "grad_norm": 2.3284267657849793,
+      "learning_rate": 1.9333122550582197e-05,
+      "loss": 0.7756,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2870696250956389,
+      "grad_norm": 2.6275948038792323,
+      "learning_rate": 1.9332232433268718e-05,
+      "loss": 0.7993,
+      "step": 1876
+    },
+    {
+      "epoch": 0.28722264728385616,
+      "grad_norm": 2.568517017557821,
+      "learning_rate": 1.9331341742826337e-05,
+      "loss": 0.7721,
+      "step": 1877
+    },
+    {
+      "epoch": 0.28737566947207344,
+      "grad_norm": 2.334517381546241,
+      "learning_rate": 1.933045047930976e-05,
+      "loss": 0.8133,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28752869166029077,
+      "grad_norm": 2.3005139977871942,
+      "learning_rate": 1.932955864277371e-05,
+      "loss": 0.8128,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28768171384850805,
+      "grad_norm": 2.7207446187801017,
+      "learning_rate": 1.932866623327297e-05,
+      "loss": 0.8604,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2878347360367253,
+      "grad_norm": 2.492245625318649,
+      "learning_rate": 1.9327773250862344e-05,
+      "loss": 0.845,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2879877582249426,
+      "grad_norm": 2.467518676739865,
+      "learning_rate": 1.932687969559667e-05,
+      "loss": 0.9091,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2881407804131599,
+      "grad_norm": 2.538560762578847,
+      "learning_rate": 1.9325985567530825e-05,
+      "loss": 0.7151,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2882938026013772,
+      "grad_norm": 2.3038764537536305,
+      "learning_rate": 1.9325090866719726e-05,
+      "loss": 0.8092,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2884468247895945,
+      "grad_norm": 2.703874276713587,
+      "learning_rate": 1.9324195593218315e-05,
+      "loss": 0.7405,
+      "step": 1885
+    },
+    {
+      "epoch": 0.28859984697781177,
+      "grad_norm": 2.2950766344341047,
+      "learning_rate": 1.932329974708158e-05,
+      "loss": 0.6749,
+      "step": 1886
+    },
+    {
+      "epoch": 0.28875286916602905,
+      "grad_norm": 2.4093547511917826,
+      "learning_rate": 1.932240332836453e-05,
+      "loss": 0.9128,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2889058913542464,
+      "grad_norm": 2.743832608962814,
+      "learning_rate": 1.9321506337122224e-05,
+      "loss": 0.8467,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28905891354246366,
+      "grad_norm": 2.418044806593421,
+      "learning_rate": 1.932060877340975e-05,
+      "loss": 0.7427,
+      "step": 1889
+    },
+    {
+      "epoch": 0.28921193573068094,
+      "grad_norm": 2.4645967587420987,
+      "learning_rate": 1.9319710637282227e-05,
+      "loss": 0.8417,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2893649579188982,
+      "grad_norm": 2.6800041375280763,
+      "learning_rate": 1.9318811928794817e-05,
+      "loss": 0.8604,
+      "step": 1891
+    },
+    {
+      "epoch": 0.28951798010711555,
+      "grad_norm": 2.8142534613435033,
+      "learning_rate": 1.9317912648002708e-05,
+      "loss": 0.8685,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2896710022953328,
+      "grad_norm": 2.517363713662083,
+      "learning_rate": 1.931701279496113e-05,
+      "loss": 0.8123,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2898240244835501,
+      "grad_norm": 2.559876231147071,
+      "learning_rate": 1.9316112369725354e-05,
+      "loss": 0.8617,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2899770466717674,
+      "grad_norm": 2.247438925725813,
+      "learning_rate": 1.9315211372350667e-05,
+      "loss": 0.8346,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2901300688599847,
+      "grad_norm": 2.4545703617872365,
+      "learning_rate": 1.9314309802892407e-05,
+      "loss": 0.7889,
+      "step": 1896
+    },
+    {
+      "epoch": 0.290283091048202,
+      "grad_norm": 2.3768578395555835,
+      "learning_rate": 1.931340766140595e-05,
+      "loss": 0.8748,
+      "step": 1897
+    },
+    {
+      "epoch": 0.29043611323641927,
+      "grad_norm": 2.429358667095976,
+      "learning_rate": 1.931250494794669e-05,
+      "loss": 0.8916,
+      "step": 1898
+    },
+    {
+      "epoch": 0.29058913542463655,
+      "grad_norm": 2.494900204607473,
+      "learning_rate": 1.9311601662570072e-05,
+      "loss": 0.8617,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2907421576128539,
+      "grad_norm": 2.4468745440891295,
+      "learning_rate": 1.931069780533157e-05,
+      "loss": 0.7456,
+      "step": 1900
+    },
+    {
+      "epoch": 0.29089517980107116,
+      "grad_norm": 2.736762375685349,
+      "learning_rate": 1.930979337628669e-05,
+      "loss": 0.7943,
+      "step": 1901
+    },
+    {
+      "epoch": 0.29104820198928844,
+      "grad_norm": 2.702694904834764,
+      "learning_rate": 1.930888837549098e-05,
+      "loss": 0.8252,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2912012241775057,
+      "grad_norm": 2.527780804908085,
+      "learning_rate": 1.9307982803000017e-05,
+      "loss": 0.7839,
+      "step": 1903
+    },
+    {
+      "epoch": 0.29135424636572305,
+      "grad_norm": 2.579662212769781,
+      "learning_rate": 1.9307076658869417e-05,
+      "loss": 0.8185,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2915072685539403,
+      "grad_norm": 2.3463765755521186,
+      "learning_rate": 1.9306169943154832e-05,
+      "loss": 0.7127,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2916602907421576,
+      "grad_norm": 2.5764100564188106,
+      "learning_rate": 1.9305262655911946e-05,
+      "loss": 0.8555,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2918133129303749,
+      "grad_norm": 2.7225480761530623,
+      "learning_rate": 1.9304354797196475e-05,
+      "loss": 0.8489,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2919663351185922,
+      "grad_norm": 2.9225439859512687,
+      "learning_rate": 1.9303446367064175e-05,
+      "loss": 0.8204,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2921193573068095,
+      "grad_norm": 2.7728976761121276,
+      "learning_rate": 1.9302537365570843e-05,
+      "loss": 0.8533,
+      "step": 1909
+    },
+    {
+      "epoch": 0.29227237949502677,
+      "grad_norm": 2.788291953812961,
+      "learning_rate": 1.9301627792772297e-05,
+      "loss": 0.9378,
+      "step": 1910
+    },
+    {
+      "epoch": 0.29242540168324405,
+      "grad_norm": 2.4399042097338235,
+      "learning_rate": 1.93007176487244e-05,
+      "loss": 0.7785,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2925784238714614,
+      "grad_norm": 2.428125304895574,
+      "learning_rate": 1.929980693348305e-05,
+      "loss": 0.7808,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29273144605967866,
+      "grad_norm": 2.472689699702656,
+      "learning_rate": 1.929889564710417e-05,
+      "loss": 0.8331,
+      "step": 1913
+    },
+    {
+      "epoch": 0.29288446824789593,
+      "grad_norm": 2.5262689416061037,
+      "learning_rate": 1.9297983789643735e-05,
+      "loss": 0.8749,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2930374904361132,
+      "grad_norm": 2.638386784656946,
+      "learning_rate": 1.929707136115774e-05,
+      "loss": 0.919,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29319051262433055,
+      "grad_norm": 2.4550326677156344,
+      "learning_rate": 1.9296158361702223e-05,
+      "loss": 0.7279,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2933435348125478,
+      "grad_norm": 2.643293435117193,
+      "learning_rate": 1.9295244791333255e-05,
+      "loss": 0.8779,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2934965570007651,
+      "grad_norm": 2.5727735075626557,
+      "learning_rate": 1.929433065010694e-05,
+      "loss": 0.7984,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2936495791889824,
+      "grad_norm": 2.434329291686872,
+      "learning_rate": 1.929341593807942e-05,
+      "loss": 0.858,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2938026013771997,
+      "grad_norm": 2.4388397392820393,
+      "learning_rate": 1.9292500655306872e-05,
+      "loss": 0.8588,
+      "step": 1920
+    },
+    {
+      "epoch": 0.293955623565417,
+      "grad_norm": 2.469418880655886,
+      "learning_rate": 1.9291584801845508e-05,
+      "loss": 0.8674,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29410864575363427,
+      "grad_norm": 2.300386835534169,
+      "learning_rate": 1.929066837775157e-05,
+      "loss": 0.8688,
+      "step": 1922
+    },
+    {
+      "epoch": 0.29426166794185155,
+      "grad_norm": 2.317847305611336,
+      "learning_rate": 1.9289751383081342e-05,
+      "loss": 0.7381,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2944146901300689,
+      "grad_norm": 2.4895403782250067,
+      "learning_rate": 1.928883381789114e-05,
+      "loss": 0.9119,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29456771231828616,
+      "grad_norm": 2.56795549530539,
+      "learning_rate": 1.9287915682237314e-05,
+      "loss": 0.7846,
+      "step": 1925
+    },
+    {
+      "epoch": 0.29472073450650343,
+      "grad_norm": 2.549866674572653,
+      "learning_rate": 1.928699697617625e-05,
+      "loss": 0.7652,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2948737566947207,
+      "grad_norm": 2.4390351626614057,
+      "learning_rate": 1.9286077699764376e-05,
+      "loss": 0.7653,
+      "step": 1927
+    },
+    {
+      "epoch": 0.29502677888293805,
+      "grad_norm": 2.4418868117867394,
+      "learning_rate": 1.9285157853058142e-05,
+      "loss": 0.8644,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2951798010711553,
+      "grad_norm": 2.502305903557577,
+      "learning_rate": 1.9284237436114038e-05,
+      "loss": 0.9264,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2953328232593726,
+      "grad_norm": 2.2963754527727933,
+      "learning_rate": 1.9283316448988593e-05,
+      "loss": 0.7938,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2954858454475899,
+      "grad_norm": 3.0614462809673526,
+      "learning_rate": 1.928239489173837e-05,
+      "loss": 0.8869,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2956388676358072,
+      "grad_norm": 2.4080850709017385,
+      "learning_rate": 1.9281472764419962e-05,
+      "loss": 0.8448,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2957918898240245,
+      "grad_norm": 2.5403826284863418,
+      "learning_rate": 1.9280550067090003e-05,
+      "loss": 0.7956,
+      "step": 1933
+    },
+    {
+      "epoch": 0.29594491201224177,
+      "grad_norm": 2.2358612452947093,
+      "learning_rate": 1.927962679980516e-05,
+      "loss": 0.7314,
+      "step": 1934
+    },
+    {
+      "epoch": 0.29609793420045905,
+      "grad_norm": 2.43089626487857,
+      "learning_rate": 1.927870296262213e-05,
+      "loss": 0.8507,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2962509563886764,
+      "grad_norm": 2.455578482886986,
+      "learning_rate": 1.927777855559766e-05,
+      "loss": 0.8509,
+      "step": 1936
+    },
+    {
+      "epoch": 0.29640397857689366,
+      "grad_norm": 2.564323058941907,
+      "learning_rate": 1.9276853578788502e-05,
+      "loss": 0.9223,
+      "step": 1937
+    },
+    {
+      "epoch": 0.29655700076511093,
+      "grad_norm": 2.7623433756645994,
+      "learning_rate": 1.9275928032251484e-05,
+      "loss": 0.9499,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2967100229533282,
+      "grad_norm": 2.3216848528978558,
+      "learning_rate": 1.9275001916043436e-05,
+      "loss": 0.7305,
+      "step": 1939
+    },
+    {
+      "epoch": 0.29686304514154555,
+      "grad_norm": 2.388246616293578,
+      "learning_rate": 1.927407523022123e-05,
+      "loss": 0.928,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2970160673297628,
+      "grad_norm": 2.3536391454598906,
+      "learning_rate": 1.927314797484179e-05,
+      "loss": 0.7234,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2971690895179801,
+      "grad_norm": 2.424189518386324,
+      "learning_rate": 1.9272220149962057e-05,
+      "loss": 0.8198,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2973221117061974,
+      "grad_norm": 2.763169194414181,
+      "learning_rate": 1.927129175563901e-05,
+      "loss": 0.8644,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2974751338944147,
+      "grad_norm": 2.4385903706286274,
+      "learning_rate": 1.9270362791929663e-05,
+      "loss": 0.7624,
+      "step": 1944
+    },
+    {
+      "epoch": 0.297628156082632,
+      "grad_norm": 2.3830375578538048,
+      "learning_rate": 1.9269433258891076e-05,
+      "loss": 0.785,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29778117827084927,
+      "grad_norm": 3.284351176237654,
+      "learning_rate": 1.926850315658033e-05,
+      "loss": 0.9291,
+      "step": 1946
+    },
+    {
+      "epoch": 0.29793420045906654,
+      "grad_norm": 2.3165311177119188,
+      "learning_rate": 1.9267572485054544e-05,
+      "loss": 0.8254,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2980872226472839,
+      "grad_norm": 2.4524168762163483,
+      "learning_rate": 1.926664124437088e-05,
+      "loss": 0.7993,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29824024483550116,
+      "grad_norm": 2.6058768862175388,
+      "learning_rate": 1.9265709434586522e-05,
+      "loss": 0.816,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29839326702371843,
+      "grad_norm": 2.243298273276326,
+      "learning_rate": 1.9264777055758704e-05,
+      "loss": 0.7164,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2985462892119357,
+      "grad_norm": 2.62737281453155,
+      "learning_rate": 1.926384410794468e-05,
+      "loss": 0.9408,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29869931140015304,
+      "grad_norm": 2.661164387856838,
+      "learning_rate": 1.9262910591201752e-05,
+      "loss": 0.8959,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2988523335883703,
+      "grad_norm": 2.463256788156899,
+      "learning_rate": 1.9261976505587245e-05,
+      "loss": 0.7347,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2990053557765876,
+      "grad_norm": 2.696961971747593,
+      "learning_rate": 1.9261041851158526e-05,
+      "loss": 0.8956,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2991583779648049,
+      "grad_norm": 2.3188961750528403,
+      "learning_rate": 1.9260106627973e-05,
+      "loss": 0.8279,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2993114001530222,
+      "grad_norm": 2.4269478964045117,
+      "learning_rate": 1.92591708360881e-05,
+      "loss": 0.7931,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2994644223412395,
+      "grad_norm": 2.48052465786968,
+      "learning_rate": 1.92582344755613e-05,
+      "loss": 0.8876,
+      "step": 1957
+    },
+    {
+      "epoch": 0.29961744452945677,
+      "grad_norm": 2.309190634912905,
+      "learning_rate": 1.9257297546450097e-05,
+      "loss": 0.8796,
+      "step": 1958
+    },
+    {
+      "epoch": 0.29977046671767404,
+      "grad_norm": 2.5168712148937717,
+      "learning_rate": 1.925636004881204e-05,
+      "loss": 0.9515,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2999234889058914,
+      "grad_norm": 2.3043960080089723,
+      "learning_rate": 1.92554219827047e-05,
+      "loss": 0.8175,
+      "step": 1960
+    },
+    {
+      "epoch": 0.30007651109410866,
+      "grad_norm": 2.5426296255764176,
+      "learning_rate": 1.925448334818569e-05,
+      "loss": 0.7521,
+      "step": 1961
+    },
+    {
+      "epoch": 0.30022953328232593,
+      "grad_norm": 2.414666262662447,
+      "learning_rate": 1.9253544145312654e-05,
+      "loss": 0.9249,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3003825554705432,
+      "grad_norm": 2.5867144779294184,
+      "learning_rate": 1.925260437414327e-05,
+      "loss": 0.9349,
+      "step": 1963
+    },
+    {
+      "epoch": 0.30053557765876054,
+      "grad_norm": 2.473431576610012,
+      "learning_rate": 1.9251664034735258e-05,
+      "loss": 0.8798,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3006885998469778,
+      "grad_norm": 2.1385680671542278,
+      "learning_rate": 1.9250723127146362e-05,
+      "loss": 0.7429,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3008416220351951,
+      "grad_norm": 2.5981076958560956,
+      "learning_rate": 1.9249781651434372e-05,
+      "loss": 0.7934,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3009946442234124,
+      "grad_norm": 2.455515129458916,
+      "learning_rate": 1.92488396076571e-05,
+      "loss": 0.905,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3011476664116297,
+      "grad_norm": 2.8389366912830303,
+      "learning_rate": 1.9247896995872413e-05,
+      "loss": 0.843,
+      "step": 1968
+    },
+    {
+      "epoch": 0.301300688599847,
+      "grad_norm": 2.420967448885132,
+      "learning_rate": 1.924695381613819e-05,
+      "loss": 0.8977,
+      "step": 1969
+    },
+    {
+      "epoch": 0.30145371078806427,
+      "grad_norm": 2.2695379545380314,
+      "learning_rate": 1.924601006851236e-05,
+      "loss": 0.6859,
+      "step": 1970
+    },
+    {
+      "epoch": 0.30160673297628154,
+      "grad_norm": 2.2636926091187144,
+      "learning_rate": 1.924506575305288e-05,
+      "loss": 0.7773,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3017597551644989,
+      "grad_norm": 2.7898273917009755,
+      "learning_rate": 1.9244120869817746e-05,
+      "loss": 0.8155,
+      "step": 1972
+    },
+    {
+      "epoch": 0.30191277735271616,
+      "grad_norm": 2.4340034177552856,
+      "learning_rate": 1.924317541886499e-05,
+      "loss": 0.8266,
+      "step": 1973
+    },
+    {
+      "epoch": 0.30206579954093343,
+      "grad_norm": 2.2766416109438583,
+      "learning_rate": 1.9242229400252666e-05,
+      "loss": 0.8102,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3022188217291507,
+      "grad_norm": 2.5561178004217955,
+      "learning_rate": 1.924128281403888e-05,
+      "loss": 0.8036,
+      "step": 1975
+    },
+    {
+      "epoch": 0.30237184391736804,
+      "grad_norm": 2.379553828425003,
+      "learning_rate": 1.9240335660281764e-05,
+      "loss": 0.7768,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3025248661055853,
+      "grad_norm": 2.5252891466757292,
+      "learning_rate": 1.923938793903949e-05,
+      "loss": 0.7895,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3026778882938026,
+      "grad_norm": 2.2715040776578252,
+      "learning_rate": 1.9238439650370256e-05,
+      "loss": 0.8965,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3028309104820199,
+      "grad_norm": 2.465898560897087,
+      "learning_rate": 1.92374907943323e-05,
+      "loss": 0.7972,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3029839326702372,
+      "grad_norm": 2.628192379084707,
+      "learning_rate": 1.92365413709839e-05,
+      "loss": 0.8772,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3031369548584545,
+      "grad_norm": 2.59756485098828,
+      "learning_rate": 1.9235591380383362e-05,
+      "loss": 0.8435,
+      "step": 1981
+    },
+    {
+      "epoch": 0.30328997704667177,
+      "grad_norm": 2.3284828288876116,
+      "learning_rate": 1.9234640822589023e-05,
+      "loss": 0.7544,
+      "step": 1982
+    },
+    {
+      "epoch": 0.30344299923488904,
+      "grad_norm": 2.603774040163367,
+      "learning_rate": 1.9233689697659268e-05,
+      "loss": 0.8142,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3035960214231064,
+      "grad_norm": 2.4923640409532393,
+      "learning_rate": 1.923273800565251e-05,
+      "loss": 0.9354,
+      "step": 1984
+    },
+    {
+      "epoch": 0.30374904361132365,
+      "grad_norm": 2.6499802758692765,
+      "learning_rate": 1.9231785746627186e-05,
+      "loss": 0.8094,
+      "step": 1985
+    },
+    {
+      "epoch": 0.30390206579954093,
+      "grad_norm": 2.2595162035840697,
+      "learning_rate": 1.9230832920641784e-05,
+      "loss": 0.8003,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3040550879877582,
+      "grad_norm": 2.4192559742903246,
+      "learning_rate": 1.922987952775482e-05,
+      "loss": 0.804,
+      "step": 1987
+    },
+    {
+      "epoch": 0.30420811017597554,
+      "grad_norm": 2.316981550914932,
+      "learning_rate": 1.9228925568024855e-05,
+      "loss": 0.754,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3043611323641928,
+      "grad_norm": 7.151782661755483,
+      "learning_rate": 1.9227971041510463e-05,
+      "loss": 0.8844,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3045141545524101,
+      "grad_norm": 2.734107657629692,
+      "learning_rate": 1.9227015948270266e-05,
+      "loss": 0.7944,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3046671767406274,
+      "grad_norm": 2.529668625783798,
+      "learning_rate": 1.9226060288362927e-05,
+      "loss": 0.8836,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3048201989288447,
+      "grad_norm": 2.546234327275014,
+      "learning_rate": 1.922510406184713e-05,
+      "loss": 0.7898,
+      "step": 1992
+    },
+    {
+      "epoch": 0.304973221117062,
+      "grad_norm": 2.395162331895941,
+      "learning_rate": 1.922414726878161e-05,
+      "loss": 0.7806,
+      "step": 1993
+    },
+    {
+      "epoch": 0.30512624330527927,
+      "grad_norm": 2.55658252083043,
+      "learning_rate": 1.9223189909225114e-05,
+      "loss": 0.8855,
+      "step": 1994
+    },
+    {
+      "epoch": 0.30527926549349654,
+      "grad_norm": 2.4501674002626976,
+      "learning_rate": 1.9222231983236447e-05,
+      "loss": 0.8802,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3054322876817138,
+      "grad_norm": 2.4889603134355314,
+      "learning_rate": 1.9221273490874436e-05,
+      "loss": 0.7977,
+      "step": 1996
+    },
+    {
+      "epoch": 0.30558530986993115,
+      "grad_norm": 2.3944327612655725,
+      "learning_rate": 1.922031443219795e-05,
+      "loss": 0.832,
+      "step": 1997
+    },
+    {
+      "epoch": 0.30573833205814843,
+      "grad_norm": 2.995473969176262,
+      "learning_rate": 1.921935480726588e-05,
+      "loss": 0.7878,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3058913542463657,
+      "grad_norm": 2.7168338445829474,
+      "learning_rate": 1.9218394616137166e-05,
+      "loss": 0.8424,
+      "step": 1999
+    },
+    {
+      "epoch": 0.306044376434583,
+      "grad_norm": 2.5441476948697854,
+      "learning_rate": 1.921743385887078e-05,
+      "loss": 0.8708,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3061973986228003,
+      "grad_norm": 2.4494392535414096,
+      "learning_rate": 1.9216472535525717e-05,
+      "loss": 0.8481,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3063504208110176,
+      "grad_norm": 2.6112325729047923,
+      "learning_rate": 1.921551064616102e-05,
+      "loss": 0.8245,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3065034429992349,
+      "grad_norm": 2.3444443429432416,
+      "learning_rate": 1.9214548190835766e-05,
+      "loss": 0.7449,
+      "step": 2003
+    },
+    {
+      "epoch": 0.30665646518745215,
+      "grad_norm": 2.446188822684606,
+      "learning_rate": 1.9213585169609058e-05,
+      "loss": 0.831,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3068094873756695,
+      "grad_norm": 2.2145856338125602,
+      "learning_rate": 1.9212621582540042e-05,
+      "loss": 0.7718,
+      "step": 2005
+    },
+    {
+      "epoch": 0.30696250956388677,
+      "grad_norm": 2.5819277402890526,
+      "learning_rate": 1.9211657429687895e-05,
+      "loss": 0.8231,
+      "step": 2006
+    },
+    {
+      "epoch": 0.30711553175210404,
+      "grad_norm": 2.0650856048224617,
+      "learning_rate": 1.9210692711111827e-05,
+      "loss": 0.8361,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3072685539403213,
+      "grad_norm": 2.6944527828836615,
+      "learning_rate": 1.920972742687109e-05,
+      "loss": 0.8581,
+      "step": 2008
+    },
+    {
+      "epoch": 0.30742157612853865,
+      "grad_norm": 2.1937400834532825,
+      "learning_rate": 1.920876157702496e-05,
+      "loss": 0.7833,
+      "step": 2009
+    },
+    {
+      "epoch": 0.30757459831675593,
+      "grad_norm": 2.5777698912582205,
+      "learning_rate": 1.920779516163276e-05,
+      "loss": 0.8232,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3077276205049732,
+      "grad_norm": 2.2556884773655463,
+      "learning_rate": 1.920682818075384e-05,
+      "loss": 0.7792,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3078806426931905,
+      "grad_norm": 2.682048126514191,
+      "learning_rate": 1.920586063444758e-05,
+      "loss": 0.8919,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3080336648814078,
+      "grad_norm": 2.570180102807476,
+      "learning_rate": 1.9204892522773405e-05,
+      "loss": 0.8524,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3081866870696251,
+      "grad_norm": 2.563307171833568,
+      "learning_rate": 1.9203923845790767e-05,
+      "loss": 0.9107,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3083397092578424,
+      "grad_norm": 2.2104842092303336,
+      "learning_rate": 1.9202954603559164e-05,
+      "loss": 0.843,
+      "step": 2015
+    },
+    {
+      "epoch": 0.30849273144605965,
+      "grad_norm": 2.5360024853513417,
+      "learning_rate": 1.920198479613812e-05,
+      "loss": 0.7056,
+      "step": 2016
+    },
+    {
+      "epoch": 0.308645753634277,
+      "grad_norm": 2.7399669049444255,
+      "learning_rate": 1.9201014423587187e-05,
+      "loss": 0.8708,
+      "step": 2017
+    },
+    {
+      "epoch": 0.30879877582249426,
+      "grad_norm": 2.4553097150606447,
+      "learning_rate": 1.9200043485965963e-05,
+      "loss": 0.8834,
+      "step": 2018
+    },
+    {
+      "epoch": 0.30895179801071154,
+      "grad_norm": 2.8655692998053697,
+      "learning_rate": 1.919907198333408e-05,
+      "loss": 0.9893,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3091048201989288,
+      "grad_norm": 2.2561640988275546,
+      "learning_rate": 1.9198099915751198e-05,
+      "loss": 0.8193,
+      "step": 2020
+    },
+    {
+      "epoch": 0.30925784238714615,
+      "grad_norm": 2.4836616435229,
+      "learning_rate": 1.9197127283277017e-05,
+      "loss": 0.8635,
+      "step": 2021
+    },
+    {
+      "epoch": 0.30941086457536343,
+      "grad_norm": 2.5715982339637065,
+      "learning_rate": 1.919615408597127e-05,
+      "loss": 0.8791,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3095638867635807,
+      "grad_norm": 2.5943107227567848,
+      "learning_rate": 1.9195180323893728e-05,
+      "loss": 0.8121,
+      "step": 2023
+    },
+    {
+      "epoch": 0.309716908951798,
+      "grad_norm": 2.2801541193319492,
+      "learning_rate": 1.9194205997104187e-05,
+      "loss": 0.7939,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3098699311400153,
+      "grad_norm": 2.459839184463675,
+      "learning_rate": 1.919323110566249e-05,
+      "loss": 0.9424,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3100229533282326,
+      "grad_norm": 2.1911119079151993,
+      "learning_rate": 1.9192255649628503e-05,
+      "loss": 0.8289,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3101759755164499,
+      "grad_norm": 2.4644860950022722,
+      "learning_rate": 1.919127962906214e-05,
+      "loss": 0.8372,
+      "step": 2027
+    },
+    {
+      "epoch": 0.31032899770466715,
+      "grad_norm": 2.2356742468806026,
+      "learning_rate": 1.919030304402334e-05,
+      "loss": 0.7998,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3104820198928845,
+      "grad_norm": 2.5171160782594795,
+      "learning_rate": 1.9189325894572076e-05,
+      "loss": 0.8296,
+      "step": 2029
+    },
+    {
+      "epoch": 0.31063504208110176,
+      "grad_norm": 2.7014831046380934,
+      "learning_rate": 1.9188348180768356e-05,
+      "loss": 0.9374,
+      "step": 2030
+    },
+    {
+      "epoch": 0.31078806426931904,
+      "grad_norm": 2.734471374350287,
+      "learning_rate": 1.918736990267223e-05,
+      "loss": 0.7857,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3109410864575363,
+      "grad_norm": 2.7131784176496487,
+      "learning_rate": 1.918639106034378e-05,
+      "loss": 0.8601,
+      "step": 2032
+    },
+    {
+      "epoch": 0.31109410864575365,
+      "grad_norm": 2.513092380080759,
+      "learning_rate": 1.918541165384312e-05,
+      "loss": 0.7936,
+      "step": 2033
+    },
+    {
+      "epoch": 0.31124713083397093,
+      "grad_norm": 2.927728563084092,
+      "learning_rate": 1.9184431683230392e-05,
+      "loss": 1.0046,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3114001530221882,
+      "grad_norm": 2.404772452838863,
+      "learning_rate": 1.9183451148565787e-05,
+      "loss": 0.7403,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3115531752104055,
+      "grad_norm": 2.516603840983081,
+      "learning_rate": 1.918247004990952e-05,
+      "loss": 0.7745,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3117061973986228,
+      "grad_norm": 2.386770013882607,
+      "learning_rate": 1.9181488387321848e-05,
+      "loss": 0.8298,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3118592195868401,
+      "grad_norm": 2.6428767252854555,
+      "learning_rate": 1.918050616086305e-05,
+      "loss": 0.8006,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3120122417750574,
+      "grad_norm": 3.1631302400947825,
+      "learning_rate": 1.917952337059346e-05,
+      "loss": 0.8649,
+      "step": 2039
+    },
+    {
+      "epoch": 0.31216526396327465,
+      "grad_norm": 2.442779992040568,
+      "learning_rate": 1.9178540016573425e-05,
+      "loss": 0.6726,
+      "step": 2040
+    },
+    {
+      "epoch": 0.312318286151492,
+      "grad_norm": 2.44707016093417,
+      "learning_rate": 1.9177556098863345e-05,
+      "loss": 0.8137,
+      "step": 2041
+    },
+    {
+      "epoch": 0.31247130833970926,
+      "grad_norm": 2.525282513459153,
+      "learning_rate": 1.9176571617523637e-05,
+      "loss": 0.686,
+      "step": 2042
+    },
+    {
+      "epoch": 0.31262433052792654,
+      "grad_norm": 2.5982517000505405,
+      "learning_rate": 1.9175586572614772e-05,
+      "loss": 0.7722,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3127773527161438,
+      "grad_norm": 2.452454622090508,
+      "learning_rate": 1.9174600964197243e-05,
+      "loss": 0.8334,
+      "step": 2044
+    },
+    {
+      "epoch": 0.31293037490436115,
+      "grad_norm": 2.618928484678063,
+      "learning_rate": 1.917361479233157e-05,
+      "loss": 0.9241,
+      "step": 2045
+    },
+    {
+      "epoch": 0.31308339709257843,
+      "grad_norm": 2.435076329407817,
+      "learning_rate": 1.9172628057078328e-05,
+      "loss": 0.7933,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3132364192807957,
+      "grad_norm": 2.533700576320131,
+      "learning_rate": 1.9171640758498117e-05,
+      "loss": 0.7797,
+      "step": 2047
+    },
+    {
+      "epoch": 0.313389441469013,
+      "grad_norm": 2.4696336812088546,
+      "learning_rate": 1.917065289665156e-05,
+      "loss": 0.7301,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3135424636572303,
+      "grad_norm": 2.4661352148431868,
+      "learning_rate": 1.9169664471599338e-05,
+      "loss": 0.8389,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3136954858454476,
+      "grad_norm": 2.3361325316370767,
+      "learning_rate": 1.916867548340215e-05,
+      "loss": 0.7848,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3138485080336649,
+      "grad_norm": 2.7039151927382106,
+      "learning_rate": 1.916768593212073e-05,
+      "loss": 0.7517,
+      "step": 2051
+    },
+    {
+      "epoch": 0.31400153022188215,
+      "grad_norm": 2.302389279273721,
+      "learning_rate": 1.9166695817815854e-05,
+      "loss": 0.816,
+      "step": 2052
+    },
+    {
+      "epoch": 0.3141545524100995,
+      "grad_norm": 2.33484318748721,
+      "learning_rate": 1.9165705140548325e-05,
+      "loss": 0.7547,
+      "step": 2053
+    },
+    {
+      "epoch": 0.31430757459831676,
+      "grad_norm": 2.4695189957920336,
+      "learning_rate": 1.916471390037899e-05,
+      "loss": 0.8419,
+      "step": 2054
+    },
+    {
+      "epoch": 0.31446059678653404,
+      "grad_norm": 2.6640125173002573,
+      "learning_rate": 1.916372209736872e-05,
+      "loss": 0.8851,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3146136189747513,
+      "grad_norm": 2.5342638647426154,
+      "learning_rate": 1.9162729731578428e-05,
+      "loss": 0.7527,
+      "step": 2056
+    },
+    {
+      "epoch": 0.31476664116296865,
+      "grad_norm": 2.534955417695161,
+      "learning_rate": 1.916173680306906e-05,
+      "loss": 0.8765,
+      "step": 2057
+    },
+    {
+      "epoch": 0.31491966335118593,
+      "grad_norm": 2.6830494811488,
+      "learning_rate": 1.9160743311901592e-05,
+      "loss": 0.7909,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3150726855394032,
+      "grad_norm": 2.550549374015315,
+      "learning_rate": 1.915974925813704e-05,
+      "loss": 0.7659,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3152257077276205,
+      "grad_norm": 2.3928412556709624,
+      "learning_rate": 1.9158754641836452e-05,
+      "loss": 0.7826,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3153787299158378,
+      "grad_norm": 2.9224489475145274,
+      "learning_rate": 1.9157759463060914e-05,
+      "loss": 1.0251,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3155317521040551,
+      "grad_norm": 2.5484811016781466,
+      "learning_rate": 1.9156763721871542e-05,
+      "loss": 0.8764,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3156847742922724,
+      "grad_norm": 2.446852756534984,
+      "learning_rate": 1.9155767418329488e-05,
+      "loss": 0.8835,
+      "step": 2063
+    },
+    {
+      "epoch": 0.31583779648048965,
+      "grad_norm": 2.72442600547987,
+      "learning_rate": 1.915477055249594e-05,
+      "loss": 0.8934,
+      "step": 2064
+    },
+    {
+      "epoch": 0.315990818668707,
+      "grad_norm": 2.4582101077154195,
+      "learning_rate": 1.9153773124432117e-05,
+      "loss": 0.7802,
+      "step": 2065
+    },
+    {
+      "epoch": 0.31614384085692426,
+      "grad_norm": 2.5396120010948358,
+      "learning_rate": 1.9152775134199278e-05,
+      "loss": 0.867,
+      "step": 2066
+    },
+    {
+      "epoch": 0.31629686304514154,
+      "grad_norm": 2.3762790446329487,
+      "learning_rate": 1.9151776581858708e-05,
+      "loss": 0.7767,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3164498852333588,
+      "grad_norm": 2.433257226366239,
+      "learning_rate": 1.915077746747174e-05,
+      "loss": 0.791,
+      "step": 2068
+    },
+    {
+      "epoch": 0.31660290742157615,
+      "grad_norm": 2.4692823299271867,
+      "learning_rate": 1.914977779109973e-05,
+      "loss": 0.8087,
+      "step": 2069
+    },
+    {
+      "epoch": 0.31675592960979343,
+      "grad_norm": 2.131604538235996,
+      "learning_rate": 1.9148777552804075e-05,
+      "loss": 0.7566,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3169089517980107,
+      "grad_norm": 2.4472979026288533,
+      "learning_rate": 1.9147776752646193e-05,
+      "loss": 0.9065,
+      "step": 2071
+    },
+    {
+      "epoch": 0.317061973986228,
+      "grad_norm": 2.484643272652111,
+      "learning_rate": 1.914677539068756e-05,
+      "loss": 0.8602,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3172149961744453,
+      "grad_norm": 2.5281446531443112,
+      "learning_rate": 1.9145773466989665e-05,
+      "loss": 0.8182,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3173680183626626,
+      "grad_norm": 2.5908840665780755,
+      "learning_rate": 1.9144770981614043e-05,
+      "loss": 0.7987,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3175210405508799,
+      "grad_norm": 2.0941908773437943,
+      "learning_rate": 1.914376793462226e-05,
+      "loss": 0.896,
+      "step": 2075
+    },
+    {
+      "epoch": 0.31767406273909715,
+      "grad_norm": 2.2421817228071705,
+      "learning_rate": 1.914276432607592e-05,
+      "loss": 0.6662,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3178270849273145,
+      "grad_norm": 2.4922409133172394,
+      "learning_rate": 1.9141760156036656e-05,
+      "loss": 0.7889,
+      "step": 2077
+    },
+    {
+      "epoch": 0.31798010711553176,
+      "grad_norm": 2.3788296857697695,
+      "learning_rate": 1.9140755424566135e-05,
+      "loss": 0.7257,
+      "step": 2078
+    },
+    {
+      "epoch": 0.31813312930374904,
+      "grad_norm": 2.3379923933589017,
+      "learning_rate": 1.9139750131726067e-05,
+      "loss": 0.7745,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3182861514919663,
+      "grad_norm": 2.873252027227465,
+      "learning_rate": 1.9138744277578186e-05,
+      "loss": 0.8744,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31843917368018365,
+      "grad_norm": 2.4216698005628503,
+      "learning_rate": 1.9137737862184272e-05,
+      "loss": 0.8559,
+      "step": 2081
+    },
+    {
+      "epoch": 0.31859219586840093,
+      "grad_norm": 2.4418905760060454,
+      "learning_rate": 1.9136730885606124e-05,
+      "loss": 0.7539,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3187452180566182,
+      "grad_norm": 2.4020958445199936,
+      "learning_rate": 1.9135723347905593e-05,
+      "loss": 0.827,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3188982402448355,
+      "grad_norm": 2.5406101984226908,
+      "learning_rate": 1.913471524914455e-05,
+      "loss": 0.8508,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3190512624330528,
+      "grad_norm": 2.2918266846443736,
+      "learning_rate": 1.913370658938491e-05,
+      "loss": 0.7455,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3192042846212701,
+      "grad_norm": 2.480568139223291,
+      "learning_rate": 1.9132697368688616e-05,
+      "loss": 0.8116,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3193573068094874,
+      "grad_norm": 2.854465154555059,
+      "learning_rate": 1.913168758711765e-05,
+      "loss": 0.9847,
+      "step": 2087
+    },
+    {
+      "epoch": 0.31951032899770465,
+      "grad_norm": 2.572434816180969,
+      "learning_rate": 1.9130677244734026e-05,
+      "loss": 0.8183,
+      "step": 2088
+    },
+    {
+      "epoch": 0.319663351185922,
+      "grad_norm": 3.003205537027778,
+      "learning_rate": 1.9129666341599793e-05,
+      "loss": 0.853,
+      "step": 2089
+    },
+    {
+      "epoch": 0.31981637337413926,
+      "grad_norm": 2.4337824036824953,
+      "learning_rate": 1.9128654877777034e-05,
+      "loss": 0.7883,
+      "step": 2090
+    },
+    {
+      "epoch": 0.31996939556235654,
+      "grad_norm": 2.5391007309566302,
+      "learning_rate": 1.9127642853327867e-05,
+      "loss": 0.7448,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3201224177505738,
+      "grad_norm": 2.5479704491291986,
+      "learning_rate": 1.9126630268314447e-05,
+      "loss": 0.7959,
+      "step": 2092
+    },
+    {
+      "epoch": 0.32027543993879115,
+      "grad_norm": 2.7326391324081016,
+      "learning_rate": 1.9125617122798952e-05,
+      "loss": 0.8832,
+      "step": 2093
+    },
+    {
+      "epoch": 0.32042846212700843,
+      "grad_norm": 2.1860153989983298,
+      "learning_rate": 1.9124603416843617e-05,
+      "loss": 0.8377,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3205814843152257,
+      "grad_norm": 2.503472081997963,
+      "learning_rate": 1.912358915051069e-05,
+      "loss": 0.7138,
+      "step": 2095
+    },
+    {
+      "epoch": 0.320734506503443,
+      "grad_norm": 2.2062061908006845,
+      "learning_rate": 1.912257432386246e-05,
+      "loss": 0.7035,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3208875286916603,
+      "grad_norm": 2.077975827552243,
+      "learning_rate": 1.912155893696125e-05,
+      "loss": 0.71,
+      "step": 2097
+    },
+    {
+      "epoch": 0.3210405508798776,
+      "grad_norm": 2.4162267436191307,
+      "learning_rate": 1.9120542989869427e-05,
+      "loss": 0.8393,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3211935730680949,
+      "grad_norm": 2.5988343283265154,
+      "learning_rate": 1.911952648264938e-05,
+      "loss": 0.8368,
+      "step": 2099
+    },
+    {
+      "epoch": 0.32134659525631215,
+      "grad_norm": 2.2723363777964662,
+      "learning_rate": 1.911850941536353e-05,
+      "loss": 0.7547,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3214996174445295,
+      "grad_norm": 2.413632966823172,
+      "learning_rate": 1.9117491788074348e-05,
+      "loss": 0.774,
+      "step": 2101
+    },
+    {
+      "epoch": 0.32165263963274676,
+      "grad_norm": 2.439605514523114,
+      "learning_rate": 1.9116473600844327e-05,
+      "loss": 0.7549,
+      "step": 2102
+    },
+    {
+      "epoch": 0.32180566182096404,
+      "grad_norm": 2.2988123659235473,
+      "learning_rate": 1.9115454853736e-05,
+      "loss": 0.7875,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3219586840091813,
+      "grad_norm": 2.3470597989174697,
+      "learning_rate": 1.9114435546811928e-05,
+      "loss": 0.7903,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3221117061973986,
+      "grad_norm": 2.562614663596328,
+      "learning_rate": 1.9113415680134717e-05,
+      "loss": 0.8583,
+      "step": 2105
+    },
+    {
+      "epoch": 0.32226472838561593,
+      "grad_norm": 2.6369756053615907,
+      "learning_rate": 1.9112395253766995e-05,
+      "loss": 0.9672,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3224177505738332,
+      "grad_norm": 2.914943083424258,
+      "learning_rate": 1.911137426777143e-05,
+      "loss": 0.827,
+      "step": 2107
+    },
+    {
+      "epoch": 0.3225707727620505,
+      "grad_norm": 2.3463201198738863,
+      "learning_rate": 1.911035272221073e-05,
+      "loss": 0.8843,
+      "step": 2108
+    },
+    {
+      "epoch": 0.32272379495026776,
+      "grad_norm": 2.5753832932412637,
+      "learning_rate": 1.910933061714763e-05,
+      "loss": 0.7845,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3228768171384851,
+      "grad_norm": 2.677331183762575,
+      "learning_rate": 1.91083079526449e-05,
+      "loss": 0.8315,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3230298393267024,
+      "grad_norm": 2.619466419949009,
+      "learning_rate": 1.9107284728765347e-05,
+      "loss": 0.9082,
+      "step": 2111
+    },
+    {
+      "epoch": 0.32318286151491965,
+      "grad_norm": 2.318593309042997,
+      "learning_rate": 1.910626094557181e-05,
+      "loss": 0.7603,
+      "step": 2112
+    },
+    {
+      "epoch": 0.32333588370313693,
+      "grad_norm": 2.5023945668186625,
+      "learning_rate": 1.9105236603127167e-05,
+      "loss": 0.8891,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32348890589135426,
+      "grad_norm": 2.6435011702222755,
+      "learning_rate": 1.910421170149432e-05,
+      "loss": 0.8287,
+      "step": 2114
+    },
+    {
+      "epoch": 0.32364192807957154,
+      "grad_norm": 2.3861826759572966,
+      "learning_rate": 1.910318624073622e-05,
+      "loss": 0.7984,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3237949502677888,
+      "grad_norm": 2.5742746321849106,
+      "learning_rate": 1.910216022091584e-05,
+      "loss": 0.9795,
+      "step": 2116
+    },
+    {
+      "epoch": 0.3239479724560061,
+      "grad_norm": 2.471229350103612,
+      "learning_rate": 1.910113364209619e-05,
+      "loss": 0.7576,
+      "step": 2117
+    },
+    {
+      "epoch": 0.32410099464422343,
+      "grad_norm": 2.4970893298757546,
+      "learning_rate": 1.9100106504340324e-05,
+      "loss": 0.7732,
+      "step": 2118
+    },
+    {
+      "epoch": 0.3242540168324407,
+      "grad_norm": 2.268656403612154,
+      "learning_rate": 1.9099078807711318e-05,
+      "loss": 0.7906,
+      "step": 2119
+    },
+    {
+      "epoch": 0.324407039020658,
+      "grad_norm": 2.4314592035006557,
+      "learning_rate": 1.9098050552272284e-05,
+      "loss": 0.9111,
+      "step": 2120
+    },
+    {
+      "epoch": 0.32456006120887526,
+      "grad_norm": 2.4708369395051046,
+      "learning_rate": 1.909702173808637e-05,
+      "loss": 0.8278,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3247130833970926,
+      "grad_norm": 2.6888386351450113,
+      "learning_rate": 1.909599236521677e-05,
+      "loss": 0.785,
+      "step": 2122
+    },
+    {
+      "epoch": 0.3248661055853099,
+      "grad_norm": 2.530033705647798,
+      "learning_rate": 1.9094962433726692e-05,
+      "loss": 0.7601,
+      "step": 2123
+    },
+    {
+      "epoch": 0.32501912777352715,
+      "grad_norm": 2.577311046308443,
+      "learning_rate": 1.9093931943679395e-05,
+      "loss": 0.8827,
+      "step": 2124
+    },
+    {
+      "epoch": 0.32517214996174443,
+      "grad_norm": 2.682264853957065,
+      "learning_rate": 1.909290089513816e-05,
+      "loss": 0.8972,
+      "step": 2125
+    },
+    {
+      "epoch": 0.32532517214996176,
+      "grad_norm": 2.292095938273781,
+      "learning_rate": 1.9091869288166305e-05,
+      "loss": 0.8293,
+      "step": 2126
+    },
+    {
+      "epoch": 0.32547819433817904,
+      "grad_norm": 2.5223280792400704,
+      "learning_rate": 1.9090837122827195e-05,
+      "loss": 0.8509,
+      "step": 2127
+    },
+    {
+      "epoch": 0.3256312165263963,
+      "grad_norm": 2.447119418811232,
+      "learning_rate": 1.908980439918421e-05,
+      "loss": 0.7685,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3257842387146136,
+      "grad_norm": 2.6601569588663048,
+      "learning_rate": 1.9088771117300778e-05,
+      "loss": 0.9052,
+      "step": 2129
+    },
+    {
+      "epoch": 0.3259372609028309,
+      "grad_norm": 2.536018852055662,
+      "learning_rate": 1.908773727724036e-05,
+      "loss": 0.8709,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3260902830910482,
+      "grad_norm": 2.330125546657301,
+      "learning_rate": 1.908670287906644e-05,
+      "loss": 0.8163,
+      "step": 2131
+    },
+    {
+      "epoch": 0.3262433052792655,
+      "grad_norm": 2.368910748289313,
+      "learning_rate": 1.9085667922842547e-05,
+      "loss": 0.8446,
+      "step": 2132
+    },
+    {
+      "epoch": 0.32639632746748276,
+      "grad_norm": 2.460618804621041,
+      "learning_rate": 1.908463240863225e-05,
+      "loss": 0.8018,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3265493496557001,
+      "grad_norm": 2.5082175611570063,
+      "learning_rate": 1.9083596336499133e-05,
+      "loss": 0.7482,
+      "step": 2134
+    },
+    {
+      "epoch": 0.32670237184391737,
+      "grad_norm": 2.8829890496898294,
+      "learning_rate": 1.908255970650683e-05,
+      "loss": 0.7766,
+      "step": 2135
+    },
+    {
+      "epoch": 0.32685539403213465,
+      "grad_norm": 2.6191018339121346,
+      "learning_rate": 1.9081522518719006e-05,
+      "loss": 0.8283,
+      "step": 2136
+    },
+    {
+      "epoch": 0.3270084162203519,
+      "grad_norm": 2.675185201207612,
+      "learning_rate": 1.9080484773199356e-05,
+      "loss": 1.0072,
+      "step": 2137
+    },
+    {
+      "epoch": 0.32716143840856926,
+      "grad_norm": 2.5802166600723293,
+      "learning_rate": 1.9079446470011615e-05,
+      "loss": 0.789,
+      "step": 2138
+    },
+    {
+      "epoch": 0.32731446059678654,
+      "grad_norm": 2.7923015418179107,
+      "learning_rate": 1.9078407609219543e-05,
+      "loss": 0.7883,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3274674827850038,
+      "grad_norm": 2.418163628175102,
+      "learning_rate": 1.9077368190886946e-05,
+      "loss": 0.7245,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3276205049732211,
+      "grad_norm": 2.598496284037849,
+      "learning_rate": 1.907632821507766e-05,
+      "loss": 0.7575,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3277735271614384,
+      "grad_norm": 2.7343117555473846,
+      "learning_rate": 1.907528768185555e-05,
+      "loss": 0.874,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3279265493496557,
+      "grad_norm": 2.642891739247081,
+      "learning_rate": 1.9074246591284522e-05,
+      "loss": 0.8379,
+      "step": 2143
+    },
+    {
+      "epoch": 0.328079571537873,
+      "grad_norm": 2.608322969653126,
+      "learning_rate": 1.9073204943428512e-05,
+      "loss": 0.8056,
+      "step": 2144
+    },
+    {
+      "epoch": 0.32823259372609026,
+      "grad_norm": 2.6873976836473767,
+      "learning_rate": 1.907216273835149e-05,
+      "loss": 0.7994,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3283856159143076,
+      "grad_norm": 2.4514711844365658,
+      "learning_rate": 1.9071119976117465e-05,
+      "loss": 0.8881,
+      "step": 2146
+    },
+    {
+      "epoch": 0.32853863810252487,
+      "grad_norm": 2.5267854318495515,
+      "learning_rate": 1.9070076656790475e-05,
+      "loss": 0.8897,
+      "step": 2147
+    },
+    {
+      "epoch": 0.32869166029074215,
+      "grad_norm": 2.1105651231627647,
+      "learning_rate": 1.9069032780434596e-05,
+      "loss": 0.7618,
+      "step": 2148
+    },
+    {
+      "epoch": 0.3288446824789594,
+      "grad_norm": 2.3346993598677765,
+      "learning_rate": 1.9067988347113934e-05,
+      "loss": 0.785,
+      "step": 2149
+    },
+    {
+      "epoch": 0.32899770466717676,
+      "grad_norm": 2.4214920221821203,
+      "learning_rate": 1.9066943356892636e-05,
+      "loss": 0.8432,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32915072685539404,
+      "grad_norm": 2.558431202242363,
+      "learning_rate": 1.9065897809834877e-05,
+      "loss": 0.9265,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3293037490436113,
+      "grad_norm": 2.411042024636327,
+      "learning_rate": 1.9064851706004862e-05,
+      "loss": 0.8417,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3294567712318286,
+      "grad_norm": 2.832038379528674,
+      "learning_rate": 1.9063805045466847e-05,
+      "loss": 0.8054,
+      "step": 2153
+    },
+    {
+      "epoch": 0.3296097934200459,
+      "grad_norm": 2.6592576047547123,
+      "learning_rate": 1.9062757828285105e-05,
+      "loss": 0.9434,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3297628156082632,
+      "grad_norm": 2.352342451790287,
+      "learning_rate": 1.9061710054523953e-05,
+      "loss": 0.7586,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3299158377964805,
+      "grad_norm": 2.544209759964932,
+      "learning_rate": 1.9060661724247734e-05,
+      "loss": 0.8578,
+      "step": 2156
+    },
+    {
+      "epoch": 0.33006885998469776,
+      "grad_norm": 2.276504238923556,
+      "learning_rate": 1.9059612837520834e-05,
+      "loss": 0.8868,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3302218821729151,
+      "grad_norm": 2.5733234300853938,
+      "learning_rate": 1.9058563394407668e-05,
+      "loss": 0.9171,
+      "step": 2158
+    },
+    {
+      "epoch": 0.33037490436113237,
+      "grad_norm": 2.431640470434858,
+      "learning_rate": 1.9057513394972687e-05,
+      "loss": 0.6812,
+      "step": 2159
+    },
+    {
+      "epoch": 0.33052792654934965,
+      "grad_norm": 2.603271164903475,
+      "learning_rate": 1.9056462839280376e-05,
+      "loss": 0.8457,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3306809487375669,
+      "grad_norm": 2.731126948671394,
+      "learning_rate": 1.905541172739525e-05,
+      "loss": 0.7824,
+      "step": 2161
+    },
+    {
+      "epoch": 0.33083397092578426,
+      "grad_norm": 2.753119967092687,
+      "learning_rate": 1.905436005938187e-05,
+      "loss": 0.8542,
+      "step": 2162
+    },
+    {
+      "epoch": 0.33098699311400154,
+      "grad_norm": 2.337685299070537,
+      "learning_rate": 1.9053307835304812e-05,
+      "loss": 0.8189,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3311400153022188,
+      "grad_norm": 2.320976989090415,
+      "learning_rate": 1.9052255055228707e-05,
+      "loss": 0.6544,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3312930374904361,
+      "grad_norm": 2.5843477334717653,
+      "learning_rate": 1.9051201719218207e-05,
+      "loss": 0.8549,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3314460596786534,
+      "grad_norm": 2.129202487167354,
+      "learning_rate": 1.9050147827337996e-05,
+      "loss": 0.7141,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3315990818668707,
+      "grad_norm": 2.5652483228964242,
+      "learning_rate": 1.904909337965281e-05,
+      "loss": 0.7735,
+      "step": 2167
+    },
+    {
+      "epoch": 0.331752104055088,
+      "grad_norm": 2.3273524046603633,
+      "learning_rate": 1.9048038376227392e-05,
+      "loss": 0.8203,
+      "step": 2168
+    },
+    {
+      "epoch": 0.33190512624330526,
+      "grad_norm": 2.731759408744939,
+      "learning_rate": 1.9046982817126545e-05,
+      "loss": 0.7929,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3320581484315226,
+      "grad_norm": 2.395296587410217,
+      "learning_rate": 1.904592670241509e-05,
+      "loss": 0.9106,
+      "step": 2170
+    },
+    {
+      "epoch": 0.33221117061973987,
+      "grad_norm": 2.6281801579811566,
+      "learning_rate": 1.904487003215789e-05,
+      "loss": 0.8377,
+      "step": 2171
+    },
+    {
+      "epoch": 0.33236419280795715,
+      "grad_norm": 2.496937430980088,
+      "learning_rate": 1.9043812806419833e-05,
+      "loss": 0.7716,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3325172149961744,
+      "grad_norm": 2.335221367154538,
+      "learning_rate": 1.9042755025265858e-05,
+      "loss": 0.9313,
+      "step": 2173
+    },
+    {
+      "epoch": 0.33267023718439176,
+      "grad_norm": 2.4705985458013457,
+      "learning_rate": 1.904169668876092e-05,
+      "loss": 0.8674,
+      "step": 2174
+    },
+    {
+      "epoch": 0.33282325937260904,
+      "grad_norm": 2.470941759164986,
+      "learning_rate": 1.9040637796970013e-05,
+      "loss": 0.8822,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3329762815608263,
+      "grad_norm": 2.7471317200021725,
+      "learning_rate": 1.9039578349958177e-05,
+      "loss": 0.9441,
+      "step": 2176
+    },
+    {
+      "epoch": 0.3331293037490436,
+      "grad_norm": 2.561583712652535,
+      "learning_rate": 1.9038518347790468e-05,
+      "loss": 0.8602,
+      "step": 2177
+    },
+    {
+      "epoch": 0.3332823259372609,
+      "grad_norm": 2.636580914055904,
+      "learning_rate": 1.9037457790531988e-05,
+      "loss": 0.8297,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3334353481254782,
+      "grad_norm": 2.874894906101151,
+      "learning_rate": 1.9036396678247872e-05,
+      "loss": 0.7911,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3335883703136955,
+      "grad_norm": 2.449895925283834,
+      "learning_rate": 1.9035335011003286e-05,
+      "loss": 0.8366,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33374139250191276,
+      "grad_norm": 2.294977029076781,
+      "learning_rate": 1.903427278886343e-05,
+      "loss": 0.7386,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3338944146901301,
+      "grad_norm": 3.231281267707513,
+      "learning_rate": 1.903321001189354e-05,
+      "loss": 0.8684,
+      "step": 2182
+    },
+    {
+      "epoch": 0.33404743687834737,
+      "grad_norm": 2.787124174288599,
+      "learning_rate": 1.9032146680158884e-05,
+      "loss": 0.8729,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33420045906656465,
+      "grad_norm": 2.636855875087689,
+      "learning_rate": 1.903108279372477e-05,
+      "loss": 0.8681,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3343534812547819,
+      "grad_norm": 2.3818211247344054,
+      "learning_rate": 1.9030018352656525e-05,
+      "loss": 0.9681,
+      "step": 2185
+    },
+    {
+      "epoch": 0.33450650344299926,
+      "grad_norm": 2.6499490500959655,
+      "learning_rate": 1.9028953357019534e-05,
+      "loss": 0.7594,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33465952563121654,
+      "grad_norm": 2.803284880960482,
+      "learning_rate": 1.902788780687919e-05,
+      "loss": 0.8759,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3348125478194338,
+      "grad_norm": 2.6200835706423558,
+      "learning_rate": 1.9026821702300942e-05,
+      "loss": 0.8095,
+      "step": 2188
+    },
+    {
+      "epoch": 0.3349655700076511,
+      "grad_norm": 2.5244225371364255,
+      "learning_rate": 1.902575504335026e-05,
+      "loss": 0.8257,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3351185921958684,
+      "grad_norm": 2.272322663580629,
+      "learning_rate": 1.9024687830092653e-05,
+      "loss": 0.8683,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3352716143840857,
+      "grad_norm": 2.3507940993495855,
+      "learning_rate": 1.902362006259366e-05,
+      "loss": 0.7174,
+      "step": 2191
+    },
+    {
+      "epoch": 0.335424636572303,
+      "grad_norm": 2.3419236603411786,
+      "learning_rate": 1.9022551740918858e-05,
+      "loss": 0.8454,
+      "step": 2192
+    },
+    {
+      "epoch": 0.33557765876052026,
+      "grad_norm": 2.2407962986631427,
+      "learning_rate": 1.9021482865133857e-05,
+      "loss": 0.8129,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3357306809487376,
+      "grad_norm": 2.829969199508418,
+      "learning_rate": 1.90204134353043e-05,
+      "loss": 0.9139,
+      "step": 2194
+    },
+    {
+      "epoch": 0.33588370313695487,
+      "grad_norm": 2.570297450213411,
+      "learning_rate": 1.901934345149587e-05,
+      "loss": 0.8215,
+      "step": 2195
+    },
+    {
+      "epoch": 0.33603672532517215,
+      "grad_norm": 2.1742508587047356,
+      "learning_rate": 1.9018272913774272e-05,
+      "loss": 0.7921,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3361897475133894,
+      "grad_norm": 2.5516210884648354,
+      "learning_rate": 1.9017201822205257e-05,
+      "loss": 0.777,
+      "step": 2197
+    },
+    {
+      "epoch": 0.33634276970160676,
+      "grad_norm": 2.5965934611015524,
+      "learning_rate": 1.9016130176854595e-05,
+      "loss": 0.8242,
+      "step": 2198
+    },
+    {
+      "epoch": 0.33649579188982404,
+      "grad_norm": 2.633683261887388,
+      "learning_rate": 1.9015057977788115e-05,
+      "loss": 0.8029,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3366488140780413,
+      "grad_norm": 2.6200741185136533,
+      "learning_rate": 1.9013985225071657e-05,
+      "loss": 0.8098,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3368018362662586,
+      "grad_norm": 2.385459716174837,
+      "learning_rate": 1.90129119187711e-05,
+      "loss": 0.9182,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3369548584544759,
+      "grad_norm": 2.8745143161019677,
+      "learning_rate": 1.901183805895237e-05,
+      "loss": 0.833,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3371078806426932,
+      "grad_norm": 2.316707026669433,
+      "learning_rate": 1.90107636456814e-05,
+      "loss": 0.8279,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3372609028309105,
+      "grad_norm": 2.7035829713270187,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.8465,
+      "step": 2204
+    },
+    {
+      "epoch": 0.33741392501912776,
+      "grad_norm": 2.43981629724686,
+      "learning_rate": 1.9008613159046755e-05,
+      "loss": 0.8019,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3375669472073451,
+      "grad_norm": 2.2730341237566107,
+      "learning_rate": 1.900753708581514e-05,
+      "loss": 0.8162,
+      "step": 2206
+    },
+    {
+      "epoch": 0.33771996939556237,
+      "grad_norm": 2.2234217115960955,
+      "learning_rate": 1.9006460459395436e-05,
+      "loss": 0.7875,
+      "step": 2207
+    },
+    {
+      "epoch": 0.33787299158377965,
+      "grad_norm": 2.8015510519189264,
+      "learning_rate": 1.9005383279853763e-05,
+      "loss": 0.9112,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3380260137719969,
+      "grad_norm": 2.5704163897021255,
+      "learning_rate": 1.900430554725627e-05,
+      "loss": 0.904,
+      "step": 2209
+    },
+    {
+      "epoch": 0.33817903596021426,
+      "grad_norm": 2.305524302766331,
+      "learning_rate": 1.9003227261669154e-05,
+      "loss": 0.8197,
+      "step": 2210
+    },
+    {
+      "epoch": 0.33833205814843154,
+      "grad_norm": 2.5338540294728613,
+      "learning_rate": 1.9002148423158626e-05,
+      "loss": 0.7963,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3384850803366488,
+      "grad_norm": 2.767802849237396,
+      "learning_rate": 1.9001069031790948e-05,
+      "loss": 0.8008,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3386381025248661,
+      "grad_norm": 2.6297059670318785,
+      "learning_rate": 1.899998908763241e-05,
+      "loss": 0.8164,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3387911247130834,
+      "grad_norm": 2.5239777756138095,
+      "learning_rate": 1.8998908590749333e-05,
+      "loss": 0.7567,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3389441469013007,
+      "grad_norm": 2.6121503690567582,
+      "learning_rate": 1.8997827541208073e-05,
+      "loss": 0.7942,
+      "step": 2215
+    },
+    {
+      "epoch": 0.339097169089518,
+      "grad_norm": 2.513509488316226,
+      "learning_rate": 1.8996745939075024e-05,
+      "loss": 0.8487,
+      "step": 2216
+    },
+    {
+      "epoch": 0.33925019127773526,
+      "grad_norm": 2.406167309519931,
+      "learning_rate": 1.8995663784416613e-05,
+      "loss": 0.8935,
+      "step": 2217
+    },
+    {
+      "epoch": 0.33940321346595254,
+      "grad_norm": 2.2466325362978314,
+      "learning_rate": 1.8994581077299303e-05,
+      "loss": 0.764,
+      "step": 2218
+    },
+    {
+      "epoch": 0.33955623565416987,
+      "grad_norm": 2.3061195054332155,
+      "learning_rate": 1.8993497817789574e-05,
+      "loss": 0.7542,
+      "step": 2219
+    },
+    {
+      "epoch": 0.33970925784238715,
+      "grad_norm": 2.8382048184352957,
+      "learning_rate": 1.8992414005953964e-05,
+      "loss": 0.9037,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3398622800306044,
+      "grad_norm": 2.3287211280392777,
+      "learning_rate": 1.8991329641859035e-05,
+      "loss": 0.8013,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3400153022188217,
+      "grad_norm": 2.299742943168751,
+      "learning_rate": 1.8990244725571374e-05,
+      "loss": 0.7663,
+      "step": 2222
+    },
+    {
+      "epoch": 0.34016832440703904,
+      "grad_norm": 2.6471103760984764,
+      "learning_rate": 1.8989159257157616e-05,
+      "loss": 0.9033,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3403213465952563,
+      "grad_norm": 2.280805425315999,
+      "learning_rate": 1.8988073236684422e-05,
+      "loss": 0.8295,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3404743687834736,
+      "grad_norm": 2.141508514611218,
+      "learning_rate": 1.898698666421849e-05,
+      "loss": 0.7833,
+      "step": 2225
+    },
+    {
+      "epoch": 0.34062739097169087,
+      "grad_norm": 2.7267105114369685,
+      "learning_rate": 1.8985899539826547e-05,
+      "loss": 0.7954,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3407804131599082,
+      "grad_norm": 2.2761348371968584,
+      "learning_rate": 1.898481186357536e-05,
+      "loss": 0.7791,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3409334353481255,
+      "grad_norm": 2.311320182693125,
+      "learning_rate": 1.8983723635531733e-05,
+      "loss": 0.7803,
+      "step": 2228
+    },
+    {
+      "epoch": 0.34108645753634276,
+      "grad_norm": 2.416674743767647,
+      "learning_rate": 1.898263485576249e-05,
+      "loss": 0.7712,
+      "step": 2229
+    },
+    {
+      "epoch": 0.34123947972456004,
+      "grad_norm": 2.41675063992554,
+      "learning_rate": 1.8981545524334497e-05,
+      "loss": 0.8321,
+      "step": 2230
+    },
+    {
+      "epoch": 0.34139250191277737,
+      "grad_norm": 2.374216831787798,
+      "learning_rate": 1.8980455641314658e-05,
+      "loss": 0.759,
+      "step": 2231
+    },
+    {
+      "epoch": 0.34154552410099465,
+      "grad_norm": 2.5464606847654414,
+      "learning_rate": 1.897936520676991e-05,
+      "loss": 0.8562,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3416985462892119,
+      "grad_norm": 2.138460875090068,
+      "learning_rate": 1.8978274220767218e-05,
+      "loss": 0.7936,
+      "step": 2233
+    },
+    {
+      "epoch": 0.3418515684774292,
+      "grad_norm": 2.430427227076344,
+      "learning_rate": 1.8977182683373577e-05,
+      "loss": 0.7254,
+      "step": 2234
+    },
+    {
+      "epoch": 0.34200459066564654,
+      "grad_norm": 2.4379531574708335,
+      "learning_rate": 1.8976090594656034e-05,
+      "loss": 0.7694,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3421576128538638,
+      "grad_norm": 2.3778128523783413,
+      "learning_rate": 1.8974997954681652e-05,
+      "loss": 0.7247,
+      "step": 2236
+    },
+    {
+      "epoch": 0.3423106350420811,
+      "grad_norm": 2.6390102931039183,
+      "learning_rate": 1.8973904763517534e-05,
+      "loss": 0.8236,
+      "step": 2237
+    },
+    {
+      "epoch": 0.34246365723029837,
+      "grad_norm": 2.298857942777125,
+      "learning_rate": 1.8972811021230816e-05,
+      "loss": 0.7822,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3426166794185157,
+      "grad_norm": 2.1998620792842702,
+      "learning_rate": 1.8971716727888677e-05,
+      "loss": 0.7806,
+      "step": 2239
+    },
+    {
+      "epoch": 0.342769701606733,
+      "grad_norm": 2.656918161912584,
+      "learning_rate": 1.897062188355831e-05,
+      "loss": 0.9258,
+      "step": 2240
+    },
+    {
+      "epoch": 0.34292272379495026,
+      "grad_norm": 2.4514776639915032,
+      "learning_rate": 1.8969526488306965e-05,
+      "loss": 0.771,
+      "step": 2241
+    },
+    {
+      "epoch": 0.34307574598316753,
+      "grad_norm": 2.404064889745198,
+      "learning_rate": 1.8968430542201905e-05,
+      "loss": 0.7684,
+      "step": 2242
+    },
+    {
+      "epoch": 0.34322876817138487,
+      "grad_norm": 2.289169317066614,
+      "learning_rate": 1.8967334045310445e-05,
+      "loss": 0.9253,
+      "step": 2243
+    },
+    {
+      "epoch": 0.34338179035960215,
+      "grad_norm": 2.3330229020319337,
+      "learning_rate": 1.896623699769992e-05,
+      "loss": 0.8232,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3435348125478194,
+      "grad_norm": 2.513086642418052,
+      "learning_rate": 1.89651393994377e-05,
+      "loss": 0.8884,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3436878347360367,
+      "grad_norm": 2.8654016636145743,
+      "learning_rate": 1.8964041250591203e-05,
+      "loss": 0.873,
+      "step": 2246
+    },
+    {
+      "epoch": 0.34384085692425403,
+      "grad_norm": 2.565973612887949,
+      "learning_rate": 1.8962942551227862e-05,
+      "loss": 0.7987,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3439938791124713,
+      "grad_norm": 2.36055185125566,
+      "learning_rate": 1.896184330141516e-05,
+      "loss": 0.7303,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3441469013006886,
+      "grad_norm": 2.3238221779857984,
+      "learning_rate": 1.8960743501220597e-05,
+      "loss": 0.797,
+      "step": 2249
+    },
+    {
+      "epoch": 0.34429992348890587,
+      "grad_norm": 2.455737367611592,
+      "learning_rate": 1.8959643150711723e-05,
+      "loss": 0.759,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3444529456771232,
+      "grad_norm": 2.3782474528258106,
+      "learning_rate": 1.8958542249956116e-05,
+      "loss": 0.8592,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3446059678653405,
+      "grad_norm": 2.278193327077733,
+      "learning_rate": 1.8957440799021377e-05,
+      "loss": 0.7989,
+      "step": 2252
+    },
+    {
+      "epoch": 0.34475899005355776,
+      "grad_norm": 2.299702774333914,
+      "learning_rate": 1.895633879797516e-05,
+      "loss": 0.8959,
+      "step": 2253
+    },
+    {
+      "epoch": 0.34491201224177503,
+      "grad_norm": 2.050241850848599,
+      "learning_rate": 1.8955236246885143e-05,
+      "loss": 0.6608,
+      "step": 2254
+    },
+    {
+      "epoch": 0.34506503442999237,
+      "grad_norm": 2.1911195420404725,
+      "learning_rate": 1.8954133145819028e-05,
+      "loss": 0.696,
+      "step": 2255
+    },
+    {
+      "epoch": 0.34521805661820965,
+      "grad_norm": 2.505315912925584,
+      "learning_rate": 1.8953029494844575e-05,
+      "loss": 0.8803,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3453710788064269,
+      "grad_norm": 2.3649033111770716,
+      "learning_rate": 1.8951925294029553e-05,
+      "loss": 0.7616,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3455241009946442,
+      "grad_norm": 2.3759627706360256,
+      "learning_rate": 1.895082054344178e-05,
+      "loss": 0.8723,
+      "step": 2258
+    },
+    {
+      "epoch": 0.34567712318286153,
+      "grad_norm": 2.4769807539083124,
+      "learning_rate": 1.89497152431491e-05,
+      "loss": 0.8755,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3458301453710788,
+      "grad_norm": 2.372527770894994,
+      "learning_rate": 1.8948609393219394e-05,
+      "loss": 0.8229,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3459831675592961,
+      "grad_norm": 2.30265390487159,
+      "learning_rate": 1.894750299372058e-05,
+      "loss": 0.7476,
+      "step": 2261
+    },
+    {
+      "epoch": 0.34613618974751337,
+      "grad_norm": 2.355500546095671,
+      "learning_rate": 1.8946396044720607e-05,
+      "loss": 0.732,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3462892119357307,
+      "grad_norm": 2.324299123570443,
+      "learning_rate": 1.8945288546287452e-05,
+      "loss": 0.803,
+      "step": 2263
+    },
+    {
+      "epoch": 0.346442234123948,
+      "grad_norm": 2.4044763714157695,
+      "learning_rate": 1.894418049848913e-05,
+      "loss": 0.7983,
+      "step": 2264
+    },
+    {
+      "epoch": 0.34659525631216526,
+      "grad_norm": 2.6759518579838932,
+      "learning_rate": 1.8943071901393698e-05,
+      "loss": 0.8131,
+      "step": 2265
+    },
+    {
+      "epoch": 0.34674827850038253,
+      "grad_norm": 2.3991278259603535,
+      "learning_rate": 1.894196275506923e-05,
+      "loss": 0.8602,
+      "step": 2266
+    },
+    {
+      "epoch": 0.34690130068859987,
+      "grad_norm": 2.2733901741363542,
+      "learning_rate": 1.8940853059583853e-05,
+      "loss": 0.8241,
+      "step": 2267
+    },
+    {
+      "epoch": 0.34705432287681715,
+      "grad_norm": 3.8759891438201906,
+      "learning_rate": 1.8939742815005716e-05,
+      "loss": 0.8396,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3472073450650344,
+      "grad_norm": 2.3324798219158316,
+      "learning_rate": 1.8938632021402994e-05,
+      "loss": 0.676,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3473603672532517,
+      "grad_norm": 2.2829117471511373,
+      "learning_rate": 1.893752067884391e-05,
+      "loss": 0.7266,
+      "step": 2270
+    },
+    {
+      "epoch": 0.34751338944146903,
+      "grad_norm": 2.544061089313835,
+      "learning_rate": 1.8936408787396724e-05,
+      "loss": 0.7567,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3476664116296863,
+      "grad_norm": 2.8639141424306698,
+      "learning_rate": 1.893529634712971e-05,
+      "loss": 0.9786,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3478194338179036,
+      "grad_norm": 2.2951994146776244,
+      "learning_rate": 1.8934183358111194e-05,
+      "loss": 0.7286,
+      "step": 2273
+    },
+    {
+      "epoch": 0.34797245600612087,
+      "grad_norm": 2.514494070203749,
+      "learning_rate": 1.8933069820409528e-05,
+      "loss": 0.8458,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3481254781943382,
+      "grad_norm": 2.3009159384379516,
+      "learning_rate": 1.8931955734093096e-05,
+      "loss": 0.809,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3482785003825555,
+      "grad_norm": 2.6169208483293187,
+      "learning_rate": 1.893084109923032e-05,
+      "loss": 0.7312,
+      "step": 2276
+    },
+    {
+      "epoch": 0.34843152257077276,
+      "grad_norm": 2.318572642711375,
+      "learning_rate": 1.8929725915889656e-05,
+      "loss": 0.8057,
+      "step": 2277
+    },
+    {
+      "epoch": 0.34858454475899003,
+      "grad_norm": 2.4384080072672605,
+      "learning_rate": 1.892861018413959e-05,
+      "loss": 0.7845,
+      "step": 2278
+    },
+    {
+      "epoch": 0.34873756694720737,
+      "grad_norm": 2.233659076191945,
+      "learning_rate": 1.8927493904048645e-05,
+      "loss": 0.7987,
+      "step": 2279
+    },
+    {
+      "epoch": 0.34889058913542464,
+      "grad_norm": 2.8132353615209555,
+      "learning_rate": 1.8926377075685375e-05,
+      "loss": 0.792,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3490436113236419,
+      "grad_norm": 2.745124612350017,
+      "learning_rate": 1.8925259699118363e-05,
+      "loss": 0.8503,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3491966335118592,
+      "grad_norm": 2.3941958793093483,
+      "learning_rate": 1.8924141774416243e-05,
+      "loss": 0.7677,
+      "step": 2282
+    },
+    {
+      "epoch": 0.34934965570007653,
+      "grad_norm": 2.3379365924269493,
+      "learning_rate": 1.892302330164766e-05,
+      "loss": 0.8122,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3495026778882938,
+      "grad_norm": 2.533816103656603,
+      "learning_rate": 1.8921904280881316e-05,
+      "loss": 0.7815,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3496557000765111,
+      "grad_norm": 2.6931588712567947,
+      "learning_rate": 1.8920784712185925e-05,
+      "loss": 0.9097,
+      "step": 2285
+    },
+    {
+      "epoch": 0.34980872226472837,
+      "grad_norm": 2.270706751809004,
+      "learning_rate": 1.8919664595630244e-05,
+      "loss": 0.8721,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3499617444529457,
+      "grad_norm": 2.645794975569509,
+      "learning_rate": 1.8918543931283065e-05,
+      "loss": 0.8858,
+      "step": 2287
+    },
+    {
+      "epoch": 0.350114766641163,
+      "grad_norm": 2.747285576887863,
+      "learning_rate": 1.891742271921322e-05,
+      "loss": 0.87,
+      "step": 2288
+    },
+    {
+      "epoch": 0.35026778882938026,
+      "grad_norm": 2.363498122115401,
+      "learning_rate": 1.8916300959489555e-05,
+      "loss": 0.7907,
+      "step": 2289
+    },
+    {
+      "epoch": 0.35042081101759753,
+      "grad_norm": 2.4364836344114784,
+      "learning_rate": 1.8915178652180968e-05,
+      "loss": 0.8606,
+      "step": 2290
+    },
+    {
+      "epoch": 0.35057383320581487,
+      "grad_norm": 2.353223078511656,
+      "learning_rate": 1.8914055797356385e-05,
+      "loss": 0.7986,
+      "step": 2291
+    },
+    {
+      "epoch": 0.35072685539403214,
+      "grad_norm": 2.7270978738275513,
+      "learning_rate": 1.8912932395084763e-05,
+      "loss": 0.7931,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3508798775822494,
+      "grad_norm": 2.537289091735032,
+      "learning_rate": 1.8911808445435097e-05,
+      "loss": 0.9002,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3510328997704667,
+      "grad_norm": 2.6426827278918488,
+      "learning_rate": 1.8910683948476407e-05,
+      "loss": 0.7942,
+      "step": 2294
+    },
+    {
+      "epoch": 0.35118592195868403,
+      "grad_norm": 2.2600513381334273,
+      "learning_rate": 1.890955890427776e-05,
+      "loss": 0.7864,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3513389441469013,
+      "grad_norm": 2.1081811704121085,
+      "learning_rate": 1.8908433312908245e-05,
+      "loss": 0.6954,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3514919663351186,
+      "grad_norm": 2.75647258716163,
+      "learning_rate": 1.8907307174436993e-05,
+      "loss": 0.833,
+      "step": 2297
+    },
+    {
+      "epoch": 0.35164498852333587,
+      "grad_norm": 2.4321268972579246,
+      "learning_rate": 1.8906180488933157e-05,
+      "loss": 0.8105,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3517980107115532,
+      "grad_norm": 2.1495453410926504,
+      "learning_rate": 1.8905053256465938e-05,
+      "loss": 0.664,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3519510328997705,
+      "grad_norm": 2.6676960585373695,
+      "learning_rate": 1.8903925477104564e-05,
+      "loss": 0.8758,
+      "step": 2300
+    },
+    {
+      "epoch": 0.35210405508798776,
+      "grad_norm": 2.7198945884747974,
+      "learning_rate": 1.8902797150918296e-05,
+      "loss": 0.7659,
+      "step": 2301
+    },
+    {
+      "epoch": 0.35225707727620503,
+      "grad_norm": 2.6559777133855365,
+      "learning_rate": 1.8901668277976418e-05,
+      "loss": 0.8424,
+      "step": 2302
+    },
+    {
+      "epoch": 0.35241009946442237,
+      "grad_norm": 2.434513815774383,
+      "learning_rate": 1.8900538858348278e-05,
+      "loss": 0.8141,
+      "step": 2303
+    },
+    {
+      "epoch": 0.35256312165263964,
+      "grad_norm": 2.484738285350647,
+      "learning_rate": 1.889940889210322e-05,
+      "loss": 0.735,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3527161438408569,
+      "grad_norm": 2.2508751836804537,
+      "learning_rate": 1.889827837931065e-05,
+      "loss": 0.7989,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3528691660290742,
+      "grad_norm": 2.640756293380206,
+      "learning_rate": 1.8897147320039996e-05,
+      "loss": 0.881,
+      "step": 2306
+    },
+    {
+      "epoch": 0.35302218821729153,
+      "grad_norm": 2.5263231329934652,
+      "learning_rate": 1.8896015714360717e-05,
+      "loss": 0.8671,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3531752104055088,
+      "grad_norm": 2.085279323673372,
+      "learning_rate": 1.8894883562342312e-05,
+      "loss": 0.6073,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3533282325937261,
+      "grad_norm": 2.691825024482245,
+      "learning_rate": 1.8893750864054313e-05,
+      "loss": 0.8882,
+      "step": 2309
+    },
+    {
+      "epoch": 0.35348125478194337,
+      "grad_norm": 2.4796532361562282,
+      "learning_rate": 1.8892617619566277e-05,
+      "loss": 0.859,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3536342769701607,
+      "grad_norm": 2.498491204421089,
+      "learning_rate": 1.8891483828947805e-05,
+      "loss": 0.8075,
+      "step": 2311
+    },
+    {
+      "epoch": 0.353787299158378,
+      "grad_norm": 2.4618340934724237,
+      "learning_rate": 1.889034949226853e-05,
+      "loss": 0.8162,
+      "step": 2312
+    },
+    {
+      "epoch": 0.35394032134659525,
+      "grad_norm": 2.3787322282171117,
+      "learning_rate": 1.8889214609598113e-05,
+      "loss": 0.7916,
+      "step": 2313
+    },
+    {
+      "epoch": 0.35409334353481253,
+      "grad_norm": 2.331498599753699,
+      "learning_rate": 1.888807918100625e-05,
+      "loss": 0.8632,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35424636572302987,
+      "grad_norm": 2.221565693142028,
+      "learning_rate": 1.8886943206562677e-05,
+      "loss": 0.7716,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35439938791124714,
+      "grad_norm": 2.7336828188903874,
+      "learning_rate": 1.8885806686337154e-05,
+      "loss": 0.8538,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3545524100994644,
+      "grad_norm": 2.3932636463702948,
+      "learning_rate": 1.888466962039948e-05,
+      "loss": 0.8395,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3547054322876817,
+      "grad_norm": 2.2898696428789775,
+      "learning_rate": 1.888353200881949e-05,
+      "loss": 0.7866,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35485845447589903,
+      "grad_norm": 2.2783308963748494,
+      "learning_rate": 1.8882393851667046e-05,
+      "loss": 0.7497,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3550114766641163,
+      "grad_norm": 2.3945871355573454,
+      "learning_rate": 1.8881255149012047e-05,
+      "loss": 0.8037,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3551644988523336,
+      "grad_norm": 2.4599488406377965,
+      "learning_rate": 1.8880115900924426e-05,
+      "loss": 0.7507,
+      "step": 2321
+    },
+    {
+      "epoch": 0.35531752104055087,
+      "grad_norm": 2.4065925989291492,
+      "learning_rate": 1.8878976107474147e-05,
+      "loss": 0.7945,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3554705432287682,
+      "grad_norm": 2.4302388243500834,
+      "learning_rate": 1.8877835768731212e-05,
+      "loss": 0.8201,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3556235654169855,
+      "grad_norm": 2.446682716888137,
+      "learning_rate": 1.8876694884765648e-05,
+      "loss": 0.8339,
+      "step": 2324
+    },
+    {
+      "epoch": 0.35577658760520275,
+      "grad_norm": 2.656460918351645,
+      "learning_rate": 1.887555345564753e-05,
+      "loss": 0.8646,
+      "step": 2325
+    },
+    {
+      "epoch": 0.35592960979342003,
+      "grad_norm": 2.349546610049587,
+      "learning_rate": 1.887441148144695e-05,
+      "loss": 0.7222,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3560826319816373,
+      "grad_norm": 2.404762586158525,
+      "learning_rate": 1.887326896223405e-05,
+      "loss": 0.8084,
+      "step": 2327
+    },
+    {
+      "epoch": 0.35623565416985464,
+      "grad_norm": 2.283494276451758,
+      "learning_rate": 1.8872125898078985e-05,
+      "loss": 0.841,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3563886763580719,
+      "grad_norm": 2.448512269421924,
+      "learning_rate": 1.8870982289051962e-05,
+      "loss": 0.7663,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3565416985462892,
+      "grad_norm": 2.457783473444631,
+      "learning_rate": 1.8869838135223214e-05,
+      "loss": 0.763,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3566947207345065,
+      "grad_norm": 2.5155960596034785,
+      "learning_rate": 1.8868693436663006e-05,
+      "loss": 0.8754,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3568477429227238,
+      "grad_norm": 2.4245646747227085,
+      "learning_rate": 1.886754819344164e-05,
+      "loss": 0.7733,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3570007651109411,
+      "grad_norm": 2.2376909756083623,
+      "learning_rate": 1.8866402405629447e-05,
+      "loss": 0.7319,
+      "step": 2333
+    },
+    {
+      "epoch": 0.35715378729915837,
+      "grad_norm": 2.4584252279555714,
+      "learning_rate": 1.88652560732968e-05,
+      "loss": 0.8691,
+      "step": 2334
+    },
+    {
+      "epoch": 0.35730680948737564,
+      "grad_norm": 2.541807107424281,
+      "learning_rate": 1.8864109196514096e-05,
+      "loss": 0.8073,
+      "step": 2335
+    },
+    {
+      "epoch": 0.357459831675593,
+      "grad_norm": 2.359184716636419,
+      "learning_rate": 1.886296177535177e-05,
+      "loss": 0.7375,
+      "step": 2336
+    },
+    {
+      "epoch": 0.35761285386381025,
+      "grad_norm": 2.3320992648966636,
+      "learning_rate": 1.8861813809880288e-05,
+      "loss": 0.7953,
+      "step": 2337
+    },
+    {
+      "epoch": 0.35776587605202753,
+      "grad_norm": 2.304009796345526,
+      "learning_rate": 1.8860665300170155e-05,
+      "loss": 0.7176,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3579188982402448,
+      "grad_norm": 2.3503014104713094,
+      "learning_rate": 1.8859516246291897e-05,
+      "loss": 0.8045,
+      "step": 2339
+    },
+    {
+      "epoch": 0.35807192042846214,
+      "grad_norm": 2.619508152477919,
+      "learning_rate": 1.8858366648316093e-05,
+      "loss": 0.908,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3582249426166794,
+      "grad_norm": 2.4429041886742264,
+      "learning_rate": 1.8857216506313334e-05,
+      "loss": 0.7526,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3583779648048967,
+      "grad_norm": 2.543918354148765,
+      "learning_rate": 1.8856065820354267e-05,
+      "loss": 0.9139,
+      "step": 2342
+    },
+    {
+      "epoch": 0.358530986993114,
+      "grad_norm": 2.6126350593443632,
+      "learning_rate": 1.8854914590509547e-05,
+      "loss": 0.8782,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3586840091813313,
+      "grad_norm": 2.6110295147807583,
+      "learning_rate": 1.8853762816849882e-05,
+      "loss": 0.8034,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3588370313695486,
+      "grad_norm": 2.3093166827506866,
+      "learning_rate": 1.885261049944601e-05,
+      "loss": 0.8519,
+      "step": 2345
+    },
+    {
+      "epoch": 0.35899005355776586,
+      "grad_norm": 2.355282542968486,
+      "learning_rate": 1.885145763836869e-05,
+      "loss": 0.8291,
+      "step": 2346
+    },
+    {
+      "epoch": 0.35914307574598314,
+      "grad_norm": 2.347727191149938,
+      "learning_rate": 1.8850304233688735e-05,
+      "loss": 0.8327,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3592960979342005,
+      "grad_norm": 2.7381027335841504,
+      "learning_rate": 1.884915028547697e-05,
+      "loss": 0.8587,
+      "step": 2348
+    },
+    {
+      "epoch": 0.35944912012241775,
+      "grad_norm": 2.2608152358156977,
+      "learning_rate": 1.884799579380427e-05,
+      "loss": 0.7851,
+      "step": 2349
+    },
+    {
+      "epoch": 0.35960214231063503,
+      "grad_norm": 2.5461028673938064,
+      "learning_rate": 1.8846840758741533e-05,
+      "loss": 0.8457,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3597551644988523,
+      "grad_norm": 2.197749699940499,
+      "learning_rate": 1.88456851803597e-05,
+      "loss": 0.8232,
+      "step": 2351
+    },
+    {
+      "epoch": 0.35990818668706964,
+      "grad_norm": 2.2081723097087247,
+      "learning_rate": 1.8844529058729732e-05,
+      "loss": 0.8364,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3600612088752869,
+      "grad_norm": 2.2167658053229014,
+      "learning_rate": 1.8843372393922633e-05,
+      "loss": 0.7587,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3602142310635042,
+      "grad_norm": 2.5104046616375006,
+      "learning_rate": 1.8842215186009447e-05,
+      "loss": 0.8437,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3603672532517215,
+      "grad_norm": 2.4725723695877857,
+      "learning_rate": 1.884105743506123e-05,
+      "loss": 0.7279,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3605202754399388,
+      "grad_norm": 2.457372115979769,
+      "learning_rate": 1.883989914114909e-05,
+      "loss": 0.827,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3606732976281561,
+      "grad_norm": 2.410585651428132,
+      "learning_rate": 1.883874030434416e-05,
+      "loss": 0.7795,
+      "step": 2357
+    },
+    {
+      "epoch": 0.36082631981637336,
+      "grad_norm": 2.4141145834331135,
+      "learning_rate": 1.8837580924717614e-05,
+      "loss": 0.777,
+      "step": 2358
+    },
+    {
+      "epoch": 0.36097934200459064,
+      "grad_norm": 2.2617813201912673,
+      "learning_rate": 1.883642100234065e-05,
+      "loss": 0.7614,
+      "step": 2359
+    },
+    {
+      "epoch": 0.361132364192808,
+      "grad_norm": 2.2966321276793735,
+      "learning_rate": 1.88352605372845e-05,
+      "loss": 0.7941,
+      "step": 2360
+    },
+    {
+      "epoch": 0.36128538638102525,
+      "grad_norm": 2.110102263249786,
+      "learning_rate": 1.8834099529620442e-05,
+      "loss": 0.6854,
+      "step": 2361
+    },
+    {
+      "epoch": 0.36143840856924253,
+      "grad_norm": 2.7367322503438754,
+      "learning_rate": 1.883293797941977e-05,
+      "loss": 0.8223,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3615914307574598,
+      "grad_norm": 2.709621371396735,
+      "learning_rate": 1.8831775886753822e-05,
+      "loss": 0.7787,
+      "step": 2363
+    },
+    {
+      "epoch": 0.36174445294567714,
+      "grad_norm": 2.440509867786845,
+      "learning_rate": 1.883061325169397e-05,
+      "loss": 0.8569,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3618974751338944,
+      "grad_norm": 2.504540793639551,
+      "learning_rate": 1.882945007431161e-05,
+      "loss": 0.7796,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3620504973221117,
+      "grad_norm": 2.6996923030735096,
+      "learning_rate": 1.882828635467818e-05,
+      "loss": 0.8674,
+      "step": 2366
+    },
+    {
+      "epoch": 0.362203519510329,
+      "grad_norm": 2.8062029268635267,
+      "learning_rate": 1.8827122092865147e-05,
+      "loss": 0.9018,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3623565416985463,
+      "grad_norm": 2.393990682979781,
+      "learning_rate": 1.8825957288944017e-05,
+      "loss": 0.8371,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3625095638867636,
+      "grad_norm": 2.302645835173265,
+      "learning_rate": 1.882479194298632e-05,
+      "loss": 0.7554,
+      "step": 2369
+    },
+    {
+      "epoch": 0.36266258607498086,
+      "grad_norm": 2.42744947114573,
+      "learning_rate": 1.882362605506363e-05,
+      "loss": 0.8365,
+      "step": 2370
+    },
+    {
+      "epoch": 0.36281560826319814,
+      "grad_norm": 2.541019733868751,
+      "learning_rate": 1.8822459625247544e-05,
+      "loss": 0.9123,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3629686304514155,
+      "grad_norm": 1.9423446059365113,
+      "learning_rate": 1.8821292653609698e-05,
+      "loss": 0.6642,
+      "step": 2372
+    },
+    {
+      "epoch": 0.36312165263963275,
+      "grad_norm": 2.662428175077465,
+      "learning_rate": 1.8820125140221762e-05,
+      "loss": 0.934,
+      "step": 2373
+    },
+    {
+      "epoch": 0.36327467482785003,
+      "grad_norm": 2.3608324759387105,
+      "learning_rate": 1.8818957085155437e-05,
+      "loss": 0.8484,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3634276970160673,
+      "grad_norm": 2.43897927593089,
+      "learning_rate": 1.881778848848246e-05,
+      "loss": 0.9507,
+      "step": 2375
+    },
+    {
+      "epoch": 0.36358071920428464,
+      "grad_norm": 2.2402163894990617,
+      "learning_rate": 1.881661935027459e-05,
+      "loss": 0.7647,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3637337413925019,
+      "grad_norm": 2.4107917616959322,
+      "learning_rate": 1.881544967060364e-05,
+      "loss": 0.8487,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3638867635807192,
+      "grad_norm": 2.291052028819227,
+      "learning_rate": 1.881427944954144e-05,
+      "loss": 0.765,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3640397857689365,
+      "grad_norm": 2.33261212765457,
+      "learning_rate": 1.8813108687159852e-05,
+      "loss": 0.7599,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3641928079571538,
+      "grad_norm": 2.3741799543583517,
+      "learning_rate": 1.8811937383530786e-05,
+      "loss": 0.9039,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3643458301453711,
+      "grad_norm": 2.697621817440523,
+      "learning_rate": 1.8810765538726175e-05,
+      "loss": 0.8248,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36449885233358836,
+      "grad_norm": 2.317964315825029,
+      "learning_rate": 1.880959315281798e-05,
+      "loss": 0.6964,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36465187452180564,
+      "grad_norm": 2.2431665552546427,
+      "learning_rate": 1.8808420225878207e-05,
+      "loss": 0.8107,
+      "step": 2383
+    },
+    {
+      "epoch": 0.364804896710023,
+      "grad_norm": 2.4004243356517687,
+      "learning_rate": 1.8807246757978892e-05,
+      "loss": 0.7904,
+      "step": 2384
+    },
+    {
+      "epoch": 0.36495791889824025,
+      "grad_norm": 2.7187226403241267,
+      "learning_rate": 1.8806072749192096e-05,
+      "loss": 0.9686,
+      "step": 2385
+    },
+    {
+      "epoch": 0.36511094108645753,
+      "grad_norm": 2.8316217465961344,
+      "learning_rate": 1.8804898199589925e-05,
+      "loss": 0.8722,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3652639632746748,
+      "grad_norm": 2.3925054684622915,
+      "learning_rate": 1.8803723109244513e-05,
+      "loss": 0.9331,
+      "step": 2387
+    },
+    {
+      "epoch": 0.36541698546289214,
+      "grad_norm": 2.4682270072369636,
+      "learning_rate": 1.880254747822802e-05,
+      "loss": 0.7587,
+      "step": 2388
+    },
+    {
+      "epoch": 0.3655700076511094,
+      "grad_norm": 2.3194225175157093,
+      "learning_rate": 1.8801371306612652e-05,
+      "loss": 0.8076,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3657230298393267,
+      "grad_norm": 2.4399649171572486,
+      "learning_rate": 1.880019459447064e-05,
+      "loss": 0.8232,
+      "step": 2390
+    },
+    {
+      "epoch": 0.365876052027544,
+      "grad_norm": 2.3831569144484126,
+      "learning_rate": 1.8799017341874255e-05,
+      "loss": 0.7088,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3660290742157613,
+      "grad_norm": 2.3319903377417854,
+      "learning_rate": 1.879783954889579e-05,
+      "loss": 0.7227,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3661820964039786,
+      "grad_norm": 2.495408648387555,
+      "learning_rate": 1.879666121560758e-05,
+      "loss": 0.8645,
+      "step": 2393
+    },
+    {
+      "epoch": 0.36633511859219586,
+      "grad_norm": 2.008295423288152,
+      "learning_rate": 1.8795482342081995e-05,
+      "loss": 0.7434,
+      "step": 2394
+    },
+    {
+      "epoch": 0.36648814078041314,
+      "grad_norm": 2.3226281552138484,
+      "learning_rate": 1.879430292839143e-05,
+      "loss": 0.7134,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3666411629686305,
+      "grad_norm": 2.7280664670536074,
+      "learning_rate": 1.8793122974608318e-05,
+      "loss": 0.8971,
+      "step": 2396
+    },
+    {
+      "epoch": 0.36679418515684775,
+      "grad_norm": 2.417968916079334,
+      "learning_rate": 1.8791942480805127e-05,
+      "loss": 0.9469,
+      "step": 2397
+    },
+    {
+      "epoch": 0.36694720734506503,
+      "grad_norm": 2.3321600957081703,
+      "learning_rate": 1.8790761447054353e-05,
+      "loss": 0.8373,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3671002295332823,
+      "grad_norm": 2.9661860123961263,
+      "learning_rate": 1.8789579873428526e-05,
+      "loss": 0.8054,
+      "step": 2399
+    },
+    {
+      "epoch": 0.36725325172149964,
+      "grad_norm": 2.8557791816349822,
+      "learning_rate": 1.878839776000022e-05,
+      "loss": 0.8268,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3674062739097169,
+      "grad_norm": 2.5578308951105853,
+      "learning_rate": 1.8787215106842022e-05,
+      "loss": 0.7479,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3675592960979342,
+      "grad_norm": 2.9152984669056323,
+      "learning_rate": 1.878603191402657e-05,
+      "loss": 0.8661,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3677123182861515,
+      "grad_norm": 2.3207274090171843,
+      "learning_rate": 1.878484818162653e-05,
+      "loss": 0.805,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3678653404743688,
+      "grad_norm": 2.3006454257014557,
+      "learning_rate": 1.878366390971459e-05,
+      "loss": 0.7455,
+      "step": 2404
+    },
+    {
+      "epoch": 0.3680183626625861,
+      "grad_norm": 2.3525946228739545,
+      "learning_rate": 1.8782479098363494e-05,
+      "loss": 0.8276,
+      "step": 2405
+    },
+    {
+      "epoch": 0.36817138485080336,
+      "grad_norm": 2.4311958649745504,
+      "learning_rate": 1.8781293747645998e-05,
+      "loss": 0.8442,
+      "step": 2406
+    },
+    {
+      "epoch": 0.36832440703902064,
+      "grad_norm": 2.2341059242635364,
+      "learning_rate": 1.87801078576349e-05,
+      "loss": 0.7892,
+      "step": 2407
+    },
+    {
+      "epoch": 0.368477429227238,
+      "grad_norm": 2.4789478015956288,
+      "learning_rate": 1.877892142840303e-05,
+      "loss": 0.7717,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36863045141545525,
+      "grad_norm": 2.3251520271339743,
+      "learning_rate": 1.877773446002325e-05,
+      "loss": 0.9049,
+      "step": 2409
+    },
+    {
+      "epoch": 0.36878347360367253,
+      "grad_norm": 2.4284067911124843,
+      "learning_rate": 1.8776546952568456e-05,
+      "loss": 0.8632,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3689364957918898,
+      "grad_norm": 2.407972854722055,
+      "learning_rate": 1.8775358906111586e-05,
+      "loss": 0.6936,
+      "step": 2411
+    },
+    {
+      "epoch": 0.36908951798010714,
+      "grad_norm": 2.2389103408646136,
+      "learning_rate": 1.8774170320725592e-05,
+      "loss": 0.6321,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3692425401683244,
+      "grad_norm": 2.5869540597200396,
+      "learning_rate": 1.8772981196483474e-05,
+      "loss": 0.9377,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3693955623565417,
+      "grad_norm": 2.752160756736885,
+      "learning_rate": 1.8771791533458264e-05,
+      "loss": 0.8435,
+      "step": 2414
+    },
+    {
+      "epoch": 0.369548584544759,
+      "grad_norm": 2.522539089724177,
+      "learning_rate": 1.8770601331723017e-05,
+      "loss": 0.8243,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3697016067329763,
+      "grad_norm": 2.810550793971596,
+      "learning_rate": 1.8769410591350832e-05,
+      "loss": 0.9468,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3698546289211936,
+      "grad_norm": 2.469864839731524,
+      "learning_rate": 1.8768219312414838e-05,
+      "loss": 0.7879,
+      "step": 2417
+    },
+    {
+      "epoch": 0.37000765110941086,
+      "grad_norm": 2.271877626883605,
+      "learning_rate": 1.8767027494988194e-05,
+      "loss": 0.8244,
+      "step": 2418
+    },
+    {
+      "epoch": 0.37016067329762814,
+      "grad_norm": 2.6112175112151044,
+      "learning_rate": 1.8765835139144093e-05,
+      "loss": 0.775,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3703136954858455,
+      "grad_norm": 2.266093145167797,
+      "learning_rate": 1.8764642244955767e-05,
+      "loss": 0.7782,
+      "step": 2420
+    },
+    {
+      "epoch": 0.37046671767406275,
+      "grad_norm": 2.3075593915599724,
+      "learning_rate": 1.876344881249647e-05,
+      "loss": 0.8186,
+      "step": 2421
+    },
+    {
+      "epoch": 0.37061973986228003,
+      "grad_norm": 2.560441628674782,
+      "learning_rate": 1.8762254841839503e-05,
+      "loss": 0.9051,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3707727620504973,
+      "grad_norm": 2.542019562348531,
+      "learning_rate": 1.8761060333058183e-05,
+      "loss": 0.8128,
+      "step": 2423
+    },
+    {
+      "epoch": 0.37092578423871464,
+      "grad_norm": 2.6372207763697197,
+      "learning_rate": 1.8759865286225878e-05,
+      "loss": 0.8735,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3710788064269319,
+      "grad_norm": 2.5326515666594593,
+      "learning_rate": 1.8758669701415975e-05,
+      "loss": 0.7602,
+      "step": 2425
+    },
+    {
+      "epoch": 0.3712318286151492,
+      "grad_norm": 2.368763311753748,
+      "learning_rate": 1.8757473578701902e-05,
+      "loss": 0.7612,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3713848508033665,
+      "grad_norm": 2.128504255109629,
+      "learning_rate": 1.8756276918157118e-05,
+      "loss": 0.7579,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3715378729915838,
+      "grad_norm": 2.471878320157333,
+      "learning_rate": 1.8755079719855113e-05,
+      "loss": 0.8326,
+      "step": 2428
+    },
+    {
+      "epoch": 0.3716908951798011,
+      "grad_norm": 2.7949857804868428,
+      "learning_rate": 1.8753881983869414e-05,
+      "loss": 0.854,
+      "step": 2429
+    },
+    {
+      "epoch": 0.37184391736801836,
+      "grad_norm": 2.7751068465784123,
+      "learning_rate": 1.875268371027357e-05,
+      "loss": 0.8798,
+      "step": 2430
+    },
+    {
+      "epoch": 0.37199693955623564,
+      "grad_norm": 2.7780322221768863,
+      "learning_rate": 1.8751484899141185e-05,
+      "loss": 0.8468,
+      "step": 2431
+    },
+    {
+      "epoch": 0.372149961744453,
+      "grad_norm": 2.4139073729391702,
+      "learning_rate": 1.8750285550545877e-05,
+      "loss": 0.807,
+      "step": 2432
+    },
+    {
+      "epoch": 0.37230298393267025,
+      "grad_norm": 2.5798622524914503,
+      "learning_rate": 1.87490856645613e-05,
+      "loss": 0.7849,
+      "step": 2433
+    },
+    {
+      "epoch": 0.37245600612088753,
+      "grad_norm": 2.3845029258074173,
+      "learning_rate": 1.874788524126115e-05,
+      "loss": 0.7316,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3726090283091048,
+      "grad_norm": 2.2918954995904763,
+      "learning_rate": 1.8746684280719137e-05,
+      "loss": 0.7754,
+      "step": 2435
+    },
+    {
+      "epoch": 0.37276205049732214,
+      "grad_norm": 2.8944457285794725,
+      "learning_rate": 1.874548278300903e-05,
+      "loss": 0.8968,
+      "step": 2436
+    },
+    {
+      "epoch": 0.3729150726855394,
+      "grad_norm": 2.512484041846354,
+      "learning_rate": 1.8744280748204614e-05,
+      "loss": 0.8099,
+      "step": 2437
+    },
+    {
+      "epoch": 0.3730680948737567,
+      "grad_norm": 2.652834197290464,
+      "learning_rate": 1.8743078176379707e-05,
+      "loss": 0.841,
+      "step": 2438
+    },
+    {
+      "epoch": 0.373221117061974,
+      "grad_norm": 2.784629393063227,
+      "learning_rate": 1.8741875067608167e-05,
+      "loss": 0.8983,
+      "step": 2439
+    },
+    {
+      "epoch": 0.37337413925019125,
+      "grad_norm": 2.676099241658654,
+      "learning_rate": 1.874067142196388e-05,
+      "loss": 0.8388,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3735271614384086,
+      "grad_norm": 2.5217409667731534,
+      "learning_rate": 1.8739467239520767e-05,
+      "loss": 0.818,
+      "step": 2441
+    },
+    {
+      "epoch": 0.37368018362662586,
+      "grad_norm": 2.1378603895356583,
+      "learning_rate": 1.8738262520352783e-05,
+      "loss": 0.7787,
+      "step": 2442
+    },
+    {
+      "epoch": 0.37383320581484314,
+      "grad_norm": 2.5527568920487917,
+      "learning_rate": 1.8737057264533915e-05,
+      "loss": 0.8348,
+      "step": 2443
+    },
+    {
+      "epoch": 0.3739862280030604,
+      "grad_norm": 2.624247786380471,
+      "learning_rate": 1.873585147213818e-05,
+      "loss": 0.7936,
+      "step": 2444
+    },
+    {
+      "epoch": 0.37413925019127775,
+      "grad_norm": 2.3965554595283836,
+      "learning_rate": 1.873464514323963e-05,
+      "loss": 0.7522,
+      "step": 2445
+    },
+    {
+      "epoch": 0.37429227237949503,
+      "grad_norm": 2.5187732475788396,
+      "learning_rate": 1.873343827791235e-05,
+      "loss": 0.753,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3744452945677123,
+      "grad_norm": 2.415007932936088,
+      "learning_rate": 1.8732230876230463e-05,
+      "loss": 0.865,
+      "step": 2447
+    },
+    {
+      "epoch": 0.3745983167559296,
+      "grad_norm": 2.411623125230876,
+      "learning_rate": 1.8731022938268114e-05,
+      "loss": 0.7445,
+      "step": 2448
+    },
+    {
+      "epoch": 0.3747513389441469,
+      "grad_norm": 2.504678058017117,
+      "learning_rate": 1.8729814464099492e-05,
+      "loss": 0.8019,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3749043611323642,
+      "grad_norm": 2.4121077808768683,
+      "learning_rate": 1.8728605453798816e-05,
+      "loss": 0.8243,
+      "step": 2450
+    },
+    {
+      "epoch": 0.37505738332058147,
+      "grad_norm": 2.3792207835192927,
+      "learning_rate": 1.8727395907440328e-05,
+      "loss": 0.7216,
+      "step": 2451
+    },
+    {
+      "epoch": 0.37521040550879875,
+      "grad_norm": 2.7011902930310323,
+      "learning_rate": 1.8726185825098317e-05,
+      "loss": 0.834,
+      "step": 2452
+    },
+    {
+      "epoch": 0.3753634276970161,
+      "grad_norm": 2.37612362801207,
+      "learning_rate": 1.87249752068471e-05,
+      "loss": 0.7309,
+      "step": 2453
+    },
+    {
+      "epoch": 0.37551644988523336,
+      "grad_norm": 2.5429947552774097,
+      "learning_rate": 1.8723764052761018e-05,
+      "loss": 0.7774,
+      "step": 2454
+    },
+    {
+      "epoch": 0.37566947207345064,
+      "grad_norm": 2.399185893707355,
+      "learning_rate": 1.8722552362914463e-05,
+      "loss": 0.8234,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3758224942616679,
+      "grad_norm": 2.495485499967576,
+      "learning_rate": 1.872134013738184e-05,
+      "loss": 0.7609,
+      "step": 2456
+    },
+    {
+      "epoch": 0.37597551644988525,
+      "grad_norm": 2.352447545297215,
+      "learning_rate": 1.8720127376237606e-05,
+      "loss": 0.7429,
+      "step": 2457
+    },
+    {
+      "epoch": 0.3761285386381025,
+      "grad_norm": 2.040482423517943,
+      "learning_rate": 1.871891407955623e-05,
+      "loss": 0.7081,
+      "step": 2458
+    },
+    {
+      "epoch": 0.3762815608263198,
+      "grad_norm": 2.4635143143098253,
+      "learning_rate": 1.8717700247412237e-05,
+      "loss": 0.8347,
+      "step": 2459
+    },
+    {
+      "epoch": 0.3764345830145371,
+      "grad_norm": 2.4888516381763712,
+      "learning_rate": 1.8716485879880165e-05,
+      "loss": 0.7666,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3765876052027544,
+      "grad_norm": 2.3853061470198953,
+      "learning_rate": 1.87152709770346e-05,
+      "loss": 0.7858,
+      "step": 2461
+    },
+    {
+      "epoch": 0.3767406273909717,
+      "grad_norm": 2.8397222753085583,
+      "learning_rate": 1.8714055538950145e-05,
+      "loss": 0.7969,
+      "step": 2462
+    },
+    {
+      "epoch": 0.37689364957918897,
+      "grad_norm": 2.2192558442628387,
+      "learning_rate": 1.8712839565701455e-05,
+      "loss": 0.893,
+      "step": 2463
+    },
+    {
+      "epoch": 0.37704667176740625,
+      "grad_norm": 2.1752035742885645,
+      "learning_rate": 1.87116230573632e-05,
+      "loss": 0.7395,
+      "step": 2464
+    },
+    {
+      "epoch": 0.3771996939556236,
+      "grad_norm": 2.4877185407030167,
+      "learning_rate": 1.8710406014010094e-05,
+      "loss": 0.8267,
+      "step": 2465
+    },
+    {
+      "epoch": 0.37735271614384086,
+      "grad_norm": 2.5339436958201067,
+      "learning_rate": 1.870918843571688e-05,
+      "loss": 0.7972,
+      "step": 2466
+    },
+    {
+      "epoch": 0.37750573833205814,
+      "grad_norm": 2.557231748291859,
+      "learning_rate": 1.8707970322558328e-05,
+      "loss": 0.8761,
+      "step": 2467
+    },
+    {
+      "epoch": 0.3776587605202754,
+      "grad_norm": 2.2694866860882574,
+      "learning_rate": 1.8706751674609258e-05,
+      "loss": 0.8071,
+      "step": 2468
+    },
+    {
+      "epoch": 0.37781178270849275,
+      "grad_norm": 2.3669762986337104,
+      "learning_rate": 1.8705532491944505e-05,
+      "loss": 0.8245,
+      "step": 2469
+    },
+    {
+      "epoch": 0.37796480489671,
+      "grad_norm": 2.3039821470771846,
+      "learning_rate": 1.8704312774638945e-05,
+      "loss": 0.6956,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3781178270849273,
+      "grad_norm": 2.438094533399992,
+      "learning_rate": 1.8703092522767487e-05,
+      "loss": 0.8558,
+      "step": 2471
+    },
+    {
+      "epoch": 0.3782708492731446,
+      "grad_norm": 2.690744376860686,
+      "learning_rate": 1.870187173640507e-05,
+      "loss": 0.988,
+      "step": 2472
+    },
+    {
+      "epoch": 0.3784238714613619,
+      "grad_norm": 2.410059836592625,
+      "learning_rate": 1.870065041562667e-05,
+      "loss": 0.7595,
+      "step": 2473
+    },
+    {
+      "epoch": 0.3785768936495792,
+      "grad_norm": 2.3714407252038523,
+      "learning_rate": 1.8699428560507288e-05,
+      "loss": 0.8068,
+      "step": 2474
+    },
+    {
+      "epoch": 0.37872991583779647,
+      "grad_norm": 2.87103136320842,
+      "learning_rate": 1.8698206171121963e-05,
+      "loss": 0.7212,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37888293802601375,
+      "grad_norm": 2.419031011584517,
+      "learning_rate": 1.8696983247545776e-05,
+      "loss": 0.8506,
+      "step": 2476
+    },
+    {
+      "epoch": 0.3790359602142311,
+      "grad_norm": 2.3590944560091667,
+      "learning_rate": 1.869575978985382e-05,
+      "loss": 0.9662,
+      "step": 2477
+    },
+    {
+      "epoch": 0.37918898240244836,
+      "grad_norm": 2.4656630507450314,
+      "learning_rate": 1.8694535798121244e-05,
+      "loss": 0.8289,
+      "step": 2478
+    },
+    {
+      "epoch": 0.37934200459066564,
+      "grad_norm": 2.295792371436799,
+      "learning_rate": 1.8693311272423204e-05,
+      "loss": 0.8399,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3794950267788829,
+      "grad_norm": 2.961244599837281,
+      "learning_rate": 1.8692086212834912e-05,
+      "loss": 0.8367,
+      "step": 2480
+    },
+    {
+      "epoch": 0.37964804896710025,
+      "grad_norm": 2.6693969074114365,
+      "learning_rate": 1.8690860619431604e-05,
+      "loss": 0.9139,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3798010711553175,
+      "grad_norm": 2.635077494495987,
+      "learning_rate": 1.8689634492288547e-05,
+      "loss": 0.7848,
+      "step": 2482
+    },
+    {
+      "epoch": 0.3799540933435348,
+      "grad_norm": 2.561238780281626,
+      "learning_rate": 1.8688407831481037e-05,
+      "loss": 0.9398,
+      "step": 2483
+    },
+    {
+      "epoch": 0.3801071155317521,
+      "grad_norm": 2.3136951444719007,
+      "learning_rate": 1.8687180637084418e-05,
+      "loss": 0.7758,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3802601377199694,
+      "grad_norm": 2.504907279331964,
+      "learning_rate": 1.868595290917405e-05,
+      "loss": 0.8735,
+      "step": 2485
+    },
+    {
+      "epoch": 0.3804131599081867,
+      "grad_norm": 2.2085626782374854,
+      "learning_rate": 1.8684724647825333e-05,
+      "loss": 0.74,
+      "step": 2486
+    },
+    {
+      "epoch": 0.38056618209640397,
+      "grad_norm": 2.6191916388341125,
+      "learning_rate": 1.8683495853113703e-05,
+      "loss": 0.787,
+      "step": 2487
+    },
+    {
+      "epoch": 0.38071920428462125,
+      "grad_norm": 2.602426306454107,
+      "learning_rate": 1.868226652511462e-05,
+      "loss": 0.7568,
+      "step": 2488
+    },
+    {
+      "epoch": 0.3808722264728386,
+      "grad_norm": 2.3846879215375076,
+      "learning_rate": 1.8681036663903585e-05,
+      "loss": 0.7223,
+      "step": 2489
+    },
+    {
+      "epoch": 0.38102524866105586,
+      "grad_norm": 2.2431261468268686,
+      "learning_rate": 1.867980626955613e-05,
+      "loss": 0.7147,
+      "step": 2490
+    },
+    {
+      "epoch": 0.38117827084927314,
+      "grad_norm": 2.5618430976248265,
+      "learning_rate": 1.8678575342147815e-05,
+      "loss": 0.8167,
+      "step": 2491
+    },
+    {
+      "epoch": 0.3813312930374904,
+      "grad_norm": 2.7244295536027194,
+      "learning_rate": 1.867734388175424e-05,
+      "loss": 0.7683,
+      "step": 2492
+    },
+    {
+      "epoch": 0.38148431522570775,
+      "grad_norm": 2.6472413539846857,
+      "learning_rate": 1.8676111888451028e-05,
+      "loss": 0.8227,
+      "step": 2493
+    },
+    {
+      "epoch": 0.381637337413925,
+      "grad_norm": 2.308898420544816,
+      "learning_rate": 1.8674879362313843e-05,
+      "loss": 0.8609,
+      "step": 2494
+    },
+    {
+      "epoch": 0.3817903596021423,
+      "grad_norm": 2.563563794604606,
+      "learning_rate": 1.8673646303418382e-05,
+      "loss": 0.8591,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3819433817903596,
+      "grad_norm": 2.327587598177729,
+      "learning_rate": 1.867241271184037e-05,
+      "loss": 0.7396,
+      "step": 2496
+    },
+    {
+      "epoch": 0.3820964039785769,
+      "grad_norm": 2.2452666295710717,
+      "learning_rate": 1.8671178587655567e-05,
+      "loss": 0.8333,
+      "step": 2497
+    },
+    {
+      "epoch": 0.3822494261667942,
+      "grad_norm": 2.358122242033259,
+      "learning_rate": 1.8669943930939763e-05,
+      "loss": 0.715,
+      "step": 2498
+    },
+    {
+      "epoch": 0.38240244835501147,
+      "grad_norm": 2.242928896018524,
+      "learning_rate": 1.866870874176879e-05,
+      "loss": 0.8568,
+      "step": 2499
+    },
+    {
+      "epoch": 0.38255547054322875,
+      "grad_norm": 2.393411774568154,
+      "learning_rate": 1.8667473020218497e-05,
+      "loss": 0.8591,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3827084927314461,
+      "grad_norm": 2.508761300085195,
+      "learning_rate": 1.8666236766364778e-05,
+      "loss": 0.6883,
+      "step": 2501
+    },
+    {
+      "epoch": 0.38286151491966336,
+      "grad_norm": 2.273285381962697,
+      "learning_rate": 1.8664999980283558e-05,
+      "loss": 0.7873,
+      "step": 2502
+    },
+    {
+      "epoch": 0.38301453710788064,
+      "grad_norm": 2.5456143107904405,
+      "learning_rate": 1.8663762662050793e-05,
+      "loss": 0.7534,
+      "step": 2503
+    },
+    {
+      "epoch": 0.3831675592960979,
+      "grad_norm": 2.5340046766612954,
+      "learning_rate": 1.8662524811742467e-05,
+      "loss": 0.9405,
+      "step": 2504
+    },
+    {
+      "epoch": 0.38332058148431525,
+      "grad_norm": 2.2030346255099857,
+      "learning_rate": 1.8661286429434607e-05,
+      "loss": 0.8104,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3834736036725325,
+      "grad_norm": 2.0899223105089977,
+      "learning_rate": 1.8660047515203263e-05,
+      "loss": 0.7298,
+      "step": 2506
+    },
+    {
+      "epoch": 0.3836266258607498,
+      "grad_norm": 2.091576745913007,
+      "learning_rate": 1.8658808069124523e-05,
+      "loss": 0.7127,
+      "step": 2507
+    },
+    {
+      "epoch": 0.3837796480489671,
+      "grad_norm": 2.442967562778881,
+      "learning_rate": 1.8657568091274503e-05,
+      "loss": 0.928,
+      "step": 2508
+    },
+    {
+      "epoch": 0.3839326702371844,
+      "grad_norm": 2.6676064813396496,
+      "learning_rate": 1.8656327581729364e-05,
+      "loss": 0.8753,
+      "step": 2509
+    },
+    {
+      "epoch": 0.3840856924254017,
+      "grad_norm": 2.9341603959582567,
+      "learning_rate": 1.865508654056528e-05,
+      "loss": 0.9654,
+      "step": 2510
+    },
+    {
+      "epoch": 0.38423871461361897,
+      "grad_norm": 2.4344257952073156,
+      "learning_rate": 1.8653844967858475e-05,
+      "loss": 0.7526,
+      "step": 2511
+    },
+    {
+      "epoch": 0.38439173680183625,
+      "grad_norm": 2.361905617278103,
+      "learning_rate": 1.8652602863685195e-05,
+      "loss": 0.8353,
+      "step": 2512
+    },
+    {
+      "epoch": 0.3845447589900536,
+      "grad_norm": 2.218282200611466,
+      "learning_rate": 1.8651360228121724e-05,
+      "loss": 0.6705,
+      "step": 2513
+    },
+    {
+      "epoch": 0.38469778117827086,
+      "grad_norm": 2.5345150182492953,
+      "learning_rate": 1.8650117061244378e-05,
+      "loss": 0.8431,
+      "step": 2514
+    },
+    {
+      "epoch": 0.38485080336648814,
+      "grad_norm": 2.2846439326524712,
+      "learning_rate": 1.8648873363129502e-05,
+      "loss": 0.7474,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3850038255547054,
+      "grad_norm": 2.3803085886744992,
+      "learning_rate": 1.864762913385348e-05,
+      "loss": 0.7584,
+      "step": 2516
+    },
+    {
+      "epoch": 0.38515684774292275,
+      "grad_norm": 2.257800923753178,
+      "learning_rate": 1.864638437349272e-05,
+      "loss": 0.751,
+      "step": 2517
+    },
+    {
+      "epoch": 0.38530986993114,
+      "grad_norm": 2.4758559918625016,
+      "learning_rate": 1.8645139082123675e-05,
+      "loss": 0.8829,
+      "step": 2518
+    },
+    {
+      "epoch": 0.3854628921193573,
+      "grad_norm": 2.508363591784869,
+      "learning_rate": 1.8643893259822817e-05,
+      "loss": 0.8091,
+      "step": 2519
+    },
+    {
+      "epoch": 0.3856159143075746,
+      "grad_norm": 2.344123531418125,
+      "learning_rate": 1.864264690666666e-05,
+      "loss": 0.8086,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3857689364957919,
+      "grad_norm": 2.6911838867682327,
+      "learning_rate": 1.8641400022731746e-05,
+      "loss": 0.8829,
+      "step": 2521
+    },
+    {
+      "epoch": 0.3859219586840092,
+      "grad_norm": 2.454881184279251,
+      "learning_rate": 1.864015260809465e-05,
+      "loss": 0.8059,
+      "step": 2522
+    },
+    {
+      "epoch": 0.38607498087222647,
+      "grad_norm": 2.3972818653488353,
+      "learning_rate": 1.8638904662831985e-05,
+      "loss": 0.7309,
+      "step": 2523
+    },
+    {
+      "epoch": 0.38622800306044375,
+      "grad_norm": 2.2703191459881573,
+      "learning_rate": 1.8637656187020385e-05,
+      "loss": 0.7365,
+      "step": 2524
+    },
+    {
+      "epoch": 0.3863810252486611,
+      "grad_norm": 2.480521094293675,
+      "learning_rate": 1.863640718073653e-05,
+      "loss": 0.7481,
+      "step": 2525
+    },
+    {
+      "epoch": 0.38653404743687836,
+      "grad_norm": 2.32410916843462,
+      "learning_rate": 1.8635157644057124e-05,
+      "loss": 0.7864,
+      "step": 2526
+    },
+    {
+      "epoch": 0.38668706962509564,
+      "grad_norm": 2.1555637796345617,
+      "learning_rate": 1.8633907577058905e-05,
+      "loss": 0.7126,
+      "step": 2527
+    },
+    {
+      "epoch": 0.3868400918133129,
+      "grad_norm": 2.0035595797324537,
+      "learning_rate": 1.8632656979818645e-05,
+      "loss": 0.7262,
+      "step": 2528
+    },
+    {
+      "epoch": 0.38699311400153025,
+      "grad_norm": 2.500240065738295,
+      "learning_rate": 1.863140585241315e-05,
+      "loss": 0.8232,
+      "step": 2529
+    },
+    {
+      "epoch": 0.3871461361897475,
+      "grad_norm": 2.182399464955277,
+      "learning_rate": 1.8630154194919256e-05,
+      "loss": 0.7309,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3872991583779648,
+      "grad_norm": 2.547552862072076,
+      "learning_rate": 1.8628902007413835e-05,
+      "loss": 0.9328,
+      "step": 2531
+    },
+    {
+      "epoch": 0.3874521805661821,
+      "grad_norm": 2.4185769090684404,
+      "learning_rate": 1.8627649289973776e-05,
+      "loss": 0.8189,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3876052027543994,
+      "grad_norm": 2.220438695870478,
+      "learning_rate": 1.862639604267603e-05,
+      "loss": 0.7749,
+      "step": 2533
+    },
+    {
+      "epoch": 0.3877582249426167,
+      "grad_norm": 2.578841434734547,
+      "learning_rate": 1.8625142265597556e-05,
+      "loss": 0.8771,
+      "step": 2534
+    },
+    {
+      "epoch": 0.38791124713083397,
+      "grad_norm": 2.2401417785438382,
+      "learning_rate": 1.862388795881535e-05,
+      "loss": 0.7575,
+      "step": 2535
+    },
+    {
+      "epoch": 0.38806426931905125,
+      "grad_norm": 2.478671799326241,
+      "learning_rate": 1.862263312240645e-05,
+      "loss": 0.7846,
+      "step": 2536
+    },
+    {
+      "epoch": 0.3882172915072686,
+      "grad_norm": 2.2303927085740565,
+      "learning_rate": 1.8621377756447918e-05,
+      "loss": 0.8156,
+      "step": 2537
+    },
+    {
+      "epoch": 0.38837031369548586,
+      "grad_norm": 2.247486841768642,
+      "learning_rate": 1.8620121861016854e-05,
+      "loss": 0.8024,
+      "step": 2538
+    },
+    {
+      "epoch": 0.38852333588370314,
+      "grad_norm": 2.41891025833812,
+      "learning_rate": 1.861886543619038e-05,
+      "loss": 0.9377,
+      "step": 2539
+    },
+    {
+      "epoch": 0.3886763580719204,
+      "grad_norm": 2.490839576225125,
+      "learning_rate": 1.8617608482045662e-05,
+      "loss": 0.7237,
+      "step": 2540
+    },
+    {
+      "epoch": 0.38882938026013775,
+      "grad_norm": 2.204949493052091,
+      "learning_rate": 1.8616350998659895e-05,
+      "loss": 0.7152,
+      "step": 2541
+    },
+    {
+      "epoch": 0.388982402448355,
+      "grad_norm": 2.520927297137216,
+      "learning_rate": 1.8615092986110308e-05,
+      "loss": 0.5829,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3891354246365723,
+      "grad_norm": 2.0925039293849834,
+      "learning_rate": 1.861383444447416e-05,
+      "loss": 0.6984,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3892884468247896,
+      "grad_norm": 2.425641347817572,
+      "learning_rate": 1.8612575373828735e-05,
+      "loss": 0.8305,
+      "step": 2544
+    },
+    {
+      "epoch": 0.3894414690130069,
+      "grad_norm": 2.573051475202199,
+      "learning_rate": 1.8611315774251367e-05,
+      "loss": 0.7923,
+      "step": 2545
+    },
+    {
+      "epoch": 0.3895944912012242,
+      "grad_norm": 2.5068821947821625,
+      "learning_rate": 1.861005564581941e-05,
+      "loss": 0.9357,
+      "step": 2546
+    },
+    {
+      "epoch": 0.38974751338944147,
+      "grad_norm": 2.324998200221432,
+      "learning_rate": 1.8608794988610256e-05,
+      "loss": 0.8629,
+      "step": 2547
+    },
+    {
+      "epoch": 0.38990053557765875,
+      "grad_norm": 2.52960849031247,
+      "learning_rate": 1.8607533802701318e-05,
+      "loss": 0.7974,
+      "step": 2548
+    },
+    {
+      "epoch": 0.390053557765876,
+      "grad_norm": 2.4815636587464494,
+      "learning_rate": 1.860627208817006e-05,
+      "loss": 0.8583,
+      "step": 2549
+    },
+    {
+      "epoch": 0.39020657995409336,
+      "grad_norm": 2.2653029816484582,
+      "learning_rate": 1.8605009845093964e-05,
+      "loss": 0.7958,
+      "step": 2550
+    },
+    {
+      "epoch": 0.39035960214231064,
+      "grad_norm": 2.3990609889203807,
+      "learning_rate": 1.8603747073550552e-05,
+      "loss": 0.81,
+      "step": 2551
+    },
+    {
+      "epoch": 0.3905126243305279,
+      "grad_norm": 2.32455958546504,
+      "learning_rate": 1.8602483773617373e-05,
+      "loss": 0.7526,
+      "step": 2552
+    },
+    {
+      "epoch": 0.3906656465187452,
+      "grad_norm": 2.3427643871420774,
+      "learning_rate": 1.8601219945372015e-05,
+      "loss": 0.8194,
+      "step": 2553
+    },
+    {
+      "epoch": 0.3908186687069625,
+      "grad_norm": 2.2513504889406226,
+      "learning_rate": 1.8599955588892086e-05,
+      "loss": 0.7728,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3909716908951798,
+      "grad_norm": 2.338401329885293,
+      "learning_rate": 1.8598690704255245e-05,
+      "loss": 0.8088,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3911247130833971,
+      "grad_norm": 2.3014418358202473,
+      "learning_rate": 1.859742529153917e-05,
+      "loss": 0.7908,
+      "step": 2556
+    },
+    {
+      "epoch": 0.39127773527161436,
+      "grad_norm": 2.6220498441628957,
+      "learning_rate": 1.8596159350821573e-05,
+      "loss": 0.851,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3914307574598317,
+      "grad_norm": 2.5560864630883477,
+      "learning_rate": 1.8594892882180202e-05,
+      "loss": 0.7645,
+      "step": 2558
+    },
+    {
+      "epoch": 0.39158377964804897,
+      "grad_norm": 2.4441604741310794,
+      "learning_rate": 1.8593625885692835e-05,
+      "loss": 0.8209,
+      "step": 2559
+    },
+    {
+      "epoch": 0.39173680183626625,
+      "grad_norm": 2.394828726374545,
+      "learning_rate": 1.8592358361437287e-05,
+      "loss": 0.7607,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3918898240244835,
+      "grad_norm": 2.7066788691327055,
+      "learning_rate": 1.8591090309491397e-05,
+      "loss": 0.8225,
+      "step": 2561
+    },
+    {
+      "epoch": 0.39204284621270086,
+      "grad_norm": 2.466903893172232,
+      "learning_rate": 1.858982172993304e-05,
+      "loss": 0.7865,
+      "step": 2562
+    },
+    {
+      "epoch": 0.39219586840091814,
+      "grad_norm": 2.673740227798785,
+      "learning_rate": 1.858855262284013e-05,
+      "loss": 0.9003,
+      "step": 2563
+    },
+    {
+      "epoch": 0.3923488905891354,
+      "grad_norm": 2.655379391443632,
+      "learning_rate": 1.8587282988290604e-05,
+      "loss": 0.8515,
+      "step": 2564
+    },
+    {
+      "epoch": 0.3925019127773527,
+      "grad_norm": 2.387993715416807,
+      "learning_rate": 1.8586012826362437e-05,
+      "loss": 0.8344,
+      "step": 2565
+    },
+    {
+      "epoch": 0.39265493496557,
+      "grad_norm": 2.122389308766076,
+      "learning_rate": 1.8584742137133635e-05,
+      "loss": 0.7624,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3928079571537873,
+      "grad_norm": 2.4675672480936175,
+      "learning_rate": 1.8583470920682232e-05,
+      "loss": 0.7786,
+      "step": 2567
+    },
+    {
+      "epoch": 0.3929609793420046,
+      "grad_norm": 2.1616293938457827,
+      "learning_rate": 1.8582199177086302e-05,
+      "loss": 0.682,
+      "step": 2568
+    },
+    {
+      "epoch": 0.39311400153022186,
+      "grad_norm": 2.424779733933877,
+      "learning_rate": 1.8580926906423944e-05,
+      "loss": 0.7539,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3932670237184392,
+      "grad_norm": 2.468301808496714,
+      "learning_rate": 1.8579654108773296e-05,
+      "loss": 0.7523,
+      "step": 2570
+    },
+    {
+      "epoch": 0.39342004590665647,
+      "grad_norm": 2.5668439800269223,
+      "learning_rate": 1.857838078421253e-05,
+      "loss": 0.9001,
+      "step": 2571
+    },
+    {
+      "epoch": 0.39357306809487375,
+      "grad_norm": 2.6861943370437285,
+      "learning_rate": 1.857710693281984e-05,
+      "loss": 1.0394,
+      "step": 2572
+    },
+    {
+      "epoch": 0.393726090283091,
+      "grad_norm": 2.45276904000766,
+      "learning_rate": 1.8575832554673457e-05,
+      "loss": 0.7552,
+      "step": 2573
+    },
+    {
+      "epoch": 0.39387911247130836,
+      "grad_norm": 2.355317126126015,
+      "learning_rate": 1.857455764985165e-05,
+      "loss": 0.723,
+      "step": 2574
+    },
+    {
+      "epoch": 0.39403213465952563,
+      "grad_norm": 2.303305717824098,
+      "learning_rate": 1.8573282218432712e-05,
+      "loss": 0.8183,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3941851568477429,
+      "grad_norm": 2.274204248618728,
+      "learning_rate": 1.8572006260494975e-05,
+      "loss": 0.7922,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3943381790359602,
+      "grad_norm": 2.6013486170787905,
+      "learning_rate": 1.85707297761168e-05,
+      "loss": 0.8621,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3944912012241775,
+      "grad_norm": 2.6464460011195183,
+      "learning_rate": 1.856945276537658e-05,
+      "loss": 1.0042,
+      "step": 2578
+    },
+    {
+      "epoch": 0.3946442234123948,
+      "grad_norm": 2.071644475187324,
+      "learning_rate": 1.856817522835274e-05,
+      "loss": 0.7901,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3947972456006121,
+      "grad_norm": 2.1078648872166936,
+      "learning_rate": 1.8566897165123742e-05,
+      "loss": 0.6951,
+      "step": 2580
+    },
+    {
+      "epoch": 0.39495026778882936,
+      "grad_norm": 2.390748321191043,
+      "learning_rate": 1.8565618575768078e-05,
+      "loss": 0.7759,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3951032899770467,
+      "grad_norm": 2.3383337718182724,
+      "learning_rate": 1.8564339460364268e-05,
+      "loss": 0.7706,
+      "step": 2582
+    },
+    {
+      "epoch": 0.39525631216526397,
+      "grad_norm": 2.2256538171713944,
+      "learning_rate": 1.8563059818990864e-05,
+      "loss": 0.756,
+      "step": 2583
+    },
+    {
+      "epoch": 0.39540933435348125,
+      "grad_norm": 2.35477670073321,
+      "learning_rate": 1.856177965172646e-05,
+      "loss": 0.8469,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3955623565416985,
+      "grad_norm": 2.394424568314491,
+      "learning_rate": 1.8560498958649675e-05,
+      "loss": 0.8801,
+      "step": 2585
+    },
+    {
+      "epoch": 0.39571537872991586,
+      "grad_norm": 2.200020298735911,
+      "learning_rate": 1.8559217739839156e-05,
+      "loss": 0.7756,
+      "step": 2586
+    },
+    {
+      "epoch": 0.39586840091813313,
+      "grad_norm": 2.5743268374527686,
+      "learning_rate": 1.8557935995373593e-05,
+      "loss": 0.8344,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3960214231063504,
+      "grad_norm": 2.5434499031469753,
+      "learning_rate": 1.8556653725331703e-05,
+      "loss": 0.8237,
+      "step": 2588
+    },
+    {
+      "epoch": 0.3961744452945677,
+      "grad_norm": 2.2710097319079243,
+      "learning_rate": 1.8555370929792237e-05,
+      "loss": 0.7575,
+      "step": 2589
+    },
+    {
+      "epoch": 0.396327467482785,
+      "grad_norm": 2.1784171165885837,
+      "learning_rate": 1.8554087608833967e-05,
+      "loss": 0.7892,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3964804896710023,
+      "grad_norm": 2.4856988681527072,
+      "learning_rate": 1.8552803762535717e-05,
+      "loss": 0.9424,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3966335118592196,
+      "grad_norm": 2.7840026809571716,
+      "learning_rate": 1.855151939097633e-05,
+      "loss": 0.8897,
+      "step": 2592
+    },
+    {
+      "epoch": 0.39678653404743686,
+      "grad_norm": 2.5930974653909384,
+      "learning_rate": 1.855023449423468e-05,
+      "loss": 0.8272,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3969395562356542,
+      "grad_norm": 2.3916332977591277,
+      "learning_rate": 1.8548949072389684e-05,
+      "loss": 0.8303,
+      "step": 2594
+    },
+    {
+      "epoch": 0.39709257842387147,
+      "grad_norm": 2.465859918701186,
+      "learning_rate": 1.8547663125520282e-05,
+      "loss": 0.8453,
+      "step": 2595
+    },
+    {
+      "epoch": 0.39724560061208875,
+      "grad_norm": 2.5220271459852586,
+      "learning_rate": 1.8546376653705446e-05,
+      "loss": 0.9473,
+      "step": 2596
+    },
+    {
+      "epoch": 0.397398622800306,
+      "grad_norm": 2.2621031004903163,
+      "learning_rate": 1.8545089657024185e-05,
+      "loss": 0.7321,
+      "step": 2597
+    },
+    {
+      "epoch": 0.39755164498852336,
+      "grad_norm": 2.21402172022864,
+      "learning_rate": 1.8543802135555544e-05,
+      "loss": 0.7898,
+      "step": 2598
+    },
+    {
+      "epoch": 0.39770466717674063,
+      "grad_norm": 2.5160554926745142,
+      "learning_rate": 1.854251408937859e-05,
+      "loss": 0.8523,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3978576893649579,
+      "grad_norm": 2.472537617150156,
+      "learning_rate": 1.8541225518572425e-05,
+      "loss": 0.9072,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3980107115531752,
+      "grad_norm": 2.43355365679199,
+      "learning_rate": 1.8539936423216187e-05,
+      "loss": 0.8955,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3981637337413925,
+      "grad_norm": 2.352095084295794,
+      "learning_rate": 1.8538646803389048e-05,
+      "loss": 0.8204,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3983167559296098,
+      "grad_norm": 2.4611685288216885,
+      "learning_rate": 1.8537356659170204e-05,
+      "loss": 0.8142,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3984697781178271,
+      "grad_norm": 2.186102788808551,
+      "learning_rate": 1.8536065990638884e-05,
+      "loss": 0.7595,
+      "step": 2604
+    },
+    {
+      "epoch": 0.39862280030604436,
+      "grad_norm": 2.202120235361913,
+      "learning_rate": 1.8534774797874363e-05,
+      "loss": 0.7505,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3987758224942617,
+      "grad_norm": 2.362986852327419,
+      "learning_rate": 1.8533483080955938e-05,
+      "loss": 0.7464,
+      "step": 2606
+    },
+    {
+      "epoch": 0.39892884468247897,
+      "grad_norm": 2.200352843576885,
+      "learning_rate": 1.8532190839962927e-05,
+      "loss": 0.696,
+      "step": 2607
+    },
+    {
+      "epoch": 0.39908186687069624,
+      "grad_norm": 2.255800540526346,
+      "learning_rate": 1.85308980749747e-05,
+      "loss": 0.8086,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3992348890589135,
+      "grad_norm": 2.469916004202286,
+      "learning_rate": 1.852960478607065e-05,
+      "loss": 0.8218,
+      "step": 2609
+    },
+    {
+      "epoch": 0.39938791124713086,
+      "grad_norm": 2.2232561093923064,
+      "learning_rate": 1.8528310973330202e-05,
+      "loss": 0.7753,
+      "step": 2610
+    },
+    {
+      "epoch": 0.39954093343534813,
+      "grad_norm": 2.465870849875202,
+      "learning_rate": 1.8527016636832812e-05,
+      "loss": 0.8572,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3996939556235654,
+      "grad_norm": 2.5829792855369487,
+      "learning_rate": 1.8525721776657976e-05,
+      "loss": 0.8348,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3998469778117827,
+      "grad_norm": 2.3325205099538198,
+      "learning_rate": 1.8524426392885214e-05,
+      "loss": 0.8031,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.276647860896121,
+      "learning_rate": 1.8523130485594073e-05,
+      "loss": 0.8267,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4001530221882173,
+      "grad_norm": 2.3928052935724744,
+      "learning_rate": 1.8521834054864153e-05,
+      "loss": 0.8535,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4003060443764346,
+      "grad_norm": 2.511577722772227,
+      "learning_rate": 1.852053710077506e-05,
+      "loss": 0.7447,
+      "step": 2616
+    },
+    {
+      "epoch": 0.40045906656465186,
+      "grad_norm": 2.110248564429131,
+      "learning_rate": 1.8519239623406458e-05,
+      "loss": 0.7667,
+      "step": 2617
+    },
+    {
+      "epoch": 0.4006120887528692,
+      "grad_norm": 2.488627880598713,
+      "learning_rate": 1.8517941622838017e-05,
+      "loss": 0.7103,
+      "step": 2618
+    },
+    {
+      "epoch": 0.40076511094108647,
+      "grad_norm": 2.2009676567345653,
+      "learning_rate": 1.851664309914946e-05,
+      "loss": 0.7582,
+      "step": 2619
+    },
+    {
+      "epoch": 0.40091813312930374,
+      "grad_norm": 2.1310514990703013,
+      "learning_rate": 1.8515344052420537e-05,
+      "loss": 0.7643,
+      "step": 2620
+    },
+    {
+      "epoch": 0.401071155317521,
+      "grad_norm": 2.6259759882870988,
+      "learning_rate": 1.851404448273102e-05,
+      "loss": 0.9047,
+      "step": 2621
+    },
+    {
+      "epoch": 0.40122417750573836,
+      "grad_norm": 2.2943155620505915,
+      "learning_rate": 1.8512744390160723e-05,
+      "loss": 0.8079,
+      "step": 2622
+    },
+    {
+      "epoch": 0.40137719969395563,
+      "grad_norm": 2.4544127384323327,
+      "learning_rate": 1.851144377478949e-05,
+      "loss": 0.8562,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4015302218821729,
+      "grad_norm": 2.3951768753685965,
+      "learning_rate": 1.8510142636697206e-05,
+      "loss": 0.7953,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4016832440703902,
+      "grad_norm": 2.3766515796882186,
+      "learning_rate": 1.8508840975963763e-05,
+      "loss": 0.8557,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4018362662586075,
+      "grad_norm": 2.5023847534931067,
+      "learning_rate": 1.850753879266911e-05,
+      "loss": 0.812,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4019892884468248,
+      "grad_norm": 2.509151267955778,
+      "learning_rate": 1.850623608689322e-05,
+      "loss": 0.7625,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4021423106350421,
+      "grad_norm": 2.4184906288540073,
+      "learning_rate": 1.8504932858716097e-05,
+      "loss": 0.8336,
+      "step": 2628
+    },
+    {
+      "epoch": 0.40229533282325936,
+      "grad_norm": 2.3168938864394515,
+      "learning_rate": 1.8503629108217772e-05,
+      "loss": 0.7943,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4024483550114767,
+      "grad_norm": 2.3915931853920256,
+      "learning_rate": 1.8502324835478316e-05,
+      "loss": 0.8279,
+      "step": 2630
+    },
+    {
+      "epoch": 0.40260137719969397,
+      "grad_norm": 2.2349446691616706,
+      "learning_rate": 1.8501020040577838e-05,
+      "loss": 0.7348,
+      "step": 2631
+    },
+    {
+      "epoch": 0.40275439938791124,
+      "grad_norm": 2.0151409908112594,
+      "learning_rate": 1.8499714723596455e-05,
+      "loss": 0.7296,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4029074215761285,
+      "grad_norm": 2.5360099818874646,
+      "learning_rate": 1.8498408884614343e-05,
+      "loss": 0.8572,
+      "step": 2633
+    },
+    {
+      "epoch": 0.40306044376434585,
+      "grad_norm": 2.164183952890459,
+      "learning_rate": 1.8497102523711698e-05,
+      "loss": 0.6298,
+      "step": 2634
+    },
+    {
+      "epoch": 0.40321346595256313,
+      "grad_norm": 2.4630834283639,
+      "learning_rate": 1.849579564096874e-05,
+      "loss": 0.7419,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4033664881407804,
+      "grad_norm": 2.350449227312488,
+      "learning_rate": 1.849448823646574e-05,
+      "loss": 0.8939,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4035195103289977,
+      "grad_norm": 2.3427406354106886,
+      "learning_rate": 1.8493180310282985e-05,
+      "loss": 0.9118,
+      "step": 2637
+    },
+    {
+      "epoch": 0.403672532517215,
+      "grad_norm": 2.578044045167987,
+      "learning_rate": 1.8491871862500805e-05,
+      "loss": 0.8345,
+      "step": 2638
+    },
+    {
+      "epoch": 0.4038255547054323,
+      "grad_norm": 2.5999091900545612,
+      "learning_rate": 1.849056289319955e-05,
+      "loss": 0.9699,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4039785768936496,
+      "grad_norm": 2.4447201013727304,
+      "learning_rate": 1.8489253402459615e-05,
+      "loss": 0.9392,
+      "step": 2640
+    },
+    {
+      "epoch": 0.40413159908186685,
+      "grad_norm": 2.232214493716397,
+      "learning_rate": 1.848794339036142e-05,
+      "loss": 0.819,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4042846212700842,
+      "grad_norm": 2.3284688831273765,
+      "learning_rate": 1.8486632856985413e-05,
+      "loss": 0.7519,
+      "step": 2642
+    },
+    {
+      "epoch": 0.40443764345830147,
+      "grad_norm": 2.363636249635232,
+      "learning_rate": 1.848532180241208e-05,
+      "loss": 0.8482,
+      "step": 2643
+    },
+    {
+      "epoch": 0.40459066564651874,
+      "grad_norm": 2.2295062565276074,
+      "learning_rate": 1.8484010226721943e-05,
+      "loss": 0.8025,
+      "step": 2644
+    },
+    {
+      "epoch": 0.404743687834736,
+      "grad_norm": 2.358853456902877,
+      "learning_rate": 1.8482698129995552e-05,
+      "loss": 0.7455,
+      "step": 2645
+    },
+    {
+      "epoch": 0.40489671002295335,
+      "grad_norm": 2.150877684604761,
+      "learning_rate": 1.848138551231348e-05,
+      "loss": 0.7321,
+      "step": 2646
+    },
+    {
+      "epoch": 0.40504973221117063,
+      "grad_norm": 2.4426513307013624,
+      "learning_rate": 1.8480072373756344e-05,
+      "loss": 0.8077,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4052027543993879,
+      "grad_norm": 2.2088244537437114,
+      "learning_rate": 1.847875871440479e-05,
+      "loss": 0.7363,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4053557765876052,
+      "grad_norm": 2.4812489057806553,
+      "learning_rate": 1.8477444534339494e-05,
+      "loss": 0.8045,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4055087987758225,
+      "grad_norm": 2.540939066788791,
+      "learning_rate": 1.8476129833641167e-05,
+      "loss": 0.8186,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4056618209640398,
+      "grad_norm": 2.212331345125525,
+      "learning_rate": 1.847481461239055e-05,
+      "loss": 0.809,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4058148431522571,
+      "grad_norm": 2.401064578558019,
+      "learning_rate": 1.847349887066841e-05,
+      "loss": 0.7607,
+      "step": 2652
+    },
+    {
+      "epoch": 0.40596786534047435,
+      "grad_norm": 2.3926401890762485,
+      "learning_rate": 1.8472182608555554e-05,
+      "loss": 0.8287,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4061208875286917,
+      "grad_norm": 2.4175965911225785,
+      "learning_rate": 1.8470865826132823e-05,
+      "loss": 0.9406,
+      "step": 2654
+    },
+    {
+      "epoch": 0.40627390971690897,
+      "grad_norm": 2.4696218970242025,
+      "learning_rate": 1.8469548523481084e-05,
+      "loss": 0.8894,
+      "step": 2655
+    },
+    {
+      "epoch": 0.40642693190512624,
+      "grad_norm": 2.6511801074082726,
+      "learning_rate": 1.846823070068124e-05,
+      "loss": 0.8922,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4065799540933435,
+      "grad_norm": 2.314650500785454,
+      "learning_rate": 1.846691235781422e-05,
+      "loss": 0.7068,
+      "step": 2657
+    },
+    {
+      "epoch": 0.40673297628156085,
+      "grad_norm": 2.4866883666060637,
+      "learning_rate": 1.8465593494960984e-05,
+      "loss": 0.8283,
+      "step": 2658
+    },
+    {
+      "epoch": 0.40688599846977813,
+      "grad_norm": 2.1590639354930192,
+      "learning_rate": 1.846427411220254e-05,
+      "loss": 0.7625,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4070390206579954,
+      "grad_norm": 2.3846505274108063,
+      "learning_rate": 1.8462954209619906e-05,
+      "loss": 0.9031,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4071920428462127,
+      "grad_norm": 2.2267645532444775,
+      "learning_rate": 1.846163378729415e-05,
+      "loss": 0.7436,
+      "step": 2661
+    },
+    {
+      "epoch": 0.40734506503442997,
+      "grad_norm": 2.398986928697425,
+      "learning_rate": 1.8460312845306355e-05,
+      "loss": 0.8195,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4074980872226473,
+      "grad_norm": 2.61225008925308,
+      "learning_rate": 1.8458991383737658e-05,
+      "loss": 0.7914,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4076511094108646,
+      "grad_norm": 2.356902302425031,
+      "learning_rate": 1.8457669402669204e-05,
+      "loss": 0.8502,
+      "step": 2664
+    },
+    {
+      "epoch": 0.40780413159908185,
+      "grad_norm": 2.560117797837266,
+      "learning_rate": 1.8456346902182186e-05,
+      "loss": 0.781,
+      "step": 2665
+    },
+    {
+      "epoch": 0.40795715378729913,
+      "grad_norm": 2.4632168985463303,
+      "learning_rate": 1.8455023882357828e-05,
+      "loss": 0.8861,
+      "step": 2666
+    },
+    {
+      "epoch": 0.40811017597551646,
+      "grad_norm": 2.3380920013005424,
+      "learning_rate": 1.845370034327737e-05,
+      "loss": 0.7989,
+      "step": 2667
+    },
+    {
+      "epoch": 0.40826319816373374,
+      "grad_norm": 2.718768158973965,
+      "learning_rate": 1.845237628502211e-05,
+      "loss": 0.8724,
+      "step": 2668
+    },
+    {
+      "epoch": 0.408416220351951,
+      "grad_norm": 2.2335921822951366,
+      "learning_rate": 1.8451051707673354e-05,
+      "loss": 0.786,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4085692425401683,
+      "grad_norm": 2.2166311232529865,
+      "learning_rate": 1.8449726611312448e-05,
+      "loss": 0.7835,
+      "step": 2670
+    },
+    {
+      "epoch": 0.40872226472838563,
+      "grad_norm": 2.46656392359808,
+      "learning_rate": 1.844840099602078e-05,
+      "loss": 0.8221,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4088752869166029,
+      "grad_norm": 2.222395842583903,
+      "learning_rate": 1.8447074861879755e-05,
+      "loss": 0.8084,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4090283091048202,
+      "grad_norm": 2.4012567726291816,
+      "learning_rate": 1.8445748208970817e-05,
+      "loss": 0.8552,
+      "step": 2673
+    },
+    {
+      "epoch": 0.40918133129303746,
+      "grad_norm": 2.5844110968126883,
+      "learning_rate": 1.844442103737544e-05,
+      "loss": 0.8058,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4093343534812548,
+      "grad_norm": 2.4874765360529185,
+      "learning_rate": 1.8443093347175136e-05,
+      "loss": 0.8288,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4094873756694721,
+      "grad_norm": 2.3078682180408956,
+      "learning_rate": 1.8441765138451436e-05,
+      "loss": 0.929,
+      "step": 2676
+    },
+    {
+      "epoch": 0.40964039785768935,
+      "grad_norm": 2.2629872570756553,
+      "learning_rate": 1.8440436411285917e-05,
+      "loss": 0.7427,
+      "step": 2677
+    },
+    {
+      "epoch": 0.40979342004590663,
+      "grad_norm": 2.4094935497777974,
+      "learning_rate": 1.843910716576018e-05,
+      "loss": 0.7792,
+      "step": 2678
+    },
+    {
+      "epoch": 0.40994644223412396,
+      "grad_norm": 2.2940908840934267,
+      "learning_rate": 1.8437777401955855e-05,
+      "loss": 0.7945,
+      "step": 2679
+    },
+    {
+      "epoch": 0.41009946442234124,
+      "grad_norm": 2.4885993708595526,
+      "learning_rate": 1.8436447119954614e-05,
+      "loss": 0.8593,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4102524866105585,
+      "grad_norm": 2.193032443148415,
+      "learning_rate": 1.843511631983815e-05,
+      "loss": 0.7174,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4104055087987758,
+      "grad_norm": 2.3278540581719143,
+      "learning_rate": 1.8433785001688193e-05,
+      "loss": 0.7327,
+      "step": 2682
+    },
+    {
+      "epoch": 0.41055853098699313,
+      "grad_norm": 3.241194420703757,
+      "learning_rate": 1.843245316558651e-05,
+      "loss": 0.6707,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4107115531752104,
+      "grad_norm": 2.4359746230564747,
+      "learning_rate": 1.843112081161489e-05,
+      "loss": 0.8612,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4108645753634277,
+      "grad_norm": 2.3943737897219366,
+      "learning_rate": 1.8429787939855153e-05,
+      "loss": 0.695,
+      "step": 2685
+    },
+    {
+      "epoch": 0.41101759755164496,
+      "grad_norm": 2.4256681273927287,
+      "learning_rate": 1.8428454550389167e-05,
+      "loss": 0.7882,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4111706197398623,
+      "grad_norm": 3.3705908991777704,
+      "learning_rate": 1.842712064329881e-05,
+      "loss": 0.8501,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4113236419280796,
+      "grad_norm": 2.419989046073004,
+      "learning_rate": 1.8425786218666008e-05,
+      "loss": 0.9159,
+      "step": 2688
+    },
+    {
+      "epoch": 0.41147666411629685,
+      "grad_norm": 2.4530844333412904,
+      "learning_rate": 1.8424451276572712e-05,
+      "loss": 0.7439,
+      "step": 2689
+    },
+    {
+      "epoch": 0.41162968630451413,
+      "grad_norm": 2.7154613196392163,
+      "learning_rate": 1.8423115817100908e-05,
+      "loss": 0.7944,
+      "step": 2690
+    },
+    {
+      "epoch": 0.41178270849273146,
+      "grad_norm": 2.0993592442625464,
+      "learning_rate": 1.842177984033261e-05,
+      "loss": 0.6573,
+      "step": 2691
+    },
+    {
+      "epoch": 0.41193573068094874,
+      "grad_norm": 2.251977839855377,
+      "learning_rate": 1.8420443346349866e-05,
+      "loss": 0.7723,
+      "step": 2692
+    },
+    {
+      "epoch": 0.412088752869166,
+      "grad_norm": 2.4973317883727355,
+      "learning_rate": 1.8419106335234757e-05,
+      "loss": 0.8563,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4122417750573833,
+      "grad_norm": 2.16198116036908,
+      "learning_rate": 1.8417768807069388e-05,
+      "loss": 0.7834,
+      "step": 2694
+    },
+    {
+      "epoch": 0.41239479724560063,
+      "grad_norm": 2.2356466035192177,
+      "learning_rate": 1.8416430761935912e-05,
+      "loss": 0.8431,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4125478194338179,
+      "grad_norm": 2.3601155991963574,
+      "learning_rate": 1.8415092199916493e-05,
+      "loss": 0.7505,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4127008416220352,
+      "grad_norm": 2.4839307976282297,
+      "learning_rate": 1.8413753121093342e-05,
+      "loss": 0.834,
+      "step": 2697
+    },
+    {
+      "epoch": 0.41285386381025246,
+      "grad_norm": 2.5380141603149733,
+      "learning_rate": 1.84124135255487e-05,
+      "loss": 0.8884,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4130068859984698,
+      "grad_norm": 2.359684411575416,
+      "learning_rate": 1.841107341336483e-05,
+      "loss": 0.8136,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4131599081866871,
+      "grad_norm": 2.246915060713202,
+      "learning_rate": 1.840973278462404e-05,
+      "loss": 0.853,
+      "step": 2700
+    },
+    {
+      "epoch": 0.41331293037490435,
+      "grad_norm": 2.743278992870679,
+      "learning_rate": 1.8408391639408662e-05,
+      "loss": 0.9538,
+      "step": 2701
+    },
+    {
+      "epoch": 0.41346595256312163,
+      "grad_norm": 2.3414868128353445,
+      "learning_rate": 1.8407049977801057e-05,
+      "loss": 0.7642,
+      "step": 2702
+    },
+    {
+      "epoch": 0.41361897475133896,
+      "grad_norm": 2.2150403808055095,
+      "learning_rate": 1.8405707799883625e-05,
+      "loss": 0.8568,
+      "step": 2703
+    },
+    {
+      "epoch": 0.41377199693955624,
+      "grad_norm": 2.1600067523897595,
+      "learning_rate": 1.8404365105738793e-05,
+      "loss": 0.7698,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4139250191277735,
+      "grad_norm": 2.5630873870166027,
+      "learning_rate": 1.8403021895449022e-05,
+      "loss": 0.8071,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4140780413159908,
+      "grad_norm": 2.2015946582709724,
+      "learning_rate": 1.8401678169096803e-05,
+      "loss": 0.7721,
+      "step": 2706
+    },
+    {
+      "epoch": 0.41423106350420813,
+      "grad_norm": 2.3404124500020616,
+      "learning_rate": 1.840033392676466e-05,
+      "loss": 0.7017,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4143840856924254,
+      "grad_norm": 2.5603216912077142,
+      "learning_rate": 1.8398989168535153e-05,
+      "loss": 0.7995,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4145371078806427,
+      "grad_norm": 2.588436369527036,
+      "learning_rate": 1.8397643894490857e-05,
+      "loss": 0.8638,
+      "step": 2709
+    },
+    {
+      "epoch": 0.41469013006885996,
+      "grad_norm": 2.2405037181742418,
+      "learning_rate": 1.8396298104714404e-05,
+      "loss": 0.7631,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4148431522570773,
+      "grad_norm": 2.4837864680181805,
+      "learning_rate": 1.8394951799288433e-05,
+      "loss": 0.8381,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4149961744452946,
+      "grad_norm": 2.434315227598334,
+      "learning_rate": 1.8393604978295634e-05,
+      "loss": 0.7599,
+      "step": 2712
+    },
+    {
+      "epoch": 0.41514919663351185,
+      "grad_norm": 2.3929964810627875,
+      "learning_rate": 1.8392257641818717e-05,
+      "loss": 0.8392,
+      "step": 2713
+    },
+    {
+      "epoch": 0.41530221882172913,
+      "grad_norm": 2.3769240010713784,
+      "learning_rate": 1.8390909789940424e-05,
+      "loss": 0.914,
+      "step": 2714
+    },
+    {
+      "epoch": 0.41545524100994646,
+      "grad_norm": 2.4156353290483996,
+      "learning_rate": 1.8389561422743537e-05,
+      "loss": 0.8067,
+      "step": 2715
+    },
+    {
+      "epoch": 0.41560826319816374,
+      "grad_norm": 2.3762181100367727,
+      "learning_rate": 1.838821254031087e-05,
+      "loss": 0.8145,
+      "step": 2716
+    },
+    {
+      "epoch": 0.415761285386381,
+      "grad_norm": 2.7032932387797994,
+      "learning_rate": 1.8386863142725245e-05,
+      "loss": 0.7724,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4159143075745983,
+      "grad_norm": 2.4199311872817675,
+      "learning_rate": 1.8385513230069552e-05,
+      "loss": 0.796,
+      "step": 2718
+    },
+    {
+      "epoch": 0.41606732976281563,
+      "grad_norm": 2.611915068916634,
+      "learning_rate": 1.8384162802426683e-05,
+      "loss": 0.8,
+      "step": 2719
+    },
+    {
+      "epoch": 0.4162203519510329,
+      "grad_norm": 2.4719883493550663,
+      "learning_rate": 1.838281185987958e-05,
+      "loss": 0.7554,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4163733741392502,
+      "grad_norm": 6.780799641945747,
+      "learning_rate": 1.8381460402511206e-05,
+      "loss": 0.8375,
+      "step": 2721
+    },
+    {
+      "epoch": 0.41652639632746746,
+      "grad_norm": 2.2876882146123765,
+      "learning_rate": 1.8380108430404558e-05,
+      "loss": 0.8064,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4166794185156848,
+      "grad_norm": 2.451624523809711,
+      "learning_rate": 1.8378755943642674e-05,
+      "loss": 0.7935,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4168324407039021,
+      "grad_norm": 2.8942584708078827,
+      "learning_rate": 1.8377402942308606e-05,
+      "loss": 0.7752,
+      "step": 2724
+    },
+    {
+      "epoch": 0.41698546289211935,
+      "grad_norm": 2.4380212312342024,
+      "learning_rate": 1.837604942648545e-05,
+      "loss": 0.8369,
+      "step": 2725
+    },
+    {
+      "epoch": 0.41713848508033663,
+      "grad_norm": 2.42312875758766,
+      "learning_rate": 1.8374695396256333e-05,
+      "loss": 0.8675,
+      "step": 2726
+    },
+    {
+      "epoch": 0.41729150726855396,
+      "grad_norm": 2.822238056417803,
+      "learning_rate": 1.837334085170441e-05,
+      "loss": 0.8107,
+      "step": 2727
+    },
+    {
+      "epoch": 0.41744452945677124,
+      "grad_norm": 2.7255201999810397,
+      "learning_rate": 1.8371985792912867e-05,
+      "loss": 0.8395,
+      "step": 2728
+    },
+    {
+      "epoch": 0.4175975516449885,
+      "grad_norm": 2.7738793587791295,
+      "learning_rate": 1.8370630219964924e-05,
+      "loss": 0.7883,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4177505738332058,
+      "grad_norm": 3.0113058281040987,
+      "learning_rate": 1.8369274132943836e-05,
+      "loss": 0.928,
+      "step": 2730
+    },
+    {
+      "epoch": 0.41790359602142313,
+      "grad_norm": 2.3332835609228626,
+      "learning_rate": 1.8367917531932883e-05,
+      "loss": 0.9316,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4180566182096404,
+      "grad_norm": 2.52833709344253,
+      "learning_rate": 1.8366560417015376e-05,
+      "loss": 0.7773,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4182096403978577,
+      "grad_norm": 2.776855292461947,
+      "learning_rate": 1.8365202788274665e-05,
+      "loss": 0.9908,
+      "step": 2733
+    },
+    {
+      "epoch": 0.41836266258607496,
+      "grad_norm": 3.1038722511338848,
+      "learning_rate": 1.8363844645794127e-05,
+      "loss": 0.7926,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4185156847742923,
+      "grad_norm": 2.6755227443722966,
+      "learning_rate": 1.836248598965717e-05,
+      "loss": 0.7397,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4186687069625096,
+      "grad_norm": 2.193456846703546,
+      "learning_rate": 1.836112681994723e-05,
+      "loss": 0.7661,
+      "step": 2736
+    },
+    {
+      "epoch": 0.41882172915072685,
+      "grad_norm": 2.4328277091982704,
+      "learning_rate": 1.8359767136747787e-05,
+      "loss": 0.8149,
+      "step": 2737
+    },
+    {
+      "epoch": 0.41897475133894413,
+      "grad_norm": 2.5640952513553996,
+      "learning_rate": 1.835840694014234e-05,
+      "loss": 0.8672,
+      "step": 2738
+    },
+    {
+      "epoch": 0.41912777352716146,
+      "grad_norm": 2.372258279722657,
+      "learning_rate": 1.8357046230214424e-05,
+      "loss": 0.7194,
+      "step": 2739
+    },
+    {
+      "epoch": 0.41928079571537874,
+      "grad_norm": 2.694024313287626,
+      "learning_rate": 1.8355685007047602e-05,
+      "loss": 0.8198,
+      "step": 2740
+    },
+    {
+      "epoch": 0.419433817903596,
+      "grad_norm": 2.717663032912964,
+      "learning_rate": 1.8354323270725483e-05,
+      "loss": 0.754,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4195868400918133,
+      "grad_norm": 2.1220023192705466,
+      "learning_rate": 1.8352961021331686e-05,
+      "loss": 0.7195,
+      "step": 2742
+    },
+    {
+      "epoch": 0.41973986228003063,
+      "grad_norm": 2.5635730910519428,
+      "learning_rate": 1.8351598258949878e-05,
+      "loss": 0.6638,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4198928844682479,
+      "grad_norm": 2.307124143928578,
+      "learning_rate": 1.8350234983663745e-05,
+      "loss": 0.779,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4200459066564652,
+      "grad_norm": 2.581024604321476,
+      "learning_rate": 1.834887119555702e-05,
+      "loss": 0.9084,
+      "step": 2745
+    },
+    {
+      "epoch": 0.42019892884468246,
+      "grad_norm": 2.486303856122351,
+      "learning_rate": 1.834750689471345e-05,
+      "loss": 0.8191,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4203519510328998,
+      "grad_norm": 2.794251757152545,
+      "learning_rate": 1.8346142081216828e-05,
+      "loss": 0.7747,
+      "step": 2747
+    },
+    {
+      "epoch": 0.4205049732211171,
+      "grad_norm": 2.5361298102326923,
+      "learning_rate": 1.8344776755150972e-05,
+      "loss": 0.8419,
+      "step": 2748
+    },
+    {
+      "epoch": 0.42065799540933435,
+      "grad_norm": 2.453075570214805,
+      "learning_rate": 1.8343410916599728e-05,
+      "loss": 0.7913,
+      "step": 2749
+    },
+    {
+      "epoch": 0.42081101759755163,
+      "grad_norm": 2.658949348190422,
+      "learning_rate": 1.8342044565646985e-05,
+      "loss": 0.8207,
+      "step": 2750
+    },
+    {
+      "epoch": 0.42096403978576896,
+      "grad_norm": 2.6080813836023204,
+      "learning_rate": 1.8340677702376646e-05,
+      "loss": 0.9163,
+      "step": 2751
+    },
+    {
+      "epoch": 0.42111706197398624,
+      "grad_norm": 2.4580878290478765,
+      "learning_rate": 1.833931032687266e-05,
+      "loss": 0.7212,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4212700841622035,
+      "grad_norm": 2.6866209108407992,
+      "learning_rate": 1.8337942439219002e-05,
+      "loss": 0.9074,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4214231063504208,
+      "grad_norm": 2.5069055372513565,
+      "learning_rate": 1.8336574039499686e-05,
+      "loss": 0.8655,
+      "step": 2754
+    },
+    {
+      "epoch": 0.42157612853863813,
+      "grad_norm": 2.8233674919495506,
+      "learning_rate": 1.8335205127798743e-05,
+      "loss": 0.8258,
+      "step": 2755
+    },
+    {
+      "epoch": 0.4217291507268554,
+      "grad_norm": 2.455363331998748,
+      "learning_rate": 1.8333835704200244e-05,
+      "loss": 0.7243,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4218821729150727,
+      "grad_norm": 2.6976338544948764,
+      "learning_rate": 1.8332465768788294e-05,
+      "loss": 0.8472,
+      "step": 2757
+    },
+    {
+      "epoch": 0.42203519510328996,
+      "grad_norm": 2.2772022318054317,
+      "learning_rate": 1.8331095321647024e-05,
+      "loss": 0.7334,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4221882172915073,
+      "grad_norm": 2.332527092805009,
+      "learning_rate": 1.83297243628606e-05,
+      "loss": 0.8488,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4223412394797246,
+      "grad_norm": 2.65635386373047,
+      "learning_rate": 1.8328352892513213e-05,
+      "loss": 0.8725,
+      "step": 2760
+    },
+    {
+      "epoch": 0.42249426166794185,
+      "grad_norm": 2.595147960241488,
+      "learning_rate": 1.8326980910689097e-05,
+      "loss": 0.8825,
+      "step": 2761
+    },
+    {
+      "epoch": 0.42264728385615913,
+      "grad_norm": 2.4590965620170357,
+      "learning_rate": 1.832560841747251e-05,
+      "loss": 0.7872,
+      "step": 2762
+    },
+    {
+      "epoch": 0.42280030604437646,
+      "grad_norm": 2.105599760748347,
+      "learning_rate": 1.8324235412947734e-05,
+      "loss": 0.7948,
+      "step": 2763
+    },
+    {
+      "epoch": 0.42295332823259374,
+      "grad_norm": 2.1926721287195607,
+      "learning_rate": 1.83228618971991e-05,
+      "loss": 0.6939,
+      "step": 2764
+    },
+    {
+      "epoch": 0.423106350420811,
+      "grad_norm": 2.3227283833432444,
+      "learning_rate": 1.832148787031096e-05,
+      "loss": 0.7924,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4232593726090283,
+      "grad_norm": 2.439579480687344,
+      "learning_rate": 1.8320113332367695e-05,
+      "loss": 0.798,
+      "step": 2766
+    },
+    {
+      "epoch": 0.42341239479724563,
+      "grad_norm": 2.4905389517694707,
+      "learning_rate": 1.831873828345372e-05,
+      "loss": 0.9101,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4235654169854629,
+      "grad_norm": 1.9986218789081267,
+      "learning_rate": 1.8317362723653486e-05,
+      "loss": 0.6391,
+      "step": 2768
+    },
+    {
+      "epoch": 0.4237184391736802,
+      "grad_norm": 2.265857474114456,
+      "learning_rate": 1.8315986653051466e-05,
+      "loss": 0.8241,
+      "step": 2769
+    },
+    {
+      "epoch": 0.42387146136189746,
+      "grad_norm": 2.4080058443982058,
+      "learning_rate": 1.8314610071732178e-05,
+      "loss": 0.7932,
+      "step": 2770
+    },
+    {
+      "epoch": 0.42402448355011474,
+      "grad_norm": 2.3055478158698404,
+      "learning_rate": 1.8313232979780154e-05,
+      "loss": 0.7345,
+      "step": 2771
+    },
+    {
+      "epoch": 0.4241775057383321,
+      "grad_norm": 2.5461071258196064,
+      "learning_rate": 1.8311855377279974e-05,
+      "loss": 0.6186,
+      "step": 2772
+    },
+    {
+      "epoch": 0.42433052792654935,
+      "grad_norm": 2.4759293912836555,
+      "learning_rate": 1.831047726431624e-05,
+      "loss": 0.8608,
+      "step": 2773
+    },
+    {
+      "epoch": 0.42448355011476663,
+      "grad_norm": 2.228142862655671,
+      "learning_rate": 1.8309098640973584e-05,
+      "loss": 0.7126,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4246365723029839,
+      "grad_norm": 2.5373975946691507,
+      "learning_rate": 1.8307719507336675e-05,
+      "loss": 0.7881,
+      "step": 2775
+    },
+    {
+      "epoch": 0.42478959449120124,
+      "grad_norm": 2.33822674451866,
+      "learning_rate": 1.8306339863490216e-05,
+      "loss": 0.8431,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4249426166794185,
+      "grad_norm": 2.7554536395279507,
+      "learning_rate": 1.8304959709518923e-05,
+      "loss": 0.8821,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4250956388676358,
+      "grad_norm": 2.5830841255768244,
+      "learning_rate": 1.830357904550757e-05,
+      "loss": 0.8288,
+      "step": 2778
+    },
+    {
+      "epoch": 0.42524866105585307,
+      "grad_norm": 2.5029825041231204,
+      "learning_rate": 1.8302197871540945e-05,
+      "loss": 0.8809,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4254016832440704,
+      "grad_norm": 2.575665600700266,
+      "learning_rate": 1.830081618770387e-05,
+      "loss": 0.8533,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4255547054322877,
+      "grad_norm": 2.5700770047121124,
+      "learning_rate": 1.8299433994081197e-05,
+      "loss": 0.8074,
+      "step": 2781
+    },
+    {
+      "epoch": 0.42570772762050496,
+      "grad_norm": 2.0623106933950304,
+      "learning_rate": 1.8298051290757818e-05,
+      "loss": 0.7773,
+      "step": 2782
+    },
+    {
+      "epoch": 0.42586074980872224,
+      "grad_norm": 2.18537150872912,
+      "learning_rate": 1.8296668077818644e-05,
+      "loss": 0.8344,
+      "step": 2783
+    },
+    {
+      "epoch": 0.42601377199693957,
+      "grad_norm": 2.3942580809651477,
+      "learning_rate": 1.8295284355348628e-05,
+      "loss": 0.7428,
+      "step": 2784
+    },
+    {
+      "epoch": 0.42616679418515685,
+      "grad_norm": 2.780737215240603,
+      "learning_rate": 1.829390012343275e-05,
+      "loss": 0.8756,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4263198163733741,
+      "grad_norm": 2.2278400641722795,
+      "learning_rate": 1.8292515382156017e-05,
+      "loss": 0.8758,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4264728385615914,
+      "grad_norm": 2.3121894916116164,
+      "learning_rate": 1.8291130131603476e-05,
+      "loss": 0.7272,
+      "step": 2787
+    },
+    {
+      "epoch": 0.42662586074980874,
+      "grad_norm": 2.69045702826044,
+      "learning_rate": 1.8289744371860196e-05,
+      "loss": 0.7238,
+      "step": 2788
+    },
+    {
+      "epoch": 0.426778882938026,
+      "grad_norm": 2.315130684005548,
+      "learning_rate": 1.8288358103011286e-05,
+      "loss": 0.7843,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4269319051262433,
+      "grad_norm": 2.18678882527759,
+      "learning_rate": 1.828697132514188e-05,
+      "loss": 0.8087,
+      "step": 2790
+    },
+    {
+      "epoch": 0.42708492731446057,
+      "grad_norm": 2.381064095904689,
+      "learning_rate": 1.828558403833715e-05,
+      "loss": 0.8279,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4272379495026779,
+      "grad_norm": 2.577619698682901,
+      "learning_rate": 1.8284196242682287e-05,
+      "loss": 0.8842,
+      "step": 2792
+    },
+    {
+      "epoch": 0.4273909716908952,
+      "grad_norm": 2.4459585233481103,
+      "learning_rate": 1.8282807938262525e-05,
+      "loss": 0.7657,
+      "step": 2793
+    },
+    {
+      "epoch": 0.42754399387911246,
+      "grad_norm": 2.4999096948175894,
+      "learning_rate": 1.8281419125163124e-05,
+      "loss": 0.8371,
+      "step": 2794
+    },
+    {
+      "epoch": 0.42769701606732974,
+      "grad_norm": 2.10695554914119,
+      "learning_rate": 1.8280029803469383e-05,
+      "loss": 0.7368,
+      "step": 2795
+    },
+    {
+      "epoch": 0.42785003825554707,
+      "grad_norm": 2.568702428809719,
+      "learning_rate": 1.8278639973266614e-05,
+      "loss": 0.8826,
+      "step": 2796
+    },
+    {
+      "epoch": 0.42800306044376435,
+      "grad_norm": 2.4571418370595373,
+      "learning_rate": 1.827724963464018e-05,
+      "loss": 0.7719,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4281560826319816,
+      "grad_norm": 2.321507335931533,
+      "learning_rate": 1.8275858787675464e-05,
+      "loss": 0.801,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4283091048201989,
+      "grad_norm": 1.9921891114138417,
+      "learning_rate": 1.8274467432457888e-05,
+      "loss": 0.6825,
+      "step": 2799
+    },
+    {
+      "epoch": 0.42846212700841624,
+      "grad_norm": 2.4680977264565325,
+      "learning_rate": 1.8273075569072893e-05,
+      "loss": 0.7475,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4286151491966335,
+      "grad_norm": 2.3974672040643776,
+      "learning_rate": 1.8271683197605966e-05,
+      "loss": 0.79,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4287681713848508,
+      "grad_norm": 2.448891743243999,
+      "learning_rate": 1.8270290318142613e-05,
+      "loss": 0.8843,
+      "step": 2802
+    },
+    {
+      "epoch": 0.42892119357306807,
+      "grad_norm": 2.9211845208427603,
+      "learning_rate": 1.8268896930768378e-05,
+      "loss": 0.9263,
+      "step": 2803
+    },
+    {
+      "epoch": 0.4290742157612854,
+      "grad_norm": 2.2479016422193916,
+      "learning_rate": 1.8267503035568834e-05,
+      "loss": 0.7761,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4292272379495027,
+      "grad_norm": 2.256379799241374,
+      "learning_rate": 1.8266108632629585e-05,
+      "loss": 0.8842,
+      "step": 2805
+    },
+    {
+      "epoch": 0.42938026013771996,
+      "grad_norm": 2.230328900853721,
+      "learning_rate": 1.8264713722036272e-05,
+      "loss": 0.7677,
+      "step": 2806
+    },
+    {
+      "epoch": 0.42953328232593724,
+      "grad_norm": 2.4734719612301808,
+      "learning_rate": 1.8263318303874555e-05,
+      "loss": 0.8867,
+      "step": 2807
+    },
+    {
+      "epoch": 0.42968630451415457,
+      "grad_norm": 2.4024715893301445,
+      "learning_rate": 1.8261922378230134e-05,
+      "loss": 0.7662,
+      "step": 2808
+    },
+    {
+      "epoch": 0.42983932670237185,
+      "grad_norm": 2.3019580177421495,
+      "learning_rate": 1.8260525945188742e-05,
+      "loss": 0.7525,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4299923488905891,
+      "grad_norm": 2.268610624691826,
+      "learning_rate": 1.825912900483613e-05,
+      "loss": 0.8512,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4301453710788064,
+      "grad_norm": 2.50834037125462,
+      "learning_rate": 1.82577315572581e-05,
+      "loss": 0.7428,
+      "step": 2811
+    },
+    {
+      "epoch": 0.43029839326702374,
+      "grad_norm": 1.9801505295055397,
+      "learning_rate": 1.8256333602540472e-05,
+      "loss": 0.6802,
+      "step": 2812
+    },
+    {
+      "epoch": 0.430451415455241,
+      "grad_norm": 2.2031680211345988,
+      "learning_rate": 1.82549351407691e-05,
+      "loss": 0.7427,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4306044376434583,
+      "grad_norm": 2.07270281399312,
+      "learning_rate": 1.8253536172029863e-05,
+      "loss": 0.7541,
+      "step": 2814
+    },
+    {
+      "epoch": 0.43075745983167557,
+      "grad_norm": 2.4736090185078616,
+      "learning_rate": 1.8252136696408683e-05,
+      "loss": 0.8778,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4309104820198929,
+      "grad_norm": 2.1084099756052574,
+      "learning_rate": 1.8250736713991505e-05,
+      "loss": 0.7589,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4310635042081102,
+      "grad_norm": 2.377251552697672,
+      "learning_rate": 1.824933622486431e-05,
+      "loss": 0.8503,
+      "step": 2817
+    },
+    {
+      "epoch": 0.43121652639632746,
+      "grad_norm": 2.5075890326248964,
+      "learning_rate": 1.8247935229113106e-05,
+      "loss": 0.8015,
+      "step": 2818
+    },
+    {
+      "epoch": 0.43136954858454474,
+      "grad_norm": 2.4737690463297124,
+      "learning_rate": 1.824653372682393e-05,
+      "loss": 0.9242,
+      "step": 2819
+    },
+    {
+      "epoch": 0.43152257077276207,
+      "grad_norm": 2.5168772301247424,
+      "learning_rate": 1.824513171808286e-05,
+      "loss": 0.6882,
+      "step": 2820
+    },
+    {
+      "epoch": 0.43167559296097935,
+      "grad_norm": 2.4850955909543324,
+      "learning_rate": 1.8243729202975998e-05,
+      "loss": 0.8031,
+      "step": 2821
+    },
+    {
+      "epoch": 0.4318286151491966,
+      "grad_norm": 2.4310792570377364,
+      "learning_rate": 1.8242326181589472e-05,
+      "loss": 0.7786,
+      "step": 2822
+    },
+    {
+      "epoch": 0.4319816373374139,
+      "grad_norm": 2.2162562659795566,
+      "learning_rate": 1.8240922654009453e-05,
+      "loss": 0.8593,
+      "step": 2823
+    },
+    {
+      "epoch": 0.43213465952563124,
+      "grad_norm": 2.1604512997876615,
+      "learning_rate": 1.823951862032214e-05,
+      "loss": 0.7262,
+      "step": 2824
+    },
+    {
+      "epoch": 0.4322876817138485,
+      "grad_norm": 2.2860096253234334,
+      "learning_rate": 1.8238114080613748e-05,
+      "loss": 0.8108,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4324407039020658,
+      "grad_norm": 2.1896817856146926,
+      "learning_rate": 1.8236709034970547e-05,
+      "loss": 0.7637,
+      "step": 2826
+    },
+    {
+      "epoch": 0.43259372609028307,
+      "grad_norm": 2.5100011119270365,
+      "learning_rate": 1.8235303483478816e-05,
+      "loss": 0.8717,
+      "step": 2827
+    },
+    {
+      "epoch": 0.4327467482785004,
+      "grad_norm": 2.2395244764068494,
+      "learning_rate": 1.8233897426224885e-05,
+      "loss": 0.8137,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4328997704667177,
+      "grad_norm": 2.454140156073576,
+      "learning_rate": 1.8232490863295104e-05,
+      "loss": 0.7323,
+      "step": 2829
+    },
+    {
+      "epoch": 0.43305279265493496,
+      "grad_norm": 2.782749694102962,
+      "learning_rate": 1.8231083794775855e-05,
+      "loss": 0.8337,
+      "step": 2830
+    },
+    {
+      "epoch": 0.43320581484315224,
+      "grad_norm": 2.4708319224270787,
+      "learning_rate": 1.8229676220753544e-05,
+      "loss": 0.7952,
+      "step": 2831
+    },
+    {
+      "epoch": 0.43335883703136957,
+      "grad_norm": 2.450763340830767,
+      "learning_rate": 1.8228268141314625e-05,
+      "loss": 0.7822,
+      "step": 2832
+    },
+    {
+      "epoch": 0.43351185921958685,
+      "grad_norm": 2.6846953429174265,
+      "learning_rate": 1.8226859556545567e-05,
+      "loss": 0.9821,
+      "step": 2833
+    },
+    {
+      "epoch": 0.4336648814078041,
+      "grad_norm": 2.384263297031218,
+      "learning_rate": 1.8225450466532886e-05,
+      "loss": 0.7387,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4338179035960214,
+      "grad_norm": 2.5933253496031314,
+      "learning_rate": 1.8224040871363108e-05,
+      "loss": 0.8023,
+      "step": 2835
+    },
+    {
+      "epoch": 0.43397092578423874,
+      "grad_norm": 2.1470594209931164,
+      "learning_rate": 1.8222630771122813e-05,
+      "loss": 0.7342,
+      "step": 2836
+    },
+    {
+      "epoch": 0.434123947972456,
+      "grad_norm": 2.177838898350675,
+      "learning_rate": 1.822122016589859e-05,
+      "loss": 0.7216,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4342769701606733,
+      "grad_norm": 2.318251382225496,
+      "learning_rate": 1.8219809055777078e-05,
+      "loss": 0.8282,
+      "step": 2838
+    },
+    {
+      "epoch": 0.43442999234889057,
+      "grad_norm": 2.485311148535998,
+      "learning_rate": 1.8218397440844935e-05,
+      "loss": 0.9077,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4345830145371079,
+      "grad_norm": 2.187991898973603,
+      "learning_rate": 1.8216985321188857e-05,
+      "loss": 0.6987,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4347360367253252,
+      "grad_norm": 2.954983399291798,
+      "learning_rate": 1.821557269689556e-05,
+      "loss": 0.8829,
+      "step": 2841
+    },
+    {
+      "epoch": 0.43488905891354246,
+      "grad_norm": 2.3435852188557136,
+      "learning_rate": 1.8214159568051808e-05,
+      "loss": 0.7581,
+      "step": 2842
+    },
+    {
+      "epoch": 0.43504208110175974,
+      "grad_norm": 2.506988531798837,
+      "learning_rate": 1.821274593474438e-05,
+      "loss": 0.9141,
+      "step": 2843
+    },
+    {
+      "epoch": 0.43519510328997707,
+      "grad_norm": 2.376637972611891,
+      "learning_rate": 1.82113317970601e-05,
+      "loss": 0.7158,
+      "step": 2844
+    },
+    {
+      "epoch": 0.43534812547819435,
+      "grad_norm": 2.426861380496052,
+      "learning_rate": 1.8209917155085807e-05,
+      "loss": 0.8952,
+      "step": 2845
+    },
+    {
+      "epoch": 0.4355011476664116,
+      "grad_norm": 2.217284799803432,
+      "learning_rate": 1.820850200890839e-05,
+      "loss": 0.7568,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4356541698546289,
+      "grad_norm": 3.406641979935441,
+      "learning_rate": 1.8207086358614748e-05,
+      "loss": 0.7537,
+      "step": 2847
+    },
+    {
+      "epoch": 0.43580719204284624,
+      "grad_norm": 2.2708735332971397,
+      "learning_rate": 1.8205670204291827e-05,
+      "loss": 0.7447,
+      "step": 2848
+    },
+    {
+      "epoch": 0.4359602142310635,
+      "grad_norm": 2.4379444362854796,
+      "learning_rate": 1.82042535460266e-05,
+      "loss": 0.8063,
+      "step": 2849
+    },
+    {
+      "epoch": 0.4361132364192808,
+      "grad_norm": 2.644713852715883,
+      "learning_rate": 1.8202836383906065e-05,
+      "loss": 0.8981,
+      "step": 2850
+    },
+    {
+      "epoch": 0.43626625860749807,
+      "grad_norm": 2.3110592853637324,
+      "learning_rate": 1.8201418718017257e-05,
+      "loss": 0.828,
+      "step": 2851
+    },
+    {
+      "epoch": 0.4364192807957154,
+      "grad_norm": 2.19094109374234,
+      "learning_rate": 1.8200000548447242e-05,
+      "loss": 0.7952,
+      "step": 2852
+    },
+    {
+      "epoch": 0.4365723029839327,
+      "grad_norm": 2.4567016099138814,
+      "learning_rate": 1.8198581875283116e-05,
+      "loss": 0.8291,
+      "step": 2853
+    },
+    {
+      "epoch": 0.43672532517214996,
+      "grad_norm": 2.5230151299441665,
+      "learning_rate": 1.8197162698612008e-05,
+      "loss": 0.8983,
+      "step": 2854
+    },
+    {
+      "epoch": 0.43687834736036724,
+      "grad_norm": 2.2014429530314477,
+      "learning_rate": 1.8195743018521067e-05,
+      "loss": 0.8416,
+      "step": 2855
+    },
+    {
+      "epoch": 0.43703136954858457,
+      "grad_norm": 2.798858313321146,
+      "learning_rate": 1.8194322835097484e-05,
+      "loss": 0.6999,
+      "step": 2856
+    },
+    {
+      "epoch": 0.43718439173680185,
+      "grad_norm": 2.559519260396577,
+      "learning_rate": 1.8192902148428482e-05,
+      "loss": 1.0143,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4373374139250191,
+      "grad_norm": 2.3153587914032254,
+      "learning_rate": 1.8191480958601307e-05,
+      "loss": 0.8853,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4374904361132364,
+      "grad_norm": 2.6106845850972498,
+      "learning_rate": 1.819005926570324e-05,
+      "loss": 0.8967,
+      "step": 2859
+    },
+    {
+      "epoch": 0.43764345830145374,
+      "grad_norm": 2.6529185075899244,
+      "learning_rate": 1.8188637069821595e-05,
+      "loss": 0.7725,
+      "step": 2860
+    },
+    {
+      "epoch": 0.437796480489671,
+      "grad_norm": 2.4152913946558248,
+      "learning_rate": 1.8187214371043717e-05,
+      "loss": 0.7995,
+      "step": 2861
+    },
+    {
+      "epoch": 0.4379495026778883,
+      "grad_norm": 2.7423652149993165,
+      "learning_rate": 1.818579116945697e-05,
+      "loss": 0.9248,
+      "step": 2862
+    },
+    {
+      "epoch": 0.43810252486610557,
+      "grad_norm": 2.1886173607384904,
+      "learning_rate": 1.8184367465148768e-05,
+      "loss": 0.769,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4382555470543229,
+      "grad_norm": 2.2565851532333654,
+      "learning_rate": 1.818294325820654e-05,
+      "loss": 0.8638,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4384085692425402,
+      "grad_norm": 2.476506526281387,
+      "learning_rate": 1.8181518548717757e-05,
+      "loss": 0.8676,
+      "step": 2865
+    },
+    {
+      "epoch": 0.43856159143075746,
+      "grad_norm": 2.4366507731535068,
+      "learning_rate": 1.818009333676991e-05,
+      "loss": 0.785,
+      "step": 2866
+    },
+    {
+      "epoch": 0.43871461361897474,
+      "grad_norm": 2.4976448948903784,
+      "learning_rate": 1.8178667622450536e-05,
+      "loss": 0.8044,
+      "step": 2867
+    },
+    {
+      "epoch": 0.43886763580719207,
+      "grad_norm": 2.2797810019467306,
+      "learning_rate": 1.8177241405847185e-05,
+      "loss": 0.7425,
+      "step": 2868
+    },
+    {
+      "epoch": 0.43902065799540935,
+      "grad_norm": 2.189360272061294,
+      "learning_rate": 1.817581468704745e-05,
+      "loss": 0.8035,
+      "step": 2869
+    },
+    {
+      "epoch": 0.4391736801836266,
+      "grad_norm": 2.189311640496233,
+      "learning_rate": 1.817438746613895e-05,
+      "loss": 0.7365,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4393267023718439,
+      "grad_norm": 2.35659633222008,
+      "learning_rate": 1.817295974320934e-05,
+      "loss": 0.8272,
+      "step": 2871
+    },
+    {
+      "epoch": 0.43947972456006124,
+      "grad_norm": 2.093541114804993,
+      "learning_rate": 1.8171531518346296e-05,
+      "loss": 0.7011,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4396327467482785,
+      "grad_norm": 2.3219154246904203,
+      "learning_rate": 1.8170102791637538e-05,
+      "loss": 0.6946,
+      "step": 2873
+    },
+    {
+      "epoch": 0.4397857689364958,
+      "grad_norm": 2.578669440132516,
+      "learning_rate": 1.81686735631708e-05,
+      "loss": 0.8617,
+      "step": 2874
+    },
+    {
+      "epoch": 0.43993879112471307,
+      "grad_norm": 2.194341864294571,
+      "learning_rate": 1.8167243833033866e-05,
+      "loss": 0.7764,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4400918133129304,
+      "grad_norm": 2.543107156092695,
+      "learning_rate": 1.8165813601314538e-05,
+      "loss": 0.8232,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4402448355011477,
+      "grad_norm": 2.1321281234592964,
+      "learning_rate": 1.8164382868100655e-05,
+      "loss": 0.8494,
+      "step": 2877
+    },
+    {
+      "epoch": 0.44039785768936496,
+      "grad_norm": 2.1116804083957996,
+      "learning_rate": 1.8162951633480076e-05,
+      "loss": 0.689,
+      "step": 2878
+    },
+    {
+      "epoch": 0.44055087987758224,
+      "grad_norm": 2.6610632363025735,
+      "learning_rate": 1.8161519897540706e-05,
+      "loss": 0.9145,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4407039020657995,
+      "grad_norm": 2.4608039381193283,
+      "learning_rate": 1.816008766037047e-05,
+      "loss": 0.8626,
+      "step": 2880
+    },
+    {
+      "epoch": 0.44085692425401685,
+      "grad_norm": 2.0608838256285527,
+      "learning_rate": 1.815865492205733e-05,
+      "loss": 0.7088,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4410099464422341,
+      "grad_norm": 2.266950103228107,
+      "learning_rate": 1.8157221682689274e-05,
+      "loss": 0.7986,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4411629686304514,
+      "grad_norm": 2.320138111880372,
+      "learning_rate": 1.8155787942354323e-05,
+      "loss": 0.7952,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4413159908186687,
+      "grad_norm": 2.030179057500993,
+      "learning_rate": 1.815435370114053e-05,
+      "loss": 0.7375,
+      "step": 2884
+    },
+    {
+      "epoch": 0.441469013006886,
+      "grad_norm": 2.277446909757116,
+      "learning_rate": 1.8152918959135972e-05,
+      "loss": 0.7952,
+      "step": 2885
+    },
+    {
+      "epoch": 0.4416220351951033,
+      "grad_norm": 2.336938364081513,
+      "learning_rate": 1.815148371642877e-05,
+      "loss": 0.7348,
+      "step": 2886
+    },
+    {
+      "epoch": 0.44177505738332057,
+      "grad_norm": 2.068616185340274,
+      "learning_rate": 1.8150047973107062e-05,
+      "loss": 0.7344,
+      "step": 2887
+    },
+    {
+      "epoch": 0.44192807957153785,
+      "grad_norm": 2.378424795705215,
+      "learning_rate": 1.8148611729259028e-05,
+      "loss": 0.8116,
+      "step": 2888
+    },
+    {
+      "epoch": 0.4420811017597552,
+      "grad_norm": 2.315366839645358,
+      "learning_rate": 1.8147174984972866e-05,
+      "loss": 0.8843,
+      "step": 2889
+    },
+    {
+      "epoch": 0.44223412394797246,
+      "grad_norm": 2.282370051057871,
+      "learning_rate": 1.814573774033682e-05,
+      "loss": 0.6782,
+      "step": 2890
+    },
+    {
+      "epoch": 0.44238714613618974,
+      "grad_norm": 2.3368147120736054,
+      "learning_rate": 1.814429999543915e-05,
+      "loss": 0.7882,
+      "step": 2891
+    },
+    {
+      "epoch": 0.442540168324407,
+      "grad_norm": 2.1649407313959386,
+      "learning_rate": 1.8142861750368157e-05,
+      "loss": 0.697,
+      "step": 2892
+    },
+    {
+      "epoch": 0.44269319051262435,
+      "grad_norm": 2.5351682813389043,
+      "learning_rate": 1.8141423005212168e-05,
+      "loss": 0.8544,
+      "step": 2893
+    },
+    {
+      "epoch": 0.4428462127008416,
+      "grad_norm": 2.4679098802704553,
+      "learning_rate": 1.8139983760059546e-05,
+      "loss": 0.8602,
+      "step": 2894
+    },
+    {
+      "epoch": 0.4429992348890589,
+      "grad_norm": 2.4500507125200657,
+      "learning_rate": 1.8138544014998676e-05,
+      "loss": 0.8638,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4431522570772762,
+      "grad_norm": 2.907302704924948,
+      "learning_rate": 1.813710377011798e-05,
+      "loss": 0.8541,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4433052792654935,
+      "grad_norm": 2.668616754615131,
+      "learning_rate": 1.813566302550591e-05,
+      "loss": 0.8274,
+      "step": 2897
+    },
+    {
+      "epoch": 0.4434583014537108,
+      "grad_norm": 2.3327329427740255,
+      "learning_rate": 1.8134221781250946e-05,
+      "loss": 0.785,
+      "step": 2898
+    },
+    {
+      "epoch": 0.44361132364192807,
+      "grad_norm": 2.276328368333612,
+      "learning_rate": 1.81327800374416e-05,
+      "loss": 0.7404,
+      "step": 2899
+    },
+    {
+      "epoch": 0.44376434583014535,
+      "grad_norm": 2.196279587222736,
+      "learning_rate": 1.8131337794166417e-05,
+      "loss": 0.7991,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4439173680183627,
+      "grad_norm": 2.4529440496173684,
+      "learning_rate": 1.812989505151397e-05,
+      "loss": 0.8399,
+      "step": 2901
+    },
+    {
+      "epoch": 0.44407039020657996,
+      "grad_norm": 2.2537061718280325,
+      "learning_rate": 1.8128451809572865e-05,
+      "loss": 0.7269,
+      "step": 2902
+    },
+    {
+      "epoch": 0.44422341239479723,
+      "grad_norm": 2.322994259587146,
+      "learning_rate": 1.8127008068431734e-05,
+      "loss": 0.7474,
+      "step": 2903
+    },
+    {
+      "epoch": 0.4443764345830145,
+      "grad_norm": 2.3975452426239405,
+      "learning_rate": 1.8125563828179244e-05,
+      "loss": 0.8157,
+      "step": 2904
+    },
+    {
+      "epoch": 0.44452945677123185,
+      "grad_norm": 2.0799712070502197,
+      "learning_rate": 1.8124119088904094e-05,
+      "loss": 0.7465,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4446824789594491,
+      "grad_norm": 2.445692299342647,
+      "learning_rate": 1.8122673850695007e-05,
+      "loss": 0.833,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4448355011476664,
+      "grad_norm": 2.347753895749705,
+      "learning_rate": 1.8121228113640744e-05,
+      "loss": 0.6599,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4449885233358837,
+      "grad_norm": 2.294931489132973,
+      "learning_rate": 1.8119781877830093e-05,
+      "loss": 0.901,
+      "step": 2908
+    },
+    {
+      "epoch": 0.445141545524101,
+      "grad_norm": 2.2286810517605655,
+      "learning_rate": 1.8118335143351868e-05,
+      "loss": 0.7316,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4452945677123183,
+      "grad_norm": 2.4912510299578057,
+      "learning_rate": 1.8116887910294925e-05,
+      "loss": 0.8443,
+      "step": 2910
+    },
+    {
+      "epoch": 0.44544758990053557,
+      "grad_norm": 2.459604830145821,
+      "learning_rate": 1.8115440178748142e-05,
+      "loss": 0.7226,
+      "step": 2911
+    },
+    {
+      "epoch": 0.44560061208875285,
+      "grad_norm": 2.6383148521577873,
+      "learning_rate": 1.811399194880043e-05,
+      "loss": 0.8549,
+      "step": 2912
+    },
+    {
+      "epoch": 0.4457536342769702,
+      "grad_norm": 2.280528485015101,
+      "learning_rate": 1.8112543220540733e-05,
+      "loss": 0.8502,
+      "step": 2913
+    },
+    {
+      "epoch": 0.44590665646518746,
+      "grad_norm": 2.4402696603587426,
+      "learning_rate": 1.8111093994058017e-05,
+      "loss": 0.8576,
+      "step": 2914
+    },
+    {
+      "epoch": 0.44605967865340473,
+      "grad_norm": 2.2925670057229977,
+      "learning_rate": 1.810964426944129e-05,
+      "loss": 0.7551,
+      "step": 2915
+    },
+    {
+      "epoch": 0.446212700841622,
+      "grad_norm": 2.2902795548425723,
+      "learning_rate": 1.8108194046779584e-05,
+      "loss": 0.7498,
+      "step": 2916
+    },
+    {
+      "epoch": 0.44636572302983935,
+      "grad_norm": 2.510625241344344,
+      "learning_rate": 1.8106743326161957e-05,
+      "loss": 0.826,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4465187452180566,
+      "grad_norm": 2.4701981018218793,
+      "learning_rate": 1.810529210767751e-05,
+      "loss": 0.8336,
+      "step": 2918
+    },
+    {
+      "epoch": 0.4466717674062739,
+      "grad_norm": 2.5259770545154865,
+      "learning_rate": 1.8103840391415372e-05,
+      "loss": 0.7502,
+      "step": 2919
+    },
+    {
+      "epoch": 0.4468247895944912,
+      "grad_norm": 2.5249562507172287,
+      "learning_rate": 1.810238817746469e-05,
+      "loss": 0.7999,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4469778117827085,
+      "grad_norm": 2.4396115738869444,
+      "learning_rate": 1.8100935465914654e-05,
+      "loss": 0.8055,
+      "step": 2921
+    },
+    {
+      "epoch": 0.4471308339709258,
+      "grad_norm": 2.3065700824046282,
+      "learning_rate": 1.8099482256854477e-05,
+      "loss": 0.8083,
+      "step": 2922
+    },
+    {
+      "epoch": 0.44728385615914307,
+      "grad_norm": 2.349949003979491,
+      "learning_rate": 1.8098028550373415e-05,
+      "loss": 0.8084,
+      "step": 2923
+    },
+    {
+      "epoch": 0.44743687834736035,
+      "grad_norm": 2.389430202044748,
+      "learning_rate": 1.8096574346560736e-05,
+      "loss": 0.782,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4475899005355777,
+      "grad_norm": 2.210746221263167,
+      "learning_rate": 1.809511964550575e-05,
+      "loss": 0.6941,
+      "step": 2925
+    },
+    {
+      "epoch": 0.44774292272379496,
+      "grad_norm": 2.1626760051564897,
+      "learning_rate": 1.8093664447297802e-05,
+      "loss": 0.796,
+      "step": 2926
+    },
+    {
+      "epoch": 0.44789594491201223,
+      "grad_norm": 2.251030501988546,
+      "learning_rate": 1.809220875202626e-05,
+      "loss": 0.8277,
+      "step": 2927
+    },
+    {
+      "epoch": 0.4480489671002295,
+      "grad_norm": 2.1946018916323435,
+      "learning_rate": 1.8090752559780515e-05,
+      "loss": 0.7804,
+      "step": 2928
+    },
+    {
+      "epoch": 0.44820198928844684,
+      "grad_norm": 2.139406362787451,
+      "learning_rate": 1.8089295870650007e-05,
+      "loss": 0.7893,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4483550114766641,
+      "grad_norm": 2.648584053744639,
+      "learning_rate": 1.8087838684724196e-05,
+      "loss": 0.7844,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4485080336648814,
+      "grad_norm": 2.8523250188221625,
+      "learning_rate": 1.8086381002092573e-05,
+      "loss": 0.8421,
+      "step": 2931
+    },
+    {
+      "epoch": 0.4486610558530987,
+      "grad_norm": 2.316275777085334,
+      "learning_rate": 1.8084922822844655e-05,
+      "loss": 0.8628,
+      "step": 2932
+    },
+    {
+      "epoch": 0.448814078041316,
+      "grad_norm": 2.7344218627285857,
+      "learning_rate": 1.8083464147069997e-05,
+      "loss": 0.7968,
+      "step": 2933
+    },
+    {
+      "epoch": 0.4489671002295333,
+      "grad_norm": 2.5515179425547747,
+      "learning_rate": 1.8082004974858186e-05,
+      "loss": 0.6988,
+      "step": 2934
+    },
+    {
+      "epoch": 0.44912012241775057,
+      "grad_norm": 2.274009925043369,
+      "learning_rate": 1.808054530629883e-05,
+      "loss": 0.8237,
+      "step": 2935
+    },
+    {
+      "epoch": 0.44927314460596784,
+      "grad_norm": 2.2219946969109006,
+      "learning_rate": 1.8079085141481577e-05,
+      "loss": 0.8523,
+      "step": 2936
+    },
+    {
+      "epoch": 0.4494261667941852,
+      "grad_norm": 2.2150504476863295,
+      "learning_rate": 1.8077624480496102e-05,
+      "loss": 0.8053,
+      "step": 2937
+    },
+    {
+      "epoch": 0.44957918898240246,
+      "grad_norm": 2.419606454998503,
+      "learning_rate": 1.8076163323432104e-05,
+      "loss": 0.8921,
+      "step": 2938
+    },
+    {
+      "epoch": 0.44973221117061973,
+      "grad_norm": 2.36212827728317,
+      "learning_rate": 1.8074701670379324e-05,
+      "loss": 0.7622,
+      "step": 2939
+    },
+    {
+      "epoch": 0.449885233358837,
+      "grad_norm": 2.3272707660273726,
+      "learning_rate": 1.8073239521427524e-05,
+      "loss": 0.8535,
+      "step": 2940
+    },
+    {
+      "epoch": 0.45003825554705434,
+      "grad_norm": 2.516712688408001,
+      "learning_rate": 1.8071776876666506e-05,
+      "loss": 0.8504,
+      "step": 2941
+    },
+    {
+      "epoch": 0.4501912777352716,
+      "grad_norm": 2.223349887585654,
+      "learning_rate": 1.807031373618609e-05,
+      "loss": 0.8479,
+      "step": 2942
+    },
+    {
+      "epoch": 0.4503442999234889,
+      "grad_norm": 2.251986025412135,
+      "learning_rate": 1.806885010007614e-05,
+      "loss": 0.7571,
+      "step": 2943
+    },
+    {
+      "epoch": 0.4504973221117062,
+      "grad_norm": 2.910578161536275,
+      "learning_rate": 1.8067385968426533e-05,
+      "loss": 0.8832,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4506503442999235,
+      "grad_norm": 2.5658246641317435,
+      "learning_rate": 1.8065921341327203e-05,
+      "loss": 0.8364,
+      "step": 2945
+    },
+    {
+      "epoch": 0.4508033664881408,
+      "grad_norm": 2.5441684589337923,
+      "learning_rate": 1.806445621886808e-05,
+      "loss": 0.8535,
+      "step": 2946
+    },
+    {
+      "epoch": 0.45095638867635807,
+      "grad_norm": 2.346003869170097,
+      "learning_rate": 1.806299060113916e-05,
+      "loss": 0.7937,
+      "step": 2947
+    },
+    {
+      "epoch": 0.45110941086457534,
+      "grad_norm": 2.3686802916882264,
+      "learning_rate": 1.8061524488230443e-05,
+      "loss": 0.9208,
+      "step": 2948
+    },
+    {
+      "epoch": 0.4512624330527927,
+      "grad_norm": 2.315318897085286,
+      "learning_rate": 1.8060057880231966e-05,
+      "loss": 0.7576,
+      "step": 2949
+    },
+    {
+      "epoch": 0.45141545524100996,
+      "grad_norm": 2.304204698051078,
+      "learning_rate": 1.8058590777233807e-05,
+      "loss": 0.8123,
+      "step": 2950
+    },
+    {
+      "epoch": 0.45156847742922723,
+      "grad_norm": 2.4172138592186547,
+      "learning_rate": 1.805712317932606e-05,
+      "loss": 0.7145,
+      "step": 2951
+    },
+    {
+      "epoch": 0.4517214996174445,
+      "grad_norm": 2.1016129747785213,
+      "learning_rate": 1.8055655086598864e-05,
+      "loss": 0.7545,
+      "step": 2952
+    },
+    {
+      "epoch": 0.45187452180566184,
+      "grad_norm": 2.49976385009864,
+      "learning_rate": 1.8054186499142373e-05,
+      "loss": 0.8997,
+      "step": 2953
+    },
+    {
+      "epoch": 0.4520275439938791,
+      "grad_norm": 2.1015201909120953,
+      "learning_rate": 1.8052717417046776e-05,
+      "loss": 0.7667,
+      "step": 2954
+    },
+    {
+      "epoch": 0.4521805661820964,
+      "grad_norm": 2.191104113892472,
+      "learning_rate": 1.8051247840402305e-05,
+      "loss": 0.7319,
+      "step": 2955
+    },
+    {
+      "epoch": 0.4523335883703137,
+      "grad_norm": 2.4834077756294,
+      "learning_rate": 1.8049777769299207e-05,
+      "loss": 0.8395,
+      "step": 2956
+    },
+    {
+      "epoch": 0.452486610558531,
+      "grad_norm": 2.6233714667161534,
+      "learning_rate": 1.8048307203827766e-05,
+      "loss": 0.7377,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4526396327467483,
+      "grad_norm": 2.3852562290363855,
+      "learning_rate": 1.804683614407829e-05,
+      "loss": 0.8138,
+      "step": 2958
+    },
+    {
+      "epoch": 0.45279265493496557,
+      "grad_norm": 2.405802744082427,
+      "learning_rate": 1.8045364590141132e-05,
+      "loss": 0.7731,
+      "step": 2959
+    },
+    {
+      "epoch": 0.45294567712318284,
+      "grad_norm": 2.281624831951309,
+      "learning_rate": 1.804389254210666e-05,
+      "loss": 0.7172,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4530986993114002,
+      "grad_norm": 2.5650189896049183,
+      "learning_rate": 1.8042420000065276e-05,
+      "loss": 0.8563,
+      "step": 2961
+    },
+    {
+      "epoch": 0.45325172149961745,
+      "grad_norm": 2.186990313937297,
+      "learning_rate": 1.8040946964107423e-05,
+      "loss": 0.7669,
+      "step": 2962
+    },
+    {
+      "epoch": 0.45340474368783473,
+      "grad_norm": 2.5993364039741147,
+      "learning_rate": 1.8039473434323555e-05,
+      "loss": 0.8328,
+      "step": 2963
+    },
+    {
+      "epoch": 0.453557765876052,
+      "grad_norm": 2.690424314638924,
+      "learning_rate": 1.803799941080418e-05,
+      "loss": 0.7906,
+      "step": 2964
+    },
+    {
+      "epoch": 0.45371078806426934,
+      "grad_norm": 2.3615307843715803,
+      "learning_rate": 1.803652489363981e-05,
+      "loss": 0.8846,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4538638102524866,
+      "grad_norm": 2.3028327800647923,
+      "learning_rate": 1.8035049882921008e-05,
+      "loss": 0.7632,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4540168324407039,
+      "grad_norm": 2.6992236760908104,
+      "learning_rate": 1.803357437873836e-05,
+      "loss": 0.8173,
+      "step": 2967
+    },
+    {
+      "epoch": 0.4541698546289212,
+      "grad_norm": 2.1987188003454072,
+      "learning_rate": 1.8032098381182483e-05,
+      "loss": 0.8117,
+      "step": 2968
+    },
+    {
+      "epoch": 0.4543228768171385,
+      "grad_norm": 2.6263119915845605,
+      "learning_rate": 1.8030621890344023e-05,
+      "loss": 0.8381,
+      "step": 2969
+    },
+    {
+      "epoch": 0.4544758990053558,
+      "grad_norm": 2.5134499182053296,
+      "learning_rate": 1.8029144906313653e-05,
+      "loss": 0.8841,
+      "step": 2970
+    },
+    {
+      "epoch": 0.45462892119357307,
+      "grad_norm": 2.3417192495256867,
+      "learning_rate": 1.8027667429182087e-05,
+      "loss": 0.765,
+      "step": 2971
+    },
+    {
+      "epoch": 0.45478194338179034,
+      "grad_norm": 2.3521797349686455,
+      "learning_rate": 1.8026189459040063e-05,
+      "loss": 0.7582,
+      "step": 2972
+    },
+    {
+      "epoch": 0.4549349655700077,
+      "grad_norm": 2.2634494087856964,
+      "learning_rate": 1.802471099597834e-05,
+      "loss": 0.7566,
+      "step": 2973
+    },
+    {
+      "epoch": 0.45508798775822495,
+      "grad_norm": 2.757423631425634,
+      "learning_rate": 1.8023232040087723e-05,
+      "loss": 0.7558,
+      "step": 2974
+    },
+    {
+      "epoch": 0.45524100994644223,
+      "grad_norm": 2.4974299117094536,
+      "learning_rate": 1.802175259145904e-05,
+      "loss": 0.904,
+      "step": 2975
+    },
+    {
+      "epoch": 0.4553940321346595,
+      "grad_norm": 2.4164896648530365,
+      "learning_rate": 1.8020272650183152e-05,
+      "loss": 0.7928,
+      "step": 2976
+    },
+    {
+      "epoch": 0.45554705432287684,
+      "grad_norm": 2.278461747453805,
+      "learning_rate": 1.801879221635094e-05,
+      "loss": 0.7486,
+      "step": 2977
+    },
+    {
+      "epoch": 0.4557000765110941,
+      "grad_norm": 2.325475613320422,
+      "learning_rate": 1.8017311290053335e-05,
+      "loss": 0.7439,
+      "step": 2978
+    },
+    {
+      "epoch": 0.4558530986993114,
+      "grad_norm": 2.5560265554303316,
+      "learning_rate": 1.8015829871381272e-05,
+      "loss": 0.8628,
+      "step": 2979
+    },
+    {
+      "epoch": 0.4560061208875287,
+      "grad_norm": 2.7690977507099417,
+      "learning_rate": 1.8014347960425742e-05,
+      "loss": 0.8506,
+      "step": 2980
+    },
+    {
+      "epoch": 0.456159143075746,
+      "grad_norm": 2.4905184954891544,
+      "learning_rate": 1.801286555727775e-05,
+      "loss": 0.7848,
+      "step": 2981
+    },
+    {
+      "epoch": 0.4563121652639633,
+      "grad_norm": 2.7241255638703317,
+      "learning_rate": 1.8011382662028343e-05,
+      "loss": 0.7943,
+      "step": 2982
+    },
+    {
+      "epoch": 0.45646518745218057,
+      "grad_norm": 2.3442771277370404,
+      "learning_rate": 1.800989927476858e-05,
+      "loss": 0.8593,
+      "step": 2983
+    },
+    {
+      "epoch": 0.45661820964039784,
+      "grad_norm": 2.3831996561185234,
+      "learning_rate": 1.800841539558957e-05,
+      "loss": 0.7683,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4567712318286152,
+      "grad_norm": 2.529846692344636,
+      "learning_rate": 1.800693102458244e-05,
+      "loss": 0.7843,
+      "step": 2985
+    },
+    {
+      "epoch": 0.45692425401683245,
+      "grad_norm": 2.5994620405027686,
+      "learning_rate": 1.8005446161838355e-05,
+      "loss": 0.7773,
+      "step": 2986
+    },
+    {
+      "epoch": 0.45707727620504973,
+      "grad_norm": 2.1902994542558947,
+      "learning_rate": 1.8003960807448505e-05,
+      "loss": 0.8147,
+      "step": 2987
+    },
+    {
+      "epoch": 0.457230298393267,
+      "grad_norm": 2.666362830144391,
+      "learning_rate": 1.800247496150411e-05,
+      "loss": 0.8922,
+      "step": 2988
+    },
+    {
+      "epoch": 0.45738332058148434,
+      "grad_norm": 2.348016100092245,
+      "learning_rate": 1.800098862409642e-05,
+      "loss": 0.8225,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4575363427697016,
+      "grad_norm": 2.6178249141604364,
+      "learning_rate": 1.7999501795316712e-05,
+      "loss": 0.7883,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4576893649579189,
+      "grad_norm": 2.6567570604699315,
+      "learning_rate": 1.7998014475256313e-05,
+      "loss": 0.8205,
+      "step": 2991
+    },
+    {
+      "epoch": 0.4578423871461362,
+      "grad_norm": 2.1047805071412014,
+      "learning_rate": 1.7996526664006553e-05,
+      "loss": 0.6739,
+      "step": 2992
+    },
+    {
+      "epoch": 0.45799540933435345,
+      "grad_norm": 2.404156937766429,
+      "learning_rate": 1.7995038361658813e-05,
+      "loss": 0.687,
+      "step": 2993
+    },
+    {
+      "epoch": 0.4581484315225708,
+      "grad_norm": 2.3096394834735277,
+      "learning_rate": 1.7993549568304485e-05,
+      "loss": 0.8982,
+      "step": 2994
+    },
+    {
+      "epoch": 0.45830145371078806,
+      "grad_norm": 2.3445700625391535,
+      "learning_rate": 1.799206028403501e-05,
+      "loss": 0.8137,
+      "step": 2995
+    },
+    {
+      "epoch": 0.45845447589900534,
+      "grad_norm": 2.370222594067083,
+      "learning_rate": 1.7990570508941845e-05,
+      "loss": 0.6931,
+      "step": 2996
+    },
+    {
+      "epoch": 0.4586074980872226,
+      "grad_norm": 2.1936059754346706,
+      "learning_rate": 1.7989080243116486e-05,
+      "loss": 0.6917,
+      "step": 2997
+    },
+    {
+      "epoch": 0.45876052027543995,
+      "grad_norm": 2.139823927624093,
+      "learning_rate": 1.7987589486650457e-05,
+      "loss": 0.7353,
+      "step": 2998
+    },
+    {
+      "epoch": 0.45891354246365723,
+      "grad_norm": 2.418548770240328,
+      "learning_rate": 1.7986098239635313e-05,
+      "loss": 0.7797,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4590665646518745,
+      "grad_norm": 2.427079798960257,
+      "learning_rate": 1.798460650216263e-05,
+      "loss": 0.8436,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4592195868400918,
+      "grad_norm": 3.182514965849544,
+      "learning_rate": 1.7983114274324026e-05,
+      "loss": 0.9386,
+      "step": 3001
+    },
+    {
+      "epoch": 0.4593726090283091,
+      "grad_norm": 2.396757540197267,
+      "learning_rate": 1.7981621556211146e-05,
+      "loss": 0.7159,
+      "step": 3002
+    },
+    {
+      "epoch": 0.4595256312165264,
+      "grad_norm": 2.5679978233582004,
+      "learning_rate": 1.7980128347915663e-05,
+      "loss": 0.8365,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4596786534047437,
+      "grad_norm": 2.38552520369705,
+      "learning_rate": 1.7978634649529276e-05,
+      "loss": 0.9041,
+      "step": 3004
+    },
+    {
+      "epoch": 0.45983167559296095,
+      "grad_norm": 2.2016042252587416,
+      "learning_rate": 1.7977140461143724e-05,
+      "loss": 0.7491,
+      "step": 3005
+    },
+    {
+      "epoch": 0.4599846977811783,
+      "grad_norm": 2.2356987100497134,
+      "learning_rate": 1.797564578285077e-05,
+      "loss": 0.7265,
+      "step": 3006
+    },
+    {
+      "epoch": 0.46013771996939556,
+      "grad_norm": 2.2342192382296,
+      "learning_rate": 1.797415061474221e-05,
+      "loss": 0.8228,
+      "step": 3007
+    },
+    {
+      "epoch": 0.46029074215761284,
+      "grad_norm": 2.4500043399004334,
+      "learning_rate": 1.7972654956909864e-05,
+      "loss": 0.755,
+      "step": 3008
+    },
+    {
+      "epoch": 0.4604437643458301,
+      "grad_norm": 2.0413908159279393,
+      "learning_rate": 1.7971158809445586e-05,
+      "loss": 0.6604,
+      "step": 3009
+    },
+    {
+      "epoch": 0.46059678653404745,
+      "grad_norm": 2.4438109905320067,
+      "learning_rate": 1.7969662172441266e-05,
+      "loss": 0.7801,
+      "step": 3010
+    },
+    {
+      "epoch": 0.46074980872226473,
+      "grad_norm": 2.5796430716040493,
+      "learning_rate": 1.796816504598881e-05,
+      "loss": 0.7737,
+      "step": 3011
+    },
+    {
+      "epoch": 0.460902830910482,
+      "grad_norm": 2.4072337562016366,
+      "learning_rate": 1.7966667430180168e-05,
+      "loss": 0.7351,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4610558530986993,
+      "grad_norm": 2.4134621653012025,
+      "learning_rate": 1.7965169325107316e-05,
+      "loss": 0.679,
+      "step": 3013
+    },
+    {
+      "epoch": 0.4612088752869166,
+      "grad_norm": 2.382826189730748,
+      "learning_rate": 1.7963670730862253e-05,
+      "loss": 0.8454,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4613618974751339,
+      "grad_norm": 2.461863361146385,
+      "learning_rate": 1.796217164753702e-05,
+      "loss": 0.8673,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4615149196633512,
+      "grad_norm": 2.162119030473225,
+      "learning_rate": 1.7960672075223674e-05,
+      "loss": 0.7094,
+      "step": 3016
+    },
+    {
+      "epoch": 0.46166794185156845,
+      "grad_norm": 2.4495640775338123,
+      "learning_rate": 1.7959172014014315e-05,
+      "loss": 0.8153,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4618209640397858,
+      "grad_norm": 2.8686927876563653,
+      "learning_rate": 1.7957671464001064e-05,
+      "loss": 0.7448,
+      "step": 3018
+    },
+    {
+      "epoch": 0.46197398622800306,
+      "grad_norm": 2.375874766610349,
+      "learning_rate": 1.7956170425276084e-05,
+      "loss": 0.7054,
+      "step": 3019
+    },
+    {
+      "epoch": 0.46212700841622034,
+      "grad_norm": 2.2043885507178445,
+      "learning_rate": 1.7954668897931545e-05,
+      "loss": 0.8179,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4622800306044376,
+      "grad_norm": 2.2472971973036433,
+      "learning_rate": 1.7953166882059677e-05,
+      "loss": 0.8027,
+      "step": 3021
+    },
+    {
+      "epoch": 0.46243305279265495,
+      "grad_norm": 2.1792083036915866,
+      "learning_rate": 1.7951664377752714e-05,
+      "loss": 0.6854,
+      "step": 3022
+    },
+    {
+      "epoch": 0.46258607498087223,
+      "grad_norm": 2.4282694085901917,
+      "learning_rate": 1.7950161385102936e-05,
+      "loss": 0.8682,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4627390971690895,
+      "grad_norm": 2.5164884020527114,
+      "learning_rate": 1.7948657904202645e-05,
+      "loss": 0.7298,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4628921193573068,
+      "grad_norm": 2.2296415956211653,
+      "learning_rate": 1.7947153935144182e-05,
+      "loss": 0.7645,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4630451415455241,
+      "grad_norm": 2.4863277949035556,
+      "learning_rate": 1.79456494780199e-05,
+      "loss": 0.8104,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4631981637337414,
+      "grad_norm": 2.5801820662776995,
+      "learning_rate": 1.79441445329222e-05,
+      "loss": 0.8021,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4633511859219587,
+      "grad_norm": 2.6857176336923536,
+      "learning_rate": 1.794263909994351e-05,
+      "loss": 0.8011,
+      "step": 3028
+    },
+    {
+      "epoch": 0.46350420811017595,
+      "grad_norm": 2.5215164941271895,
+      "learning_rate": 1.794113317917628e-05,
+      "loss": 0.9214,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4636572302983933,
+      "grad_norm": 2.3776208831265877,
+      "learning_rate": 1.7939626770712996e-05,
+      "loss": 0.7993,
+      "step": 3030
+    },
+    {
+      "epoch": 0.46381025248661056,
+      "grad_norm": 2.3370137755310134,
+      "learning_rate": 1.7938119874646173e-05,
+      "loss": 0.9326,
+      "step": 3031
+    },
+    {
+      "epoch": 0.46396327467482784,
+      "grad_norm": 2.2444810838602196,
+      "learning_rate": 1.793661249106835e-05,
+      "loss": 0.78,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4641162968630451,
+      "grad_norm": 2.626977361377603,
+      "learning_rate": 1.793510462007211e-05,
+      "loss": 0.8295,
+      "step": 3033
+    },
+    {
+      "epoch": 0.46426931905126245,
+      "grad_norm": 2.279663603934519,
+      "learning_rate": 1.793359626175005e-05,
+      "loss": 0.7931,
+      "step": 3034
+    },
+    {
+      "epoch": 0.46442234123947973,
+      "grad_norm": 2.2209566116817783,
+      "learning_rate": 1.793208741619481e-05,
+      "loss": 0.8991,
+      "step": 3035
+    },
+    {
+      "epoch": 0.464575363427697,
+      "grad_norm": 2.3367089021530467,
+      "learning_rate": 1.793057808349905e-05,
+      "loss": 0.6924,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4647283856159143,
+      "grad_norm": 2.416662478259207,
+      "learning_rate": 1.7929068263755465e-05,
+      "loss": 0.9007,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4648814078041316,
+      "grad_norm": 2.3202236234576756,
+      "learning_rate": 1.792755795705678e-05,
+      "loss": 0.8085,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4650344299923489,
+      "grad_norm": 2.2844609011360815,
+      "learning_rate": 1.792604716349575e-05,
+      "loss": 0.5956,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4651874521805662,
+      "grad_norm": 2.1019501713247006,
+      "learning_rate": 1.7924535883165154e-05,
+      "loss": 0.7306,
+      "step": 3040
+    },
+    {
+      "epoch": 0.46534047436878345,
+      "grad_norm": 2.371476838905052,
+      "learning_rate": 1.7923024116157806e-05,
+      "loss": 0.7912,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4654934965570008,
+      "grad_norm": 2.2453509974503847,
+      "learning_rate": 1.7921511862566557e-05,
+      "loss": 0.7831,
+      "step": 3042
+    },
+    {
+      "epoch": 0.46564651874521806,
+      "grad_norm": 2.516526580358889,
+      "learning_rate": 1.7919999122484273e-05,
+      "loss": 0.8007,
+      "step": 3043
+    },
+    {
+      "epoch": 0.46579954093343534,
+      "grad_norm": 2.3714993400159288,
+      "learning_rate": 1.791848589600386e-05,
+      "loss": 0.8676,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4659525631216526,
+      "grad_norm": 2.4338479274309766,
+      "learning_rate": 1.791697218321825e-05,
+      "loss": 0.7427,
+      "step": 3045
+    },
+    {
+      "epoch": 0.46610558530986995,
+      "grad_norm": 2.377571211070433,
+      "learning_rate": 1.7915457984220406e-05,
+      "loss": 0.8151,
+      "step": 3046
+    },
+    {
+      "epoch": 0.46625860749808723,
+      "grad_norm": 2.3621546987748907,
+      "learning_rate": 1.7913943299103325e-05,
+      "loss": 0.7547,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4664116296863045,
+      "grad_norm": 2.339112355183538,
+      "learning_rate": 1.7912428127960025e-05,
+      "loss": 0.83,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4665646518745218,
+      "grad_norm": 2.685139879857841,
+      "learning_rate": 1.791091247088356e-05,
+      "loss": 0.8156,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4667176740627391,
+      "grad_norm": 2.531331145019246,
+      "learning_rate": 1.790939632796701e-05,
+      "loss": 0.7796,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4668706962509564,
+      "grad_norm": 2.4861225552288864,
+      "learning_rate": 1.7907879699303494e-05,
+      "loss": 0.7717,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4670237184391737,
+      "grad_norm": 2.310760243025036,
+      "learning_rate": 1.790636258498615e-05,
+      "loss": 0.7763,
+      "step": 3052
+    },
+    {
+      "epoch": 0.46717674062739095,
+      "grad_norm": 2.3903087000299084,
+      "learning_rate": 1.7904844985108146e-05,
+      "loss": 0.7714,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4673297628156083,
+      "grad_norm": 2.2206279036603482,
+      "learning_rate": 1.790332689976269e-05,
+      "loss": 0.7524,
+      "step": 3054
+    },
+    {
+      "epoch": 0.46748278500382556,
+      "grad_norm": 2.3187476587864095,
+      "learning_rate": 1.790180832904301e-05,
+      "loss": 0.8059,
+      "step": 3055
+    },
+    {
+      "epoch": 0.46763580719204284,
+      "grad_norm": 2.710134518322272,
+      "learning_rate": 1.7900289273042367e-05,
+      "loss": 0.7208,
+      "step": 3056
+    },
+    {
+      "epoch": 0.4677888293802601,
+      "grad_norm": 2.35421789922996,
+      "learning_rate": 1.7898769731854058e-05,
+      "loss": 0.7486,
+      "step": 3057
+    },
+    {
+      "epoch": 0.46794185156847745,
+      "grad_norm": 2.286655013492438,
+      "learning_rate": 1.7897249705571397e-05,
+      "loss": 0.7359,
+      "step": 3058
+    },
+    {
+      "epoch": 0.46809487375669473,
+      "grad_norm": 2.11297135275565,
+      "learning_rate": 1.7895729194287733e-05,
+      "loss": 0.6781,
+      "step": 3059
+    },
+    {
+      "epoch": 0.468247895944912,
+      "grad_norm": 2.406508933681986,
+      "learning_rate": 1.789420819809646e-05,
+      "loss": 0.7598,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4684009181331293,
+      "grad_norm": 2.6366510236837075,
+      "learning_rate": 1.7892686717090972e-05,
+      "loss": 0.9273,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4685539403213466,
+      "grad_norm": 2.532411121338478,
+      "learning_rate": 1.789116475136472e-05,
+      "loss": 0.7786,
+      "step": 3062
+    },
+    {
+      "epoch": 0.4687069625095639,
+      "grad_norm": 2.594310936151825,
+      "learning_rate": 1.788964230101117e-05,
+      "loss": 0.8882,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4688599846977812,
+      "grad_norm": 2.358180777677221,
+      "learning_rate": 1.788811936612382e-05,
+      "loss": 0.835,
+      "step": 3064
+    },
+    {
+      "epoch": 0.46901300688599845,
+      "grad_norm": 2.2454409418760903,
+      "learning_rate": 1.7886595946796202e-05,
+      "loss": 0.8277,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4691660290742158,
+      "grad_norm": 2.329341381250108,
+      "learning_rate": 1.7885072043121874e-05,
+      "loss": 0.8119,
+      "step": 3066
+    },
+    {
+      "epoch": 0.46931905126243306,
+      "grad_norm": 1.9843107842254468,
+      "learning_rate": 1.7883547655194426e-05,
+      "loss": 0.7151,
+      "step": 3067
+    },
+    {
+      "epoch": 0.46947207345065034,
+      "grad_norm": 2.3293478291000755,
+      "learning_rate": 1.7882022783107475e-05,
+      "loss": 0.7741,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4696250956388676,
+      "grad_norm": 2.4350391900241997,
+      "learning_rate": 1.7880497426954673e-05,
+      "loss": 0.8019,
+      "step": 3069
+    },
+    {
+      "epoch": 0.46977811782708495,
+      "grad_norm": 2.144141917751492,
+      "learning_rate": 1.787897158682969e-05,
+      "loss": 0.7007,
+      "step": 3070
+    },
+    {
+      "epoch": 0.46993114001530223,
+      "grad_norm": 2.174973634706149,
+      "learning_rate": 1.787744526282625e-05,
+      "loss": 0.792,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4700841622035195,
+      "grad_norm": 2.618268443070123,
+      "learning_rate": 1.7875918455038068e-05,
+      "loss": 0.9176,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4702371843917368,
+      "grad_norm": 2.128847386701305,
+      "learning_rate": 1.7874391163558926e-05,
+      "loss": 0.7339,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4703902065799541,
+      "grad_norm": 2.4522718758498314,
+      "learning_rate": 1.787286338848262e-05,
+      "loss": 0.8111,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4705432287681714,
+      "grad_norm": 2.2474549670739736,
+      "learning_rate": 1.7871335129902974e-05,
+      "loss": 0.8114,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4706962509563887,
+      "grad_norm": 2.4366628479396724,
+      "learning_rate": 1.7869806387913845e-05,
+      "loss": 0.8638,
+      "step": 3076
+    },
+    {
+      "epoch": 0.47084927314460595,
+      "grad_norm": 2.220474175033701,
+      "learning_rate": 1.7868277162609116e-05,
+      "loss": 0.7199,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4710022953328233,
+      "grad_norm": 2.4551268926851,
+      "learning_rate": 1.786674745408271e-05,
+      "loss": 0.7692,
+      "step": 3078
+    },
+    {
+      "epoch": 0.47115531752104056,
+      "grad_norm": 2.464234476027355,
+      "learning_rate": 1.7865217262428564e-05,
+      "loss": 0.8443,
+      "step": 3079
+    },
+    {
+      "epoch": 0.47130833970925784,
+      "grad_norm": 2.422126071311725,
+      "learning_rate": 1.786368658774066e-05,
+      "loss": 0.742,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4714613618974751,
+      "grad_norm": 2.5067019319029566,
+      "learning_rate": 1.7862155430112997e-05,
+      "loss": 0.7451,
+      "step": 3081
+    },
+    {
+      "epoch": 0.47161438408569245,
+      "grad_norm": 2.260090103969428,
+      "learning_rate": 1.7860623789639614e-05,
+      "loss": 0.715,
+      "step": 3082
+    },
+    {
+      "epoch": 0.47176740627390973,
+      "grad_norm": 2.57018335192414,
+      "learning_rate": 1.7859091666414574e-05,
+      "loss": 0.8114,
+      "step": 3083
+    },
+    {
+      "epoch": 0.471920428462127,
+      "grad_norm": 2.3585692516115517,
+      "learning_rate": 1.7857559060531967e-05,
+      "loss": 0.7992,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4720734506503443,
+      "grad_norm": 2.35928380554881,
+      "learning_rate": 1.785602597208592e-05,
+      "loss": 0.8346,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4722264728385616,
+      "grad_norm": 2.462727119352491,
+      "learning_rate": 1.7854492401170587e-05,
+      "loss": 0.6754,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4723794950267789,
+      "grad_norm": 2.196818002617913,
+      "learning_rate": 1.7852958347880147e-05,
+      "loss": 0.832,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4725325172149962,
+      "grad_norm": 2.449395292618889,
+      "learning_rate": 1.7851423812308814e-05,
+      "loss": 0.8474,
+      "step": 3088
+    },
+    {
+      "epoch": 0.47268553940321345,
+      "grad_norm": 2.0127666982797643,
+      "learning_rate": 1.784988879455083e-05,
+      "loss": 0.6382,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4728385615914308,
+      "grad_norm": 2.339434631709559,
+      "learning_rate": 1.7848353294700467e-05,
+      "loss": 0.7453,
+      "step": 3090
+    },
+    {
+      "epoch": 0.47299158377964806,
+      "grad_norm": 2.36812382885095,
+      "learning_rate": 1.7846817312852025e-05,
+      "loss": 0.7293,
+      "step": 3091
+    },
+    {
+      "epoch": 0.47314460596786534,
+      "grad_norm": 2.3238439044790975,
+      "learning_rate": 1.7845280849099835e-05,
+      "loss": 0.8259,
+      "step": 3092
+    },
+    {
+      "epoch": 0.4732976281560826,
+      "grad_norm": 2.26683061558248,
+      "learning_rate": 1.7843743903538255e-05,
+      "loss": 0.7869,
+      "step": 3093
+    },
+    {
+      "epoch": 0.47345065034429995,
+      "grad_norm": 2.2685721302716657,
+      "learning_rate": 1.784220647626168e-05,
+      "loss": 0.7982,
+      "step": 3094
+    },
+    {
+      "epoch": 0.47360367253251723,
+      "grad_norm": 2.1515562913610893,
+      "learning_rate": 1.7840668567364524e-05,
+      "loss": 0.7836,
+      "step": 3095
+    },
+    {
+      "epoch": 0.4737566947207345,
+      "grad_norm": 2.2986525906713404,
+      "learning_rate": 1.783913017694124e-05,
+      "loss": 0.7999,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4739097169089518,
+      "grad_norm": 2.14804110147054,
+      "learning_rate": 1.7837591305086304e-05,
+      "loss": 0.8341,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4740627390971691,
+      "grad_norm": 2.032811223621796,
+      "learning_rate": 1.7836051951894226e-05,
+      "loss": 0.7648,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4742157612853864,
+      "grad_norm": 2.366229763538631,
+      "learning_rate": 1.7834512117459542e-05,
+      "loss": 0.7466,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4743687834736037,
+      "grad_norm": 2.4477203066920836,
+      "learning_rate": 1.783297180187682e-05,
+      "loss": 0.8449,
+      "step": 3100
+    },
+    {
+      "epoch": 0.47452180566182095,
+      "grad_norm": 2.4758677333507015,
+      "learning_rate": 1.7831431005240655e-05,
+      "loss": 0.879,
+      "step": 3101
+    },
+    {
+      "epoch": 0.47467482785003823,
+      "grad_norm": 2.175130815541815,
+      "learning_rate": 1.782988972764568e-05,
+      "loss": 0.8596,
+      "step": 3102
+    },
+    {
+      "epoch": 0.47482785003825556,
+      "grad_norm": 2.6280119624613936,
+      "learning_rate": 1.7828347969186542e-05,
+      "loss": 0.8391,
+      "step": 3103
+    },
+    {
+      "epoch": 0.47498087222647284,
+      "grad_norm": 2.3033026539480175,
+      "learning_rate": 1.782680572995793e-05,
+      "loss": 0.8477,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4751338944146901,
+      "grad_norm": 2.3324067024311006,
+      "learning_rate": 1.7825263010054565e-05,
+      "loss": 0.9281,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4752869166029074,
+      "grad_norm": 2.4416725086218873,
+      "learning_rate": 1.7823719809571183e-05,
+      "loss": 0.7544,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4754399387911247,
+      "grad_norm": 2.3461897741365663,
+      "learning_rate": 1.782217612860256e-05,
+      "loss": 0.7901,
+      "step": 3107
+    },
+    {
+      "epoch": 0.475592960979342,
+      "grad_norm": 2.258825700505424,
+      "learning_rate": 1.7820631967243502e-05,
+      "loss": 0.74,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4757459831675593,
+      "grad_norm": 2.236350731521977,
+      "learning_rate": 1.781908732558884e-05,
+      "loss": 0.8043,
+      "step": 3109
+    },
+    {
+      "epoch": 0.47589900535577656,
+      "grad_norm": 2.306541883073562,
+      "learning_rate": 1.7817542203733435e-05,
+      "loss": 0.7783,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4760520275439939,
+      "grad_norm": 2.132024776419034,
+      "learning_rate": 1.781599660177218e-05,
+      "loss": 0.7408,
+      "step": 3111
+    },
+    {
+      "epoch": 0.47620504973221117,
+      "grad_norm": 2.3182129805418006,
+      "learning_rate": 1.7814450519800003e-05,
+      "loss": 0.8,
+      "step": 3112
+    },
+    {
+      "epoch": 0.47635807192042845,
+      "grad_norm": 2.7774847658958324,
+      "learning_rate": 1.7812903957911845e-05,
+      "loss": 0.8339,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4765110941086457,
+      "grad_norm": 2.4081664390527022,
+      "learning_rate": 1.781135691620269e-05,
+      "loss": 0.8104,
+      "step": 3114
+    },
+    {
+      "epoch": 0.47666411629686306,
+      "grad_norm": 2.2680414415315977,
+      "learning_rate": 1.7809809394767553e-05,
+      "loss": 0.7734,
+      "step": 3115
+    },
+    {
+      "epoch": 0.47681713848508034,
+      "grad_norm": 2.5455145450874874,
+      "learning_rate": 1.7808261393701466e-05,
+      "loss": 0.6636,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4769701606732976,
+      "grad_norm": 2.402374800475588,
+      "learning_rate": 1.78067129130995e-05,
+      "loss": 0.8384,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4771231828615149,
+      "grad_norm": 2.3897889539372366,
+      "learning_rate": 1.7805163953056755e-05,
+      "loss": 0.8397,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4772762050497322,
+      "grad_norm": 2.4403644458499887,
+      "learning_rate": 1.7803614513668357e-05,
+      "loss": 0.8328,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4774292272379495,
+      "grad_norm": 2.298660833843685,
+      "learning_rate": 1.7802064595029463e-05,
+      "loss": 0.7801,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4775822494261668,
+      "grad_norm": 2.488715580053673,
+      "learning_rate": 1.7800514197235262e-05,
+      "loss": 0.9101,
+      "step": 3121
+    },
+    {
+      "epoch": 0.47773527161438406,
+      "grad_norm": 2.377244202377784,
+      "learning_rate": 1.7798963320380965e-05,
+      "loss": 0.7036,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4778882938026014,
+      "grad_norm": 2.4289693996129165,
+      "learning_rate": 1.7797411964561825e-05,
+      "loss": 0.814,
+      "step": 3123
+    },
+    {
+      "epoch": 0.47804131599081867,
+      "grad_norm": 2.5574585527951905,
+      "learning_rate": 1.7795860129873113e-05,
+      "loss": 0.8414,
+      "step": 3124
+    },
+    {
+      "epoch": 0.47819433817903595,
+      "grad_norm": 2.145881998828633,
+      "learning_rate": 1.779430781641013e-05,
+      "loss": 0.7363,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4783473603672532,
+      "grad_norm": 2.2443790813070317,
+      "learning_rate": 1.7792755024268213e-05,
+      "loss": 0.8263,
+      "step": 3126
+    },
+    {
+      "epoch": 0.47850038255547056,
+      "grad_norm": 2.4804357389915483,
+      "learning_rate": 1.7791201753542725e-05,
+      "loss": 0.9297,
+      "step": 3127
+    },
+    {
+      "epoch": 0.47865340474368784,
+      "grad_norm": 2.409155080634445,
+      "learning_rate": 1.778964800432906e-05,
+      "loss": 0.8459,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4788064269319051,
+      "grad_norm": 2.0586342296667213,
+      "learning_rate": 1.7788093776722634e-05,
+      "loss": 0.7397,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4789594491201224,
+      "grad_norm": 2.3936689258666943,
+      "learning_rate": 1.7786539070818903e-05,
+      "loss": 0.8269,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4791124713083397,
+      "grad_norm": 2.359822665681357,
+      "learning_rate": 1.778498388671335e-05,
+      "loss": 0.7837,
+      "step": 3131
+    },
+    {
+      "epoch": 0.479265493496557,
+      "grad_norm": 2.6016378686683455,
+      "learning_rate": 1.778342822450148e-05,
+      "loss": 0.8188,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4794185156847743,
+      "grad_norm": 2.2391289244689783,
+      "learning_rate": 1.778187208427883e-05,
+      "loss": 0.7139,
+      "step": 3133
+    },
+    {
+      "epoch": 0.47957153787299156,
+      "grad_norm": 2.4395509973947904,
+      "learning_rate": 1.7780315466140975e-05,
+      "loss": 0.6954,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4797245600612089,
+      "grad_norm": 2.3167840415266783,
+      "learning_rate": 1.777875837018351e-05,
+      "loss": 0.8495,
+      "step": 3135
+    },
+    {
+      "epoch": 0.47987758224942617,
+      "grad_norm": 2.207197768229709,
+      "learning_rate": 1.7777200796502065e-05,
+      "loss": 0.7324,
+      "step": 3136
+    },
+    {
+      "epoch": 0.48003060443764345,
+      "grad_norm": 2.3388895918578405,
+      "learning_rate": 1.7775642745192293e-05,
+      "loss": 0.8092,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4801836266258607,
+      "grad_norm": 2.5991436129658134,
+      "learning_rate": 1.777408421634988e-05,
+      "loss": 0.7951,
+      "step": 3138
+    },
+    {
+      "epoch": 0.48033664881407806,
+      "grad_norm": 2.614271482790651,
+      "learning_rate": 1.7772525210070545e-05,
+      "loss": 0.9615,
+      "step": 3139
+    },
+    {
+      "epoch": 0.48048967100229534,
+      "grad_norm": 2.280529742595915,
+      "learning_rate": 1.777096572645003e-05,
+      "loss": 0.8814,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4806426931905126,
+      "grad_norm": 2.22455747968137,
+      "learning_rate": 1.7769405765584112e-05,
+      "loss": 0.7447,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4807957153787299,
+      "grad_norm": 1.9704088622500777,
+      "learning_rate": 1.7767845327568585e-05,
+      "loss": 0.6619,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4809487375669472,
+      "grad_norm": 2.440214460615727,
+      "learning_rate": 1.7766284412499295e-05,
+      "loss": 0.829,
+      "step": 3143
+    },
+    {
+      "epoch": 0.4811017597551645,
+      "grad_norm": 2.474091480639532,
+      "learning_rate": 1.7764723020472098e-05,
+      "loss": 0.8021,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4812547819433818,
+      "grad_norm": 2.159308906503231,
+      "learning_rate": 1.7763161151582878e-05,
+      "loss": 0.7575,
+      "step": 3145
+    },
+    {
+      "epoch": 0.48140780413159906,
+      "grad_norm": 2.2634318435235814,
+      "learning_rate": 1.7761598805927564e-05,
+      "loss": 0.7508,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4815608263198164,
+      "grad_norm": 2.2808387798097574,
+      "learning_rate": 1.7760035983602107e-05,
+      "loss": 0.8432,
+      "step": 3147
+    },
+    {
+      "epoch": 0.48171384850803367,
+      "grad_norm": 2.2333734359872213,
+      "learning_rate": 1.775847268470248e-05,
+      "loss": 0.7977,
+      "step": 3148
+    },
+    {
+      "epoch": 0.48186687069625095,
+      "grad_norm": 2.175497185361418,
+      "learning_rate": 1.7756908909324697e-05,
+      "loss": 0.8576,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4820198928844682,
+      "grad_norm": 2.141222377290015,
+      "learning_rate": 1.7755344657564792e-05,
+      "loss": 0.7083,
+      "step": 3150
+    },
+    {
+      "epoch": 0.48217291507268556,
+      "grad_norm": 2.204712165087772,
+      "learning_rate": 1.7753779929518834e-05,
+      "loss": 0.7488,
+      "step": 3151
+    },
+    {
+      "epoch": 0.48232593726090284,
+      "grad_norm": 2.6909546478854023,
+      "learning_rate": 1.7752214725282912e-05,
+      "loss": 0.8871,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4824789594491201,
+      "grad_norm": 2.4007662255888182,
+      "learning_rate": 1.775064904495316e-05,
+      "loss": 0.8097,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4826319816373374,
+      "grad_norm": 2.281531360791035,
+      "learning_rate": 1.774908288862573e-05,
+      "loss": 0.6746,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4827850038255547,
+      "grad_norm": 2.2461468476777973,
+      "learning_rate": 1.7747516256396805e-05,
+      "loss": 0.7814,
+      "step": 3155
+    },
+    {
+      "epoch": 0.482938026013772,
+      "grad_norm": 2.8940583997852145,
+      "learning_rate": 1.7745949148362598e-05,
+      "loss": 0.9052,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4830910482019893,
+      "grad_norm": 2.151146909327901,
+      "learning_rate": 1.774438156461935e-05,
+      "loss": 0.6908,
+      "step": 3157
+    },
+    {
+      "epoch": 0.48324407039020656,
+      "grad_norm": 2.215843800992492,
+      "learning_rate": 1.7742813505263337e-05,
+      "loss": 0.7155,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4833970925784239,
+      "grad_norm": 2.7562581153783814,
+      "learning_rate": 1.7741244970390854e-05,
+      "loss": 0.8132,
+      "step": 3159
+    },
+    {
+      "epoch": 0.48355011476664117,
+      "grad_norm": 2.151781872160964,
+      "learning_rate": 1.7739675960098233e-05,
+      "loss": 0.6925,
+      "step": 3160
+    },
+    {
+      "epoch": 0.48370313695485845,
+      "grad_norm": 2.31679038784991,
+      "learning_rate": 1.7738106474481838e-05,
+      "loss": 0.8587,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4838561591430757,
+      "grad_norm": 2.2757888463384135,
+      "learning_rate": 1.773653651363805e-05,
+      "loss": 0.7059,
+      "step": 3162
+    },
+    {
+      "epoch": 0.48400918133129306,
+      "grad_norm": 2.6012748147156315,
+      "learning_rate": 1.7734966077663288e-05,
+      "loss": 0.7631,
+      "step": 3163
+    },
+    {
+      "epoch": 0.48416220351951034,
+      "grad_norm": 2.27116501374633,
+      "learning_rate": 1.7733395166654e-05,
+      "loss": 0.7921,
+      "step": 3164
+    },
+    {
+      "epoch": 0.4843152257077276,
+      "grad_norm": 2.463737754185167,
+      "learning_rate": 1.7731823780706664e-05,
+      "loss": 0.8805,
+      "step": 3165
+    },
+    {
+      "epoch": 0.4844682478959449,
+      "grad_norm": 2.457864756479185,
+      "learning_rate": 1.773025191991778e-05,
+      "loss": 0.8466,
+      "step": 3166
+    },
+    {
+      "epoch": 0.4846212700841622,
+      "grad_norm": 2.3691514278135695,
+      "learning_rate": 1.7728679584383884e-05,
+      "loss": 0.7611,
+      "step": 3167
+    },
+    {
+      "epoch": 0.4847742922723795,
+      "grad_norm": 2.6690788196830746,
+      "learning_rate": 1.7727106774201542e-05,
+      "loss": 0.7784,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4849273144605968,
+      "grad_norm": 3.050441302794346,
+      "learning_rate": 1.7725533489467345e-05,
+      "loss": 0.9023,
+      "step": 3169
+    },
+    {
+      "epoch": 0.48508033664881406,
+      "grad_norm": 2.242247145906977,
+      "learning_rate": 1.7723959730277912e-05,
+      "loss": 0.6486,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4852333588370314,
+      "grad_norm": 2.171151620961211,
+      "learning_rate": 1.7722385496729896e-05,
+      "loss": 0.7588,
+      "step": 3171
+    },
+    {
+      "epoch": 0.48538638102524867,
+      "grad_norm": 2.1872229724795673,
+      "learning_rate": 1.7720810788919974e-05,
+      "loss": 0.7352,
+      "step": 3172
+    },
+    {
+      "epoch": 0.48553940321346595,
+      "grad_norm": 2.2381038298991487,
+      "learning_rate": 1.7719235606944864e-05,
+      "loss": 0.777,
+      "step": 3173
+    },
+    {
+      "epoch": 0.4856924254016832,
+      "grad_norm": 2.1952939393056026,
+      "learning_rate": 1.7717659950901293e-05,
+      "loss": 0.738,
+      "step": 3174
+    },
+    {
+      "epoch": 0.48584544758990056,
+      "grad_norm": 2.4037553430751863,
+      "learning_rate": 1.771608382088603e-05,
+      "loss": 0.7518,
+      "step": 3175
+    },
+    {
+      "epoch": 0.48599846977811784,
+      "grad_norm": 2.298974912728606,
+      "learning_rate": 1.771450721699588e-05,
+      "loss": 0.7406,
+      "step": 3176
+    },
+    {
+      "epoch": 0.4861514919663351,
+      "grad_norm": 2.4675400243903214,
+      "learning_rate": 1.7712930139327663e-05,
+      "loss": 0.8112,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4863045141545524,
+      "grad_norm": 2.332643028056019,
+      "learning_rate": 1.771135258797823e-05,
+      "loss": 0.7814,
+      "step": 3178
+    },
+    {
+      "epoch": 0.4864575363427697,
+      "grad_norm": 2.3206125208022685,
+      "learning_rate": 1.7709774563044467e-05,
+      "loss": 0.7689,
+      "step": 3179
+    },
+    {
+      "epoch": 0.486610558530987,
+      "grad_norm": 2.2311701929496537,
+      "learning_rate": 1.7708196064623288e-05,
+      "loss": 0.8517,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4867635807192043,
+      "grad_norm": 1.9798751484367991,
+      "learning_rate": 1.7706617092811634e-05,
+      "loss": 0.7107,
+      "step": 3181
+    },
+    {
+      "epoch": 0.48691660290742156,
+      "grad_norm": 2.665140548734543,
+      "learning_rate": 1.7705037647706476e-05,
+      "loss": 0.8483,
+      "step": 3182
+    },
+    {
+      "epoch": 0.4870696250956389,
+      "grad_norm": 2.3060955142676525,
+      "learning_rate": 1.7703457729404814e-05,
+      "loss": 0.7595,
+      "step": 3183
+    },
+    {
+      "epoch": 0.48722264728385617,
+      "grad_norm": 2.201706180521868,
+      "learning_rate": 1.7701877338003677e-05,
+      "loss": 0.6579,
+      "step": 3184
+    },
+    {
+      "epoch": 0.48737566947207345,
+      "grad_norm": 2.4863780424056436,
+      "learning_rate": 1.7700296473600122e-05,
+      "loss": 0.8218,
+      "step": 3185
+    },
+    {
+      "epoch": 0.4875286916602907,
+      "grad_norm": 2.5234475881120972,
+      "learning_rate": 1.7698715136291238e-05,
+      "loss": 0.9158,
+      "step": 3186
+    },
+    {
+      "epoch": 0.48768171384850806,
+      "grad_norm": 2.230782070237414,
+      "learning_rate": 1.769713332617414e-05,
+      "loss": 0.8134,
+      "step": 3187
+    },
+    {
+      "epoch": 0.48783473603672534,
+      "grad_norm": 2.555470250956034,
+      "learning_rate": 1.7695551043345972e-05,
+      "loss": 0.7862,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4879877582249426,
+      "grad_norm": 2.435982366105735,
+      "learning_rate": 1.7693968287903906e-05,
+      "loss": 0.8112,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4881407804131599,
+      "grad_norm": 2.079276970212424,
+      "learning_rate": 1.7692385059945154e-05,
+      "loss": 0.7436,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4882938026013772,
+      "grad_norm": 2.503362168438321,
+      "learning_rate": 1.769080135956694e-05,
+      "loss": 0.8012,
+      "step": 3191
+    },
+    {
+      "epoch": 0.4884468247895945,
+      "grad_norm": 2.3646623422395905,
+      "learning_rate": 1.768921718686653e-05,
+      "loss": 0.8065,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4885998469778118,
+      "grad_norm": 2.428668920391647,
+      "learning_rate": 1.7687632541941208e-05,
+      "loss": 0.8703,
+      "step": 3193
+    },
+    {
+      "epoch": 0.48875286916602906,
+      "grad_norm": 2.592725000921327,
+      "learning_rate": 1.7686047424888303e-05,
+      "loss": 0.9091,
+      "step": 3194
+    },
+    {
+      "epoch": 0.4889058913542464,
+      "grad_norm": 2.3455071603658335,
+      "learning_rate": 1.7684461835805153e-05,
+      "loss": 0.7535,
+      "step": 3195
+    },
+    {
+      "epoch": 0.48905891354246367,
+      "grad_norm": 2.7711528247671837,
+      "learning_rate": 1.7682875774789143e-05,
+      "loss": 0.8585,
+      "step": 3196
+    },
+    {
+      "epoch": 0.48921193573068095,
+      "grad_norm": 2.5378645714466836,
+      "learning_rate": 1.7681289241937675e-05,
+      "loss": 0.806,
+      "step": 3197
+    },
+    {
+      "epoch": 0.4893649579188982,
+      "grad_norm": 2.302542926603842,
+      "learning_rate": 1.7679702237348185e-05,
+      "loss": 0.775,
+      "step": 3198
+    },
+    {
+      "epoch": 0.48951798010711556,
+      "grad_norm": 2.5459949407122777,
+      "learning_rate": 1.7678114761118136e-05,
+      "loss": 0.8727,
+      "step": 3199
+    },
+    {
+      "epoch": 0.48967100229533284,
+      "grad_norm": 2.448262018505535,
+      "learning_rate": 1.7676526813345024e-05,
+      "loss": 0.7077,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4898240244835501,
+      "grad_norm": 2.622412414805754,
+      "learning_rate": 1.7674938394126368e-05,
+      "loss": 1.0291,
+      "step": 3201
+    },
+    {
+      "epoch": 0.4899770466717674,
+      "grad_norm": 2.4738400913098495,
+      "learning_rate": 1.7673349503559728e-05,
+      "loss": 0.857,
+      "step": 3202
+    },
+    {
+      "epoch": 0.4901300688599847,
+      "grad_norm": 2.5134079518079253,
+      "learning_rate": 1.767176014174267e-05,
+      "loss": 0.8002,
+      "step": 3203
+    },
+    {
+      "epoch": 0.490283091048202,
+      "grad_norm": 13.524370615262251,
+      "learning_rate": 1.767017030877281e-05,
+      "loss": 0.8024,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4904361132364193,
+      "grad_norm": 2.404277059545446,
+      "learning_rate": 1.7668580004747787e-05,
+      "loss": 0.769,
+      "step": 3205
+    },
+    {
+      "epoch": 0.49058913542463656,
+      "grad_norm": 2.2606428101347795,
+      "learning_rate": 1.7666989229765265e-05,
+      "loss": 0.7535,
+      "step": 3206
+    },
+    {
+      "epoch": 0.4907421576128539,
+      "grad_norm": 2.64388311392485,
+      "learning_rate": 1.7665397983922942e-05,
+      "loss": 0.906,
+      "step": 3207
+    },
+    {
+      "epoch": 0.49089517980107117,
+      "grad_norm": 2.5906710909288977,
+      "learning_rate": 1.7663806267318538e-05,
+      "loss": 0.7953,
+      "step": 3208
+    },
+    {
+      "epoch": 0.49104820198928845,
+      "grad_norm": 2.272625903337256,
+      "learning_rate": 1.7662214080049814e-05,
+      "loss": 0.7149,
+      "step": 3209
+    },
+    {
+      "epoch": 0.4912012241775057,
+      "grad_norm": 2.078589092936746,
+      "learning_rate": 1.7660621422214546e-05,
+      "loss": 0.6869,
+      "step": 3210
+    },
+    {
+      "epoch": 0.49135424636572306,
+      "grad_norm": 2.1478637518748904,
+      "learning_rate": 1.765902829391055e-05,
+      "loss": 0.7486,
+      "step": 3211
+    },
+    {
+      "epoch": 0.49150726855394034,
+      "grad_norm": 2.3289060874748637,
+      "learning_rate": 1.7657434695235666e-05,
+      "loss": 0.7998,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4916602907421576,
+      "grad_norm": 2.4882931922578484,
+      "learning_rate": 1.7655840626287756e-05,
+      "loss": 0.7622,
+      "step": 3213
+    },
+    {
+      "epoch": 0.4918133129303749,
+      "grad_norm": 2.1951542390318712,
+      "learning_rate": 1.765424608716473e-05,
+      "loss": 0.7396,
+      "step": 3214
+    },
+    {
+      "epoch": 0.49196633511859217,
+      "grad_norm": 2.418546776586094,
+      "learning_rate": 1.7652651077964504e-05,
+      "loss": 0.8735,
+      "step": 3215
+    },
+    {
+      "epoch": 0.4921193573068095,
+      "grad_norm": 2.258923232410249,
+      "learning_rate": 1.7651055598785037e-05,
+      "loss": 0.8429,
+      "step": 3216
+    },
+    {
+      "epoch": 0.4922723794950268,
+      "grad_norm": 2.212119341949836,
+      "learning_rate": 1.7649459649724315e-05,
+      "loss": 0.691,
+      "step": 3217
+    },
+    {
+      "epoch": 0.49242540168324406,
+      "grad_norm": 2.5536534558659456,
+      "learning_rate": 1.7647863230880354e-05,
+      "loss": 0.7298,
+      "step": 3218
+    },
+    {
+      "epoch": 0.49257842387146134,
+      "grad_norm": 2.3804911347467574,
+      "learning_rate": 1.764626634235119e-05,
+      "loss": 0.857,
+      "step": 3219
+    },
+    {
+      "epoch": 0.49273144605967867,
+      "grad_norm": 2.1041892044354715,
+      "learning_rate": 1.76446689842349e-05,
+      "loss": 0.7818,
+      "step": 3220
+    },
+    {
+      "epoch": 0.49288446824789595,
+      "grad_norm": 2.3149475586078623,
+      "learning_rate": 1.764307115662958e-05,
+      "loss": 0.8303,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4930374904361132,
+      "grad_norm": 2.404275202101493,
+      "learning_rate": 1.7641472859633362e-05,
+      "loss": 0.8074,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4931905126243305,
+      "grad_norm": 2.7182580165570585,
+      "learning_rate": 1.76398740933444e-05,
+      "loss": 0.6691,
+      "step": 3223
+    },
+    {
+      "epoch": 0.49334353481254783,
+      "grad_norm": 2.2542269012648757,
+      "learning_rate": 1.7638274857860884e-05,
+      "loss": 0.7576,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4934965570007651,
+      "grad_norm": 2.6790158205282273,
+      "learning_rate": 1.763667515328103e-05,
+      "loss": 0.7793,
+      "step": 3225
+    },
+    {
+      "epoch": 0.4936495791889824,
+      "grad_norm": 2.3313374760580494,
+      "learning_rate": 1.7635074979703076e-05,
+      "loss": 0.8102,
+      "step": 3226
+    },
+    {
+      "epoch": 0.49380260137719967,
+      "grad_norm": 2.307545812369083,
+      "learning_rate": 1.76334743372253e-05,
+      "loss": 0.7408,
+      "step": 3227
+    },
+    {
+      "epoch": 0.493955623565417,
+      "grad_norm": 2.6064789126655907,
+      "learning_rate": 1.7631873225946004e-05,
+      "loss": 0.9119,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4941086457536343,
+      "grad_norm": 2.3624050343096727,
+      "learning_rate": 1.7630271645963518e-05,
+      "loss": 0.829,
+      "step": 3229
+    },
+    {
+      "epoch": 0.49426166794185156,
+      "grad_norm": 2.554313730527207,
+      "learning_rate": 1.7628669597376198e-05,
+      "loss": 0.9271,
+      "step": 3230
+    },
+    {
+      "epoch": 0.49441469013006883,
+      "grad_norm": 2.1998616907501995,
+      "learning_rate": 1.7627067080282432e-05,
+      "loss": 0.8112,
+      "step": 3231
+    },
+    {
+      "epoch": 0.49456771231828617,
+      "grad_norm": 2.363883278824949,
+      "learning_rate": 1.7625464094780646e-05,
+      "loss": 0.7588,
+      "step": 3232
+    },
+    {
+      "epoch": 0.49472073450650345,
+      "grad_norm": 2.4905261389828506,
+      "learning_rate": 1.7623860640969275e-05,
+      "loss": 0.7716,
+      "step": 3233
+    },
+    {
+      "epoch": 0.4948737566947207,
+      "grad_norm": 2.3865027559186447,
+      "learning_rate": 1.76222567189468e-05,
+      "loss": 0.7401,
+      "step": 3234
+    },
+    {
+      "epoch": 0.495026778882938,
+      "grad_norm": 2.483958830160686,
+      "learning_rate": 1.762065232881172e-05,
+      "loss": 0.8404,
+      "step": 3235
+    },
+    {
+      "epoch": 0.49517980107115533,
+      "grad_norm": 2.1902122015963448,
+      "learning_rate": 1.7619047470662565e-05,
+      "loss": 0.8056,
+      "step": 3236
+    },
+    {
+      "epoch": 0.4953328232593726,
+      "grad_norm": 2.399968282559192,
+      "learning_rate": 1.7617442144597902e-05,
+      "loss": 0.8102,
+      "step": 3237
+    },
+    {
+      "epoch": 0.4954858454475899,
+      "grad_norm": 2.1498314028722554,
+      "learning_rate": 1.7615836350716316e-05,
+      "loss": 0.7334,
+      "step": 3238
+    },
+    {
+      "epoch": 0.49563886763580717,
+      "grad_norm": 2.6122160629141424,
+      "learning_rate": 1.7614230089116428e-05,
+      "loss": 0.8052,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4957918898240245,
+      "grad_norm": 2.293516449978853,
+      "learning_rate": 1.7612623359896884e-05,
+      "loss": 0.7155,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4959449120122418,
+      "grad_norm": 2.3151533649954965,
+      "learning_rate": 1.7611016163156355e-05,
+      "loss": 0.7752,
+      "step": 3241
+    },
+    {
+      "epoch": 0.49609793420045906,
+      "grad_norm": 2.794173016381072,
+      "learning_rate": 1.7609408498993553e-05,
+      "loss": 0.9031,
+      "step": 3242
+    },
+    {
+      "epoch": 0.49625095638867633,
+      "grad_norm": 2.588196116806027,
+      "learning_rate": 1.7607800367507204e-05,
+      "loss": 0.8531,
+      "step": 3243
+    },
+    {
+      "epoch": 0.49640397857689367,
+      "grad_norm": 2.3351088057592615,
+      "learning_rate": 1.7606191768796074e-05,
+      "loss": 0.8684,
+      "step": 3244
+    },
+    {
+      "epoch": 0.49655700076511095,
+      "grad_norm": 2.32274229856413,
+      "learning_rate": 1.7604582702958953e-05,
+      "loss": 0.7669,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4967100229533282,
+      "grad_norm": 2.477436788357222,
+      "learning_rate": 1.7602973170094658e-05,
+      "loss": 0.8381,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4968630451415455,
+      "grad_norm": 2.6129214133461343,
+      "learning_rate": 1.7601363170302035e-05,
+      "loss": 0.793,
+      "step": 3247
+    },
+    {
+      "epoch": 0.49701606732976283,
+      "grad_norm": 2.172244742262806,
+      "learning_rate": 1.7599752703679966e-05,
+      "loss": 0.624,
+      "step": 3248
+    },
+    {
+      "epoch": 0.4971690895179801,
+      "grad_norm": 2.298575304630846,
+      "learning_rate": 1.7598141770327353e-05,
+      "loss": 0.7008,
+      "step": 3249
+    },
+    {
+      "epoch": 0.4973221117061974,
+      "grad_norm": 2.268141893500064,
+      "learning_rate": 1.759653037034313e-05,
+      "loss": 0.7647,
+      "step": 3250
+    },
+    {
+      "epoch": 0.49747513389441467,
+      "grad_norm": 2.222876437768816,
+      "learning_rate": 1.7594918503826255e-05,
+      "loss": 0.6546,
+      "step": 3251
+    },
+    {
+      "epoch": 0.497628156082632,
+      "grad_norm": 2.752810421202548,
+      "learning_rate": 1.7593306170875727e-05,
+      "loss": 0.9053,
+      "step": 3252
+    },
+    {
+      "epoch": 0.4977811782708493,
+      "grad_norm": 2.4122049038132767,
+      "learning_rate": 1.7591693371590563e-05,
+      "loss": 0.8456,
+      "step": 3253
+    },
+    {
+      "epoch": 0.49793420045906656,
+      "grad_norm": 2.1764013379355425,
+      "learning_rate": 1.7590080106069807e-05,
+      "loss": 0.8315,
+      "step": 3254
+    },
+    {
+      "epoch": 0.49808722264728383,
+      "grad_norm": 2.375189125899373,
+      "learning_rate": 1.758846637441254e-05,
+      "loss": 0.7593,
+      "step": 3255
+    },
+    {
+      "epoch": 0.49824024483550117,
+      "grad_norm": 2.3610123855811675,
+      "learning_rate": 1.7586852176717867e-05,
+      "loss": 0.8311,
+      "step": 3256
+    },
+    {
+      "epoch": 0.49839326702371844,
+      "grad_norm": 2.6043618815139635,
+      "learning_rate": 1.7585237513084922e-05,
+      "loss": 0.7291,
+      "step": 3257
+    },
+    {
+      "epoch": 0.4985462892119357,
+      "grad_norm": 2.3211779791516522,
+      "learning_rate": 1.758362238361287e-05,
+      "loss": 0.769,
+      "step": 3258
+    },
+    {
+      "epoch": 0.498699311400153,
+      "grad_norm": 2.2649520601965607,
+      "learning_rate": 1.7582006788400895e-05,
+      "loss": 0.705,
+      "step": 3259
+    },
+    {
+      "epoch": 0.49885233358837033,
+      "grad_norm": 2.439781656230531,
+      "learning_rate": 1.7580390727548228e-05,
+      "loss": 0.653,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4990053557765876,
+      "grad_norm": 2.448128881631369,
+      "learning_rate": 1.7578774201154107e-05,
+      "loss": 0.8099,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4991583779648049,
+      "grad_norm": 2.4505488863057416,
+      "learning_rate": 1.7577157209317817e-05,
+      "loss": 0.8893,
+      "step": 3262
+    },
+    {
+      "epoch": 0.49931140015302217,
+      "grad_norm": 2.5299595454688135,
+      "learning_rate": 1.757553975213866e-05,
+      "loss": 0.7847,
+      "step": 3263
+    },
+    {
+      "epoch": 0.4994644223412395,
+      "grad_norm": 2.7098718662487524,
+      "learning_rate": 1.757392182971597e-05,
+      "loss": 0.7942,
+      "step": 3264
+    },
+    {
+      "epoch": 0.4996174445294568,
+      "grad_norm": 2.3535479403390354,
+      "learning_rate": 1.757230344214911e-05,
+      "loss": 0.8274,
+      "step": 3265
+    },
+    {
+      "epoch": 0.49977046671767406,
+      "grad_norm": 2.1897548479389557,
+      "learning_rate": 1.7570684589537473e-05,
+      "loss": 0.8113,
+      "step": 3266
+    },
+    {
+      "epoch": 0.49992348890589133,
+      "grad_norm": 2.321407918265366,
+      "learning_rate": 1.7569065271980482e-05,
+      "loss": 0.8089,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5000765110941087,
+      "grad_norm": 2.5519453674642194,
+      "learning_rate": 1.756744548957758e-05,
+      "loss": 0.7425,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5002295332823259,
+      "grad_norm": 2.425437004623082,
+      "learning_rate": 1.756582524242825e-05,
+      "loss": 0.7746,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5003825554705432,
+      "grad_norm": 2.3121885815367405,
+      "learning_rate": 1.756420453063199e-05,
+      "loss": 0.8158,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5005355776587606,
+      "grad_norm": 2.378431932054567,
+      "learning_rate": 1.756258335428834e-05,
+      "loss": 0.7868,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5006885998469778,
+      "grad_norm": 2.007280263141698,
+      "learning_rate": 1.756096171349686e-05,
+      "loss": 0.8533,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5008416220351951,
+      "grad_norm": 2.522900136714207,
+      "learning_rate": 1.7559339608357145e-05,
+      "loss": 0.7948,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5009946442234124,
+      "grad_norm": 2.2461761449648905,
+      "learning_rate": 1.7557717038968807e-05,
+      "loss": 0.724,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5011476664116297,
+      "grad_norm": 2.492589978811216,
+      "learning_rate": 1.7556094005431507e-05,
+      "loss": 0.8158,
+      "step": 3275
+    },
+    {
+      "epoch": 0.501300688599847,
+      "grad_norm": 2.779932572876539,
+      "learning_rate": 1.7554470507844906e-05,
+      "loss": 0.7971,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5014537107880642,
+      "grad_norm": 2.726781264906075,
+      "learning_rate": 1.7552846546308724e-05,
+      "loss": 0.882,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5016067329762816,
+      "grad_norm": 2.750726908807502,
+      "learning_rate": 1.7551222120922687e-05,
+      "loss": 0.7143,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5017597551644989,
+      "grad_norm": 2.611633220188425,
+      "learning_rate": 1.7549597231786562e-05,
+      "loss": 0.9365,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5019127773527161,
+      "grad_norm": 2.193683141102787,
+      "learning_rate": 1.7547971879000133e-05,
+      "loss": 0.7702,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5020657995409334,
+      "grad_norm": 2.1799111635263384,
+      "learning_rate": 1.7546346062663224e-05,
+      "loss": 0.767,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5022188217291508,
+      "grad_norm": 2.2272737839801846,
+      "learning_rate": 1.754471978287568e-05,
+      "loss": 0.7561,
+      "step": 3282
+    },
+    {
+      "epoch": 0.502371843917368,
+      "grad_norm": 2.3756225531817337,
+      "learning_rate": 1.7543093039737384e-05,
+      "loss": 0.7519,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5025248661055853,
+      "grad_norm": 2.6892279006528206,
+      "learning_rate": 1.754146583334823e-05,
+      "loss": 0.7317,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5026778882938026,
+      "grad_norm": 2.167783051454944,
+      "learning_rate": 1.7539838163808164e-05,
+      "loss": 0.6927,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5028309104820199,
+      "grad_norm": 2.497432083363853,
+      "learning_rate": 1.7538210031217133e-05,
+      "loss": 0.7813,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5029839326702372,
+      "grad_norm": 2.42569547096876,
+      "learning_rate": 1.7536581435675136e-05,
+      "loss": 0.6925,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5031369548584544,
+      "grad_norm": 2.2400882700793336,
+      "learning_rate": 1.753495237728219e-05,
+      "loss": 0.828,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5032899770466718,
+      "grad_norm": 2.247722482445047,
+      "learning_rate": 1.7533322856138345e-05,
+      "loss": 0.7002,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5034429992348891,
+      "grad_norm": 2.507215137207525,
+      "learning_rate": 1.7531692872343672e-05,
+      "loss": 0.811,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5035960214231063,
+      "grad_norm": 2.3787834208881655,
+      "learning_rate": 1.753006242599827e-05,
+      "loss": 0.8742,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5037490436113237,
+      "grad_norm": 2.3822113508515392,
+      "learning_rate": 1.7528431517202283e-05,
+      "loss": 0.8242,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5039020657995409,
+      "grad_norm": 2.1981554588331984,
+      "learning_rate": 1.7526800146055866e-05,
+      "loss": 0.932,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5040550879877582,
+      "grad_norm": 2.2509071078466643,
+      "learning_rate": 1.7525168312659202e-05,
+      "loss": 0.7999,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5042081101759756,
+      "grad_norm": 2.4597949432094417,
+      "learning_rate": 1.7523536017112516e-05,
+      "loss": 0.7423,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5043611323641928,
+      "grad_norm": 2.2297846203516305,
+      "learning_rate": 1.7521903259516045e-05,
+      "loss": 0.6966,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5045141545524101,
+      "grad_norm": 2.4466722284928677,
+      "learning_rate": 1.752027003997008e-05,
+      "loss": 0.8328,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5046671767406274,
+      "grad_norm": 2.240130726906534,
+      "learning_rate": 1.7518636358574905e-05,
+      "loss": 0.7626,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5048201989288447,
+      "grad_norm": 2.4645099563125554,
+      "learning_rate": 1.7517002215430857e-05,
+      "loss": 0.8377,
+      "step": 3299
+    },
+    {
+      "epoch": 0.504973221117062,
+      "grad_norm": 2.173867927524907,
+      "learning_rate": 1.75153676106383e-05,
+      "loss": 0.7712,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5051262433052792,
+      "grad_norm": 2.538823450141907,
+      "learning_rate": 1.7513732544297618e-05,
+      "loss": 0.8464,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5052792654934966,
+      "grad_norm": 2.462672051119006,
+      "learning_rate": 1.7512097016509224e-05,
+      "loss": 0.8667,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5054322876817139,
+      "grad_norm": 2.4697382148859597,
+      "learning_rate": 1.7510461027373565e-05,
+      "loss": 0.726,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5055853098699311,
+      "grad_norm": 2.2938339035117727,
+      "learning_rate": 1.7508824576991115e-05,
+      "loss": 0.7386,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5057383320581484,
+      "grad_norm": 2.157524022621634,
+      "learning_rate": 1.7507187665462374e-05,
+      "loss": 0.7496,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5058913542463658,
+      "grad_norm": 2.1429953471957015,
+      "learning_rate": 1.7505550292887866e-05,
+      "loss": 0.7888,
+      "step": 3306
+    },
+    {
+      "epoch": 0.506044376434583,
+      "grad_norm": 2.1704731421991394,
+      "learning_rate": 1.7503912459368156e-05,
+      "loss": 0.8105,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5061973986228003,
+      "grad_norm": 2.5892926483115772,
+      "learning_rate": 1.7502274165003826e-05,
+      "loss": 0.8,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5063504208110176,
+      "grad_norm": 2.4993920181365654,
+      "learning_rate": 1.7500635409895488e-05,
+      "loss": 0.8512,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5065034429992349,
+      "grad_norm": 2.340329315102193,
+      "learning_rate": 1.7498996194143792e-05,
+      "loss": 0.7955,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5066564651874522,
+      "grad_norm": 2.3758682343089905,
+      "learning_rate": 1.7497356517849397e-05,
+      "loss": 0.7777,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5068094873756694,
+      "grad_norm": 2.430405212589665,
+      "learning_rate": 1.749571638111301e-05,
+      "loss": 0.9005,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5069625095638868,
+      "grad_norm": 2.295985134609166,
+      "learning_rate": 1.749407578403536e-05,
+      "loss": 0.6991,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5071155317521041,
+      "grad_norm": 2.2080978558305646,
+      "learning_rate": 1.74924347267172e-05,
+      "loss": 0.6966,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5072685539403213,
+      "grad_norm": 2.3125199843341386,
+      "learning_rate": 1.749079320925931e-05,
+      "loss": 0.9435,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5074215761285387,
+      "grad_norm": 2.3758545793328394,
+      "learning_rate": 1.7489151231762503e-05,
+      "loss": 0.7937,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5075745983167559,
+      "grad_norm": 2.327374821187704,
+      "learning_rate": 1.7487508794327627e-05,
+      "loss": 0.6983,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5077276205049732,
+      "grad_norm": 2.3625341694144413,
+      "learning_rate": 1.7485865897055537e-05,
+      "loss": 0.797,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5078806426931906,
+      "grad_norm": 2.4367362368564836,
+      "learning_rate": 1.748422254004714e-05,
+      "loss": 0.8072,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5080336648814078,
+      "grad_norm": 2.3174896950785433,
+      "learning_rate": 1.7482578723403363e-05,
+      "loss": 0.7571,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5081866870696251,
+      "grad_norm": 1.8791679422020524,
+      "learning_rate": 1.748093444722515e-05,
+      "loss": 0.7182,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5083397092578424,
+      "grad_norm": 2.1616486568516096,
+      "learning_rate": 1.7479289711613487e-05,
+      "loss": 0.7051,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5084927314460597,
+      "grad_norm": 2.230631601951075,
+      "learning_rate": 1.7477644516669387e-05,
+      "loss": 0.7449,
+      "step": 3323
+    },
+    {
+      "epoch": 0.508645753634277,
+      "grad_norm": 2.0062049354602225,
+      "learning_rate": 1.747599886249388e-05,
+      "loss": 0.7714,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5087987758224942,
+      "grad_norm": 2.5277809968626417,
+      "learning_rate": 1.747435274918804e-05,
+      "loss": 0.9102,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5089517980107116,
+      "grad_norm": 2.2964812751680723,
+      "learning_rate": 1.7472706176852957e-05,
+      "loss": 0.8331,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5091048201989289,
+      "grad_norm": 2.4469684135554983,
+      "learning_rate": 1.7471059145589755e-05,
+      "loss": 0.8294,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5092578423871461,
+      "grad_norm": 2.197412860546768,
+      "learning_rate": 1.7469411655499583e-05,
+      "loss": 0.8342,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5094108645753634,
+      "grad_norm": 2.5275013829253212,
+      "learning_rate": 1.746776370668362e-05,
+      "loss": 0.8247,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5095638867635807,
+      "grad_norm": 2.1610874400415803,
+      "learning_rate": 1.7466115299243072e-05,
+      "loss": 0.7758,
+      "step": 3330
+    },
+    {
+      "epoch": 0.509716908951798,
+      "grad_norm": 2.2429685448279733,
+      "learning_rate": 1.746446643327918e-05,
+      "loss": 0.7493,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5098699311400153,
+      "grad_norm": 2.4143447333401435,
+      "learning_rate": 1.7462817108893202e-05,
+      "loss": 0.7255,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5100229533282326,
+      "grad_norm": 2.3857507284179795,
+      "learning_rate": 1.746116732618643e-05,
+      "loss": 0.8139,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5101759755164499,
+      "grad_norm": 2.2535908652855094,
+      "learning_rate": 1.7459517085260186e-05,
+      "loss": 0.7173,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5103289977046672,
+      "grad_norm": 2.182942551591896,
+      "learning_rate": 1.7457866386215814e-05,
+      "loss": 0.6888,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5104820198928844,
+      "grad_norm": 2.297465645975313,
+      "learning_rate": 1.7456215229154693e-05,
+      "loss": 0.7634,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5106350420811018,
+      "grad_norm": 2.3307719054135156,
+      "learning_rate": 1.7454563614178223e-05,
+      "loss": 0.789,
+      "step": 3337
+    },
+    {
+      "epoch": 0.510788064269319,
+      "grad_norm": 2.380246890813595,
+      "learning_rate": 1.7452911541387844e-05,
+      "loss": 0.8229,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5109410864575363,
+      "grad_norm": 2.214607579600976,
+      "learning_rate": 1.7451259010885008e-05,
+      "loss": 0.7872,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5110941086457537,
+      "grad_norm": 2.5175143675456826,
+      "learning_rate": 1.7449606022771206e-05,
+      "loss": 0.9167,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5112471308339709,
+      "grad_norm": 2.2963224878966475,
+      "learning_rate": 1.7447952577147957e-05,
+      "loss": 0.8955,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5114001530221882,
+      "grad_norm": 2.6007771172167304,
+      "learning_rate": 1.7446298674116803e-05,
+      "loss": 0.7827,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5115531752104056,
+      "grad_norm": 2.619381696692319,
+      "learning_rate": 1.7444644313779316e-05,
+      "loss": 0.8316,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5117061973986228,
+      "grad_norm": 2.422295060929172,
+      "learning_rate": 1.74429894962371e-05,
+      "loss": 0.7311,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5118592195868401,
+      "grad_norm": 2.3663869854583544,
+      "learning_rate": 1.7441334221591783e-05,
+      "loss": 0.7173,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5120122417750573,
+      "grad_norm": 2.1570402298204225,
+      "learning_rate": 1.7439678489945017e-05,
+      "loss": 0.7325,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5121652639632747,
+      "grad_norm": 3.0964806046696944,
+      "learning_rate": 1.7438022301398495e-05,
+      "loss": 0.8387,
+      "step": 3347
+    },
+    {
+      "epoch": 0.512318286151492,
+      "grad_norm": 2.431537047730991,
+      "learning_rate": 1.743636565605392e-05,
+      "loss": 0.7897,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5124713083397092,
+      "grad_norm": 2.091834176573096,
+      "learning_rate": 1.7434708554013046e-05,
+      "loss": 0.7902,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5126243305279266,
+      "grad_norm": 2.1331625981738775,
+      "learning_rate": 1.743305099537763e-05,
+      "loss": 0.679,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5127773527161439,
+      "grad_norm": 2.5746781871427182,
+      "learning_rate": 1.7431392980249477e-05,
+      "loss": 0.8677,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5129303749043611,
+      "grad_norm": 2.30631341726066,
+      "learning_rate": 1.7429734508730404e-05,
+      "loss": 0.7585,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5130833970925784,
+      "grad_norm": 2.426958964202821,
+      "learning_rate": 1.7428075580922278e-05,
+      "loss": 0.8328,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5132364192807957,
+      "grad_norm": 2.4386540509387076,
+      "learning_rate": 1.7426416196926965e-05,
+      "loss": 0.7015,
+      "step": 3354
+    },
+    {
+      "epoch": 0.513389441469013,
+      "grad_norm": 2.3459506785389475,
+      "learning_rate": 1.742475635684638e-05,
+      "loss": 0.7653,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5135424636572303,
+      "grad_norm": 2.576562934756025,
+      "learning_rate": 1.7423096060782466e-05,
+      "loss": 0.8647,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5136954858454476,
+      "grad_norm": 2.508717422355279,
+      "learning_rate": 1.7421435308837186e-05,
+      "loss": 0.7108,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5138485080336649,
+      "grad_norm": 2.1278984177818536,
+      "learning_rate": 1.7419774101112526e-05,
+      "loss": 0.7988,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5140015302218822,
+      "grad_norm": 2.376063781505229,
+      "learning_rate": 1.741811243771051e-05,
+      "loss": 0.7497,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5141545524100994,
+      "grad_norm": 2.401044243211275,
+      "learning_rate": 1.7416450318733194e-05,
+      "loss": 0.8131,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5143075745983168,
+      "grad_norm": 2.3899515726116958,
+      "learning_rate": 1.741478774428265e-05,
+      "loss": 0.855,
+      "step": 3361
+    },
+    {
+      "epoch": 0.514460596786534,
+      "grad_norm": 2.6320361366457314,
+      "learning_rate": 1.741312471446098e-05,
+      "loss": 0.7491,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5146136189747513,
+      "grad_norm": 2.214272223614381,
+      "learning_rate": 1.7411461229370326e-05,
+      "loss": 0.7406,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5147666411629687,
+      "grad_norm": 2.446861139903529,
+      "learning_rate": 1.7409797289112842e-05,
+      "loss": 0.7961,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5149196633511859,
+      "grad_norm": 2.6934573971392064,
+      "learning_rate": 1.740813289379072e-05,
+      "loss": 0.7586,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5150726855394032,
+      "grad_norm": 2.2123795928387264,
+      "learning_rate": 1.7406468043506176e-05,
+      "loss": 0.7273,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5152257077276206,
+      "grad_norm": 2.243404809270673,
+      "learning_rate": 1.740480273836145e-05,
+      "loss": 0.7537,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5153787299158378,
+      "grad_norm": 2.250245346866966,
+      "learning_rate": 1.7403136978458828e-05,
+      "loss": 0.7529,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5155317521040551,
+      "grad_norm": 2.3461943594132966,
+      "learning_rate": 1.74014707639006e-05,
+      "loss": 0.78,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5156847742922723,
+      "grad_norm": 2.2217893071169224,
+      "learning_rate": 1.7399804094789096e-05,
+      "loss": 0.7165,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5158377964804897,
+      "grad_norm": 2.2631879850970464,
+      "learning_rate": 1.7398136971226677e-05,
+      "loss": 0.7602,
+      "step": 3371
+    },
+    {
+      "epoch": 0.515990818668707,
+      "grad_norm": 2.431585278848265,
+      "learning_rate": 1.739646939331572e-05,
+      "loss": 0.771,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5161438408569242,
+      "grad_norm": 2.4202710096861506,
+      "learning_rate": 1.7394801361158648e-05,
+      "loss": 0.7569,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5162968630451416,
+      "grad_norm": 2.535564085626204,
+      "learning_rate": 1.7393132874857894e-05,
+      "loss": 0.8373,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5164498852333589,
+      "grad_norm": 2.5135595738903467,
+      "learning_rate": 1.739146393451593e-05,
+      "loss": 0.8715,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5166029074215761,
+      "grad_norm": 2.095406374528688,
+      "learning_rate": 1.7389794540235246e-05,
+      "loss": 0.7633,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5167559296097934,
+      "grad_norm": 2.686993076718003,
+      "learning_rate": 1.7388124692118373e-05,
+      "loss": 0.7197,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5169089517980107,
+      "grad_norm": 2.4490117625560517,
+      "learning_rate": 1.7386454390267857e-05,
+      "loss": 0.8004,
+      "step": 3378
+    },
+    {
+      "epoch": 0.517061973986228,
+      "grad_norm": 2.301469614522371,
+      "learning_rate": 1.7384783634786284e-05,
+      "loss": 0.7179,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5172149961744453,
+      "grad_norm": 2.432510545681375,
+      "learning_rate": 1.738311242577626e-05,
+      "loss": 0.8886,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5173680183626626,
+      "grad_norm": 2.4190454792921963,
+      "learning_rate": 1.7381440763340416e-05,
+      "loss": 0.867,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5175210405508799,
+      "grad_norm": 2.5166664142471924,
+      "learning_rate": 1.737976864758142e-05,
+      "loss": 0.8787,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5176740627390972,
+      "grad_norm": 2.4588195664926578,
+      "learning_rate": 1.7378096078601962e-05,
+      "loss": 0.8622,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5178270849273144,
+      "grad_norm": 2.365906374128987,
+      "learning_rate": 1.7376423056504756e-05,
+      "loss": 0.7523,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5179801071155318,
+      "grad_norm": 2.2284179451077035,
+      "learning_rate": 1.7374749581392558e-05,
+      "loss": 0.7016,
+      "step": 3385
+    },
+    {
+      "epoch": 0.518133129303749,
+      "grad_norm": 2.252164274842456,
+      "learning_rate": 1.7373075653368137e-05,
+      "loss": 0.76,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5182861514919663,
+      "grad_norm": 2.3064049524967674,
+      "learning_rate": 1.7371401272534295e-05,
+      "loss": 0.8086,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5184391736801837,
+      "grad_norm": 2.3002293014541673,
+      "learning_rate": 1.7369726438993865e-05,
+      "loss": 0.8244,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5185921958684009,
+      "grad_norm": 2.495992362678814,
+      "learning_rate": 1.7368051152849704e-05,
+      "loss": 0.8959,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5187452180566182,
+      "grad_norm": 2.3947045552206445,
+      "learning_rate": 1.73663754142047e-05,
+      "loss": 0.8003,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5188982402448356,
+      "grad_norm": 2.31492330849426,
+      "learning_rate": 1.736469922316176e-05,
+      "loss": 0.6984,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5190512624330528,
+      "grad_norm": 2.237344088919252,
+      "learning_rate": 1.736302257982383e-05,
+      "loss": 0.8371,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5192042846212701,
+      "grad_norm": 2.7400758010484645,
+      "learning_rate": 1.7361345484293882e-05,
+      "loss": 0.9044,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5193573068094873,
+      "grad_norm": 2.33269965816581,
+      "learning_rate": 1.7359667936674907e-05,
+      "loss": 0.7688,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5195103289977047,
+      "grad_norm": 2.429817087511107,
+      "learning_rate": 1.7357989937069936e-05,
+      "loss": 0.8199,
+      "step": 3395
+    },
+    {
+      "epoch": 0.519663351185922,
+      "grad_norm": 2.3330509991957933,
+      "learning_rate": 1.735631148558202e-05,
+      "loss": 0.7509,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5198163733741392,
+      "grad_norm": 2.116684148235966,
+      "learning_rate": 1.7354632582314232e-05,
+      "loss": 0.7309,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5199693955623566,
+      "grad_norm": 2.4278111061121836,
+      "learning_rate": 1.735295322736969e-05,
+      "loss": 0.8164,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5201224177505739,
+      "grad_norm": 2.2950130329897944,
+      "learning_rate": 1.7351273420851522e-05,
+      "loss": 0.7755,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5202754399387911,
+      "grad_norm": 2.3013900005039796,
+      "learning_rate": 1.7349593162862896e-05,
+      "loss": 0.7645,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5204284621270084,
+      "grad_norm": 2.131498202133757,
+      "learning_rate": 1.7347912453507003e-05,
+      "loss": 0.7436,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5205814843152257,
+      "grad_norm": 2.0513736202538104,
+      "learning_rate": 1.734623129288706e-05,
+      "loss": 0.7005,
+      "step": 3402
+    },
+    {
+      "epoch": 0.520734506503443,
+      "grad_norm": 2.458358694468482,
+      "learning_rate": 1.734454968110632e-05,
+      "loss": 0.7273,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5208875286916603,
+      "grad_norm": 2.492268056512554,
+      "learning_rate": 1.7342867618268043e-05,
+      "loss": 0.7453,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5210405508798776,
+      "grad_norm": 2.2243137346199444,
+      "learning_rate": 1.7341185104475546e-05,
+      "loss": 0.7455,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5211935730680949,
+      "grad_norm": 2.3585772891771684,
+      "learning_rate": 1.733950213983215e-05,
+      "loss": 0.8491,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5213465952563122,
+      "grad_norm": 2.301805032044261,
+      "learning_rate": 1.733781872444122e-05,
+      "loss": 0.8319,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5214996174445294,
+      "grad_norm": 2.2982512317533246,
+      "learning_rate": 1.7336134858406134e-05,
+      "loss": 0.7436,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5216526396327468,
+      "grad_norm": 2.335280353022862,
+      "learning_rate": 1.7334450541830305e-05,
+      "loss": 0.7517,
+      "step": 3409
+    },
+    {
+      "epoch": 0.521805661820964,
+      "grad_norm": 2.5502609428808394,
+      "learning_rate": 1.7332765774817177e-05,
+      "loss": 0.8186,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5219586840091813,
+      "grad_norm": 2.415137367526865,
+      "learning_rate": 1.7331080557470214e-05,
+      "loss": 0.7861,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5221117061973987,
+      "grad_norm": 2.720231789733179,
+      "learning_rate": 1.732939488989292e-05,
+      "loss": 0.8509,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5222647283856159,
+      "grad_norm": 2.2490141250767057,
+      "learning_rate": 1.732770877218881e-05,
+      "loss": 0.6168,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5224177505738332,
+      "grad_norm": 2.255322676967083,
+      "learning_rate": 1.7326022204461438e-05,
+      "loss": 0.8309,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5225707727620506,
+      "grad_norm": 2.403136772714949,
+      "learning_rate": 1.7324335186814382e-05,
+      "loss": 0.7862,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5227237949502678,
+      "grad_norm": 2.4871382173190963,
+      "learning_rate": 1.7322647719351253e-05,
+      "loss": 0.7874,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5228768171384851,
+      "grad_norm": 2.0456564998057143,
+      "learning_rate": 1.732095980217568e-05,
+      "loss": 0.7629,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5230298393267023,
+      "grad_norm": 2.140116669546573,
+      "learning_rate": 1.7319271435391326e-05,
+      "loss": 0.7619,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5231828615149197,
+      "grad_norm": 2.2570260777984443,
+      "learning_rate": 1.7317582619101876e-05,
+      "loss": 0.7535,
+      "step": 3419
+    },
+    {
+      "epoch": 0.523335883703137,
+      "grad_norm": 2.3392895682248445,
+      "learning_rate": 1.7315893353411052e-05,
+      "loss": 0.824,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5234889058913542,
+      "grad_norm": 2.543168315321036,
+      "learning_rate": 1.73142036384226e-05,
+      "loss": 0.7372,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5236419280795716,
+      "grad_norm": 2.1197882622093034,
+      "learning_rate": 1.731251347424029e-05,
+      "loss": 0.772,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5237949502677889,
+      "grad_norm": 2.5217333583866535,
+      "learning_rate": 1.731082286096792e-05,
+      "loss": 0.7934,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5239479724560061,
+      "grad_norm": 2.308111842163604,
+      "learning_rate": 1.7309131798709313e-05,
+      "loss": 0.8,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5241009946442234,
+      "grad_norm": 2.1948172683467866,
+      "learning_rate": 1.7307440287568335e-05,
+      "loss": 0.7586,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5242540168324407,
+      "grad_norm": 2.349110205686944,
+      "learning_rate": 1.730574832764886e-05,
+      "loss": 0.8168,
+      "step": 3426
+    },
+    {
+      "epoch": 0.524407039020658,
+      "grad_norm": 2.478957873778854,
+      "learning_rate": 1.73040559190548e-05,
+      "loss": 0.7624,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5245600612088753,
+      "grad_norm": 2.2363519273765573,
+      "learning_rate": 1.730236306189009e-05,
+      "loss": 0.7556,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5247130833970926,
+      "grad_norm": 2.4000273014466638,
+      "learning_rate": 1.73006697562587e-05,
+      "loss": 0.631,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5248661055853099,
+      "grad_norm": 2.296172924033883,
+      "learning_rate": 1.729897600226462e-05,
+      "loss": 0.6752,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5250191277735272,
+      "grad_norm": 2.4068767186514926,
+      "learning_rate": 1.729728180001187e-05,
+      "loss": 0.768,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5251721499617444,
+      "grad_norm": 2.0931991854527574,
+      "learning_rate": 1.7295587149604496e-05,
+      "loss": 0.7415,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5253251721499618,
+      "grad_norm": 2.2446215315731455,
+      "learning_rate": 1.7293892051146575e-05,
+      "loss": 0.7315,
+      "step": 3433
+    },
+    {
+      "epoch": 0.525478194338179,
+      "grad_norm": 2.31206674583333,
+      "learning_rate": 1.729219650474221e-05,
+      "loss": 0.7596,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5256312165263963,
+      "grad_norm": 2.4780596800982195,
+      "learning_rate": 1.7290500510495532e-05,
+      "loss": 0.8166,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5257842387146137,
+      "grad_norm": 2.2182491100816977,
+      "learning_rate": 1.7288804068510695e-05,
+      "loss": 0.7766,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5259372609028309,
+      "grad_norm": 2.5993499183042323,
+      "learning_rate": 1.7287107178891886e-05,
+      "loss": 0.7479,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5260902830910482,
+      "grad_norm": 2.6164904758032295,
+      "learning_rate": 1.7285409841743318e-05,
+      "loss": 0.8316,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5262433052792654,
+      "grad_norm": 2.1482006073526083,
+      "learning_rate": 1.728371205716923e-05,
+      "loss": 0.7556,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5263963274674828,
+      "grad_norm": 2.2992580395879236,
+      "learning_rate": 1.7282013825273894e-05,
+      "loss": 0.8617,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5265493496557001,
+      "grad_norm": 2.760122409809691,
+      "learning_rate": 1.7280315146161595e-05,
+      "loss": 0.8638,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5267023718439173,
+      "grad_norm": 2.413735791441003,
+      "learning_rate": 1.727861601993667e-05,
+      "loss": 0.7804,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5268553940321347,
+      "grad_norm": 3.149252704804255,
+      "learning_rate": 1.727691644670346e-05,
+      "loss": 0.7932,
+      "step": 3443
+    },
+    {
+      "epoch": 0.527008416220352,
+      "grad_norm": 2.2225478603436235,
+      "learning_rate": 1.727521642656634e-05,
+      "loss": 0.7971,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5271614384085692,
+      "grad_norm": 2.2044358309518524,
+      "learning_rate": 1.7273515959629715e-05,
+      "loss": 0.6773,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5273144605967865,
+      "grad_norm": 2.4651261717517774,
+      "learning_rate": 1.727181504599803e-05,
+      "loss": 0.8615,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5274674827850038,
+      "grad_norm": 2.5505959614561786,
+      "learning_rate": 1.7270113685775728e-05,
+      "loss": 0.8561,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5276205049732211,
+      "grad_norm": 2.2154727272181036,
+      "learning_rate": 1.7268411879067305e-05,
+      "loss": 0.7426,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5277735271614384,
+      "grad_norm": 2.226037165166472,
+      "learning_rate": 1.7266709625977277e-05,
+      "loss": 0.7895,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5279265493496557,
+      "grad_norm": 2.533881833418401,
+      "learning_rate": 1.7265006926610183e-05,
+      "loss": 0.7938,
+      "step": 3450
+    },
+    {
+      "epoch": 0.528079571537873,
+      "grad_norm": 2.4668535803573213,
+      "learning_rate": 1.726330378107059e-05,
+      "loss": 0.7951,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5282325937260903,
+      "grad_norm": 2.213391599044407,
+      "learning_rate": 1.72616001894631e-05,
+      "loss": 0.6765,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5283856159143075,
+      "grad_norm": 2.231713283875013,
+      "learning_rate": 1.7259896151892336e-05,
+      "loss": 0.7737,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5285386381025249,
+      "grad_norm": 2.203632199838888,
+      "learning_rate": 1.7258191668462948e-05,
+      "loss": 0.7826,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5286916602907421,
+      "grad_norm": 2.2665233994578076,
+      "learning_rate": 1.725648673927961e-05,
+      "loss": 0.7484,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5288446824789594,
+      "grad_norm": 2.2420393983606353,
+      "learning_rate": 1.725478136444704e-05,
+      "loss": 0.8268,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5289977046671768,
+      "grad_norm": 2.6467064473764834,
+      "learning_rate": 1.7253075544069964e-05,
+      "loss": 0.8971,
+      "step": 3457
+    },
+    {
+      "epoch": 0.529150726855394,
+      "grad_norm": 2.8179506973708786,
+      "learning_rate": 1.7251369278253146e-05,
+      "loss": 0.7285,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5293037490436113,
+      "grad_norm": 2.0736781024821838,
+      "learning_rate": 1.724966256710137e-05,
+      "loss": 0.7716,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5294567712318287,
+      "grad_norm": 2.3075858700032037,
+      "learning_rate": 1.7247955410719454e-05,
+      "loss": 0.7077,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5296097934200459,
+      "grad_norm": 2.325588149675756,
+      "learning_rate": 1.7246247809212248e-05,
+      "loss": 0.753,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5297628156082632,
+      "grad_norm": 2.3361579258165883,
+      "learning_rate": 1.7244539762684607e-05,
+      "loss": 0.8598,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5299158377964804,
+      "grad_norm": 2.105536512175094,
+      "learning_rate": 1.7242831271241446e-05,
+      "loss": 0.8154,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5300688599846978,
+      "grad_norm": 2.3968996850665794,
+      "learning_rate": 1.7241122334987684e-05,
+      "loss": 0.8799,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5302218821729151,
+      "grad_norm": 2.484649805300681,
+      "learning_rate": 1.7239412954028268e-05,
+      "loss": 0.8878,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5303749043611323,
+      "grad_norm": 2.41188080989505,
+      "learning_rate": 1.723770312846818e-05,
+      "loss": 0.8617,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5305279265493497,
+      "grad_norm": 2.5004806531008734,
+      "learning_rate": 1.7235992858412432e-05,
+      "loss": 0.8951,
+      "step": 3467
+    },
+    {
+      "epoch": 0.530680948737567,
+      "grad_norm": 2.3251879051654063,
+      "learning_rate": 1.7234282143966056e-05,
+      "loss": 0.7946,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5308339709257842,
+      "grad_norm": 2.270779230163958,
+      "learning_rate": 1.723257098523411e-05,
+      "loss": 0.814,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5309869931140015,
+      "grad_norm": 2.318798194460353,
+      "learning_rate": 1.7230859382321686e-05,
+      "loss": 0.8017,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5311400153022188,
+      "grad_norm": 2.3221217836842274,
+      "learning_rate": 1.7229147335333906e-05,
+      "loss": 0.7764,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5312930374904361,
+      "grad_norm": 2.1502170801701377,
+      "learning_rate": 1.72274348443759e-05,
+      "loss": 0.7496,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5314460596786534,
+      "grad_norm": 2.2513830788602442,
+      "learning_rate": 1.722572190955285e-05,
+      "loss": 0.8413,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5315990818668707,
+      "grad_norm": 2.3165428770546392,
+      "learning_rate": 1.722400853096995e-05,
+      "loss": 0.7304,
+      "step": 3474
+    },
+    {
+      "epoch": 0.531752104055088,
+      "grad_norm": 2.022193837277281,
+      "learning_rate": 1.7222294708732423e-05,
+      "loss": 0.8124,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5319051262433053,
+      "grad_norm": 2.240338201650682,
+      "learning_rate": 1.7220580442945527e-05,
+      "loss": 0.7717,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5320581484315225,
+      "grad_norm": 2.279369131283573,
+      "learning_rate": 1.7218865733714538e-05,
+      "loss": 0.8536,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5322111706197399,
+      "grad_norm": 2.684390602188598,
+      "learning_rate": 1.7217150581144767e-05,
+      "loss": 0.8194,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5323641928079571,
+      "grad_norm": 1.9666810860473458,
+      "learning_rate": 1.721543498534154e-05,
+      "loss": 0.7184,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5325172149961744,
+      "grad_norm": 2.1966123033514364,
+      "learning_rate": 1.7213718946410227e-05,
+      "loss": 0.6808,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5326702371843918,
+      "grad_norm": 2.839505940014808,
+      "learning_rate": 1.721200246445621e-05,
+      "loss": 0.7687,
+      "step": 3481
+    },
+    {
+      "epoch": 0.532823259372609,
+      "grad_norm": 2.396567722242438,
+      "learning_rate": 1.7210285539584913e-05,
+      "loss": 0.7691,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5329762815608263,
+      "grad_norm": 2.4196703683338625,
+      "learning_rate": 1.7208568171901768e-05,
+      "loss": 0.7263,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5331293037490437,
+      "grad_norm": 2.3478722383711967,
+      "learning_rate": 1.720685036151226e-05,
+      "loss": 0.7947,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5332823259372609,
+      "grad_norm": 2.463647890976903,
+      "learning_rate": 1.7205132108521868e-05,
+      "loss": 0.776,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5334353481254782,
+      "grad_norm": 2.465606762967954,
+      "learning_rate": 1.7203413413036132e-05,
+      "loss": 0.8351,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5335883703136954,
+      "grad_norm": 2.4476074906053005,
+      "learning_rate": 1.72016942751606e-05,
+      "loss": 0.8652,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5337413925019128,
+      "grad_norm": 2.3371820769390217,
+      "learning_rate": 1.7199974695000844e-05,
+      "loss": 0.7619,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5338944146901301,
+      "grad_norm": 2.016240201556696,
+      "learning_rate": 1.7198254672662482e-05,
+      "loss": 0.7056,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5340474368783473,
+      "grad_norm": 2.1501317642039823,
+      "learning_rate": 1.7196534208251138e-05,
+      "loss": 0.7003,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5342004590665647,
+      "grad_norm": 2.428358228929068,
+      "learning_rate": 1.7194813301872475e-05,
+      "loss": 0.6957,
+      "step": 3491
+    },
+    {
+      "epoch": 0.534353481254782,
+      "grad_norm": 2.3692668417947824,
+      "learning_rate": 1.719309195363218e-05,
+      "loss": 0.833,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5345065034429992,
+      "grad_norm": 2.2518561612504544,
+      "learning_rate": 1.7191370163635968e-05,
+      "loss": 0.8864,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5346595256312165,
+      "grad_norm": 2.1806620122914433,
+      "learning_rate": 1.7189647931989584e-05,
+      "loss": 0.7775,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5348125478194338,
+      "grad_norm": 2.1202465956947107,
+      "learning_rate": 1.7187925258798795e-05,
+      "loss": 0.6904,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5349655700076511,
+      "grad_norm": 2.293147683612584,
+      "learning_rate": 1.7186202144169396e-05,
+      "loss": 0.7109,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5351185921958684,
+      "grad_norm": 2.387009081930552,
+      "learning_rate": 1.7184478588207208e-05,
+      "loss": 0.8928,
+      "step": 3497
+    },
+    {
+      "epoch": 0.5352716143840857,
+      "grad_norm": 2.036587619586427,
+      "learning_rate": 1.7182754591018084e-05,
+      "loss": 0.7308,
+      "step": 3498
+    },
+    {
+      "epoch": 0.535424636572303,
+      "grad_norm": 2.23617394688373,
+      "learning_rate": 1.7181030152707903e-05,
+      "loss": 0.8534,
+      "step": 3499
+    },
+    {
+      "epoch": 0.5355776587605203,
+      "grad_norm": 2.4902832725810464,
+      "learning_rate": 1.7179305273382567e-05,
+      "loss": 0.898,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5357306809487375,
+      "grad_norm": 2.2712921361457936,
+      "learning_rate": 1.717757995314801e-05,
+      "loss": 0.7424,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5358837031369549,
+      "grad_norm": 2.1869051131580823,
+      "learning_rate": 1.717585419211019e-05,
+      "loss": 0.8083,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5360367253251721,
+      "grad_norm": 2.0173761674315616,
+      "learning_rate": 1.7174127990375092e-05,
+      "loss": 0.7098,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5361897475133894,
+      "grad_norm": 2.7480203050102974,
+      "learning_rate": 1.7172401348048726e-05,
+      "loss": 0.8463,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5363427697016068,
+      "grad_norm": 2.2412224884608185,
+      "learning_rate": 1.7170674265237136e-05,
+      "loss": 0.7697,
+      "step": 3505
+    },
+    {
+      "epoch": 0.536495791889824,
+      "grad_norm": 2.317814020727311,
+      "learning_rate": 1.7168946742046388e-05,
+      "loss": 0.7449,
+      "step": 3506
+    },
+    {
+      "epoch": 0.5366488140780413,
+      "grad_norm": 2.60268449435134,
+      "learning_rate": 1.7167218778582575e-05,
+      "loss": 0.7636,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5368018362662587,
+      "grad_norm": 2.4034307976445,
+      "learning_rate": 1.716549037495182e-05,
+      "loss": 0.7584,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5369548584544759,
+      "grad_norm": 2.181778255040697,
+      "learning_rate": 1.7163761531260267e-05,
+      "loss": 0.8163,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5371078806426932,
+      "grad_norm": 2.144458168411783,
+      "learning_rate": 1.7162032247614097e-05,
+      "loss": 0.7615,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5372609028309104,
+      "grad_norm": 2.277975562378708,
+      "learning_rate": 1.7160302524119506e-05,
+      "loss": 0.7225,
+      "step": 3511
+    },
+    {
+      "epoch": 0.5374139250191278,
+      "grad_norm": 2.1812185639721324,
+      "learning_rate": 1.7158572360882723e-05,
+      "loss": 0.7146,
+      "step": 3512
+    },
+    {
+      "epoch": 0.5375669472073451,
+      "grad_norm": 2.487858008956128,
+      "learning_rate": 1.7156841758010015e-05,
+      "loss": 0.8443,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5377199693955623,
+      "grad_norm": 2.451809471203112,
+      "learning_rate": 1.715511071560765e-05,
+      "loss": 0.8188,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5378729915837797,
+      "grad_norm": 2.4906401899132127,
+      "learning_rate": 1.7153379233781947e-05,
+      "loss": 0.7324,
+      "step": 3515
+    },
+    {
+      "epoch": 0.538026013771997,
+      "grad_norm": 2.3799176640369293,
+      "learning_rate": 1.715164731263924e-05,
+      "loss": 0.7876,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5381790359602142,
+      "grad_norm": 2.256999436428998,
+      "learning_rate": 1.7149914952285896e-05,
+      "loss": 0.704,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5383320581484315,
+      "grad_norm": 2.375566540112142,
+      "learning_rate": 1.7148182152828302e-05,
+      "loss": 0.7552,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5384850803366488,
+      "grad_norm": 2.396394674063,
+      "learning_rate": 1.714644891437288e-05,
+      "loss": 0.8285,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5386381025248661,
+      "grad_norm": 2.214449654913485,
+      "learning_rate": 1.7144715237026073e-05,
+      "loss": 0.8153,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5387911247130834,
+      "grad_norm": 2.440366421084005,
+      "learning_rate": 1.714298112089435e-05,
+      "loss": 0.7928,
+      "step": 3521
+    },
+    {
+      "epoch": 0.5389441469013007,
+      "grad_norm": 2.038324486460608,
+      "learning_rate": 1.7141246566084218e-05,
+      "loss": 0.6482,
+      "step": 3522
+    },
+    {
+      "epoch": 0.539097169089518,
+      "grad_norm": 2.718259058349394,
+      "learning_rate": 1.7139511572702195e-05,
+      "loss": 0.8081,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5392501912777353,
+      "grad_norm": 2.5128057012131255,
+      "learning_rate": 1.7137776140854838e-05,
+      "loss": 0.8716,
+      "step": 3524
+    },
+    {
+      "epoch": 0.5394032134659525,
+      "grad_norm": 2.484955535252938,
+      "learning_rate": 1.713604027064872e-05,
+      "loss": 0.8966,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5395562356541699,
+      "grad_norm": 2.329587732262935,
+      "learning_rate": 1.713430396219046e-05,
+      "loss": 0.7159,
+      "step": 3526
+    },
+    {
+      "epoch": 0.5397092578423871,
+      "grad_norm": 2.2577950281919956,
+      "learning_rate": 1.713256721558668e-05,
+      "loss": 0.7494,
+      "step": 3527
+    },
+    {
+      "epoch": 0.5398622800306044,
+      "grad_norm": 2.2125935730221107,
+      "learning_rate": 1.7130830030944042e-05,
+      "loss": 0.743,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5400153022188218,
+      "grad_norm": 2.4576951883990286,
+      "learning_rate": 1.712909240836924e-05,
+      "loss": 0.7244,
+      "step": 3529
+    },
+    {
+      "epoch": 0.540168324407039,
+      "grad_norm": 2.4385213443689078,
+      "learning_rate": 1.7127354347968984e-05,
+      "loss": 0.8586,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5403213465952563,
+      "grad_norm": 2.281079066866731,
+      "learning_rate": 1.712561584985001e-05,
+      "loss": 0.8375,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5404743687834737,
+      "grad_norm": 2.114376916900364,
+      "learning_rate": 1.7123876914119097e-05,
+      "loss": 0.7872,
+      "step": 3532
+    },
+    {
+      "epoch": 0.5406273909716909,
+      "grad_norm": 2.3214355956058057,
+      "learning_rate": 1.7122137540883034e-05,
+      "loss": 0.8094,
+      "step": 3533
+    },
+    {
+      "epoch": 0.5407804131599082,
+      "grad_norm": 2.2774113626957377,
+      "learning_rate": 1.712039773024864e-05,
+      "loss": 0.7912,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5409334353481254,
+      "grad_norm": 2.3514449429395685,
+      "learning_rate": 1.7118657482322768e-05,
+      "loss": 0.8295,
+      "step": 3535
+    },
+    {
+      "epoch": 0.5410864575363428,
+      "grad_norm": 2.388457809727973,
+      "learning_rate": 1.7116916797212288e-05,
+      "loss": 0.8479,
+      "step": 3536
+    },
+    {
+      "epoch": 0.5412394797245601,
+      "grad_norm": 2.4052542394521623,
+      "learning_rate": 1.7115175675024107e-05,
+      "loss": 0.8573,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5413925019127773,
+      "grad_norm": 2.281123645319344,
+      "learning_rate": 1.7113434115865157e-05,
+      "loss": 0.7488,
+      "step": 3538
+    },
+    {
+      "epoch": 0.5415455241009947,
+      "grad_norm": 2.0479313985657197,
+      "learning_rate": 1.7111692119842388e-05,
+      "loss": 0.7098,
+      "step": 3539
+    },
+    {
+      "epoch": 0.541698546289212,
+      "grad_norm": 2.503934028957903,
+      "learning_rate": 1.7109949687062784e-05,
+      "loss": 0.7766,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5418515684774292,
+      "grad_norm": 2.2920623120918076,
+      "learning_rate": 1.7108206817633355e-05,
+      "loss": 0.7058,
+      "step": 3541
+    },
+    {
+      "epoch": 0.5420045906656465,
+      "grad_norm": 2.3190499618303715,
+      "learning_rate": 1.7106463511661136e-05,
+      "loss": 0.8249,
+      "step": 3542
+    },
+    {
+      "epoch": 0.5421576128538638,
+      "grad_norm": 2.3367282852844093,
+      "learning_rate": 1.7104719769253196e-05,
+      "loss": 0.8424,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5423106350420811,
+      "grad_norm": 2.459648815663863,
+      "learning_rate": 1.7102975590516622e-05,
+      "loss": 0.7337,
+      "step": 3544
+    },
+    {
+      "epoch": 0.5424636572302984,
+      "grad_norm": 2.1848741538774243,
+      "learning_rate": 1.710123097555853e-05,
+      "loss": 0.6866,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5426166794185157,
+      "grad_norm": 2.2894374221802813,
+      "learning_rate": 1.709948592448606e-05,
+      "loss": 1.0706,
+      "step": 3546
+    },
+    {
+      "epoch": 0.542769701606733,
+      "grad_norm": 2.1592314348992354,
+      "learning_rate": 1.7097740437406388e-05,
+      "loss": 0.79,
+      "step": 3547
+    },
+    {
+      "epoch": 0.5429227237949502,
+      "grad_norm": 2.4608898527446863,
+      "learning_rate": 1.7095994514426704e-05,
+      "loss": 0.8597,
+      "step": 3548
+    },
+    {
+      "epoch": 0.5430757459831675,
+      "grad_norm": 2.3433110444464886,
+      "learning_rate": 1.7094248155654245e-05,
+      "loss": 0.8365,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5432287681713849,
+      "grad_norm": 2.15572265762958,
+      "learning_rate": 1.709250136119625e-05,
+      "loss": 0.7956,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5433817903596021,
+      "grad_norm": 2.3729171950028247,
+      "learning_rate": 1.7090754131160002e-05,
+      "loss": 0.816,
+      "step": 3551
+    },
+    {
+      "epoch": 0.5435348125478194,
+      "grad_norm": 2.2130874362562074,
+      "learning_rate": 1.7089006465652803e-05,
+      "loss": 0.776,
+      "step": 3552
+    },
+    {
+      "epoch": 0.5436878347360368,
+      "grad_norm": 2.0911755353057626,
+      "learning_rate": 1.7087258364781983e-05,
+      "loss": 0.7587,
+      "step": 3553
+    },
+    {
+      "epoch": 0.543840856924254,
+      "grad_norm": 2.4086231529019995,
+      "learning_rate": 1.7085509828654903e-05,
+      "loss": 0.7506,
+      "step": 3554
+    },
+    {
+      "epoch": 0.5439938791124713,
+      "grad_norm": 2.472931732242185,
+      "learning_rate": 1.7083760857378944e-05,
+      "loss": 0.8069,
+      "step": 3555
+    },
+    {
+      "epoch": 0.5441469013006885,
+      "grad_norm": 2.2070865464878966,
+      "learning_rate": 1.708201145106152e-05,
+      "loss": 0.7785,
+      "step": 3556
+    },
+    {
+      "epoch": 0.5442999234889059,
+      "grad_norm": 2.2713818289159096,
+      "learning_rate": 1.708026160981007e-05,
+      "loss": 0.7081,
+      "step": 3557
+    },
+    {
+      "epoch": 0.5444529456771232,
+      "grad_norm": 2.092372395572879,
+      "learning_rate": 1.7078511333732055e-05,
+      "loss": 0.8032,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5446059678653404,
+      "grad_norm": 2.4299843117798483,
+      "learning_rate": 1.7076760622934968e-05,
+      "loss": 0.8316,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5447589900535578,
+      "grad_norm": 2.3366226179643514,
+      "learning_rate": 1.7075009477526326e-05,
+      "loss": 0.7562,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5449120122417751,
+      "grad_norm": 2.1968180719323662,
+      "learning_rate": 1.7073257897613676e-05,
+      "loss": 0.8084,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5450650344299923,
+      "grad_norm": 2.2329839134922658,
+      "learning_rate": 1.7071505883304583e-05,
+      "loss": 0.7897,
+      "step": 3562
+    },
+    {
+      "epoch": 0.5452180566182097,
+      "grad_norm": 2.1279812197279533,
+      "learning_rate": 1.706975343470665e-05,
+      "loss": 0.7008,
+      "step": 3563
+    },
+    {
+      "epoch": 0.5453710788064269,
+      "grad_norm": 2.457635225867449,
+      "learning_rate": 1.706800055192751e-05,
+      "loss": 0.8133,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5455241009946442,
+      "grad_norm": 2.5452787153874588,
+      "learning_rate": 1.70662472350748e-05,
+      "loss": 0.8601,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5456771231828615,
+      "grad_norm": 2.2487105216791896,
+      "learning_rate": 1.7064493484256203e-05,
+      "loss": 0.7402,
+      "step": 3566
+    },
+    {
+      "epoch": 0.5458301453710788,
+      "grad_norm": 1.9842725323794457,
+      "learning_rate": 1.7062739299579428e-05,
+      "loss": 0.7388,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5459831675592961,
+      "grad_norm": 2.235134268947892,
+      "learning_rate": 1.7060984681152198e-05,
+      "loss": 0.7448,
+      "step": 3568
+    },
+    {
+      "epoch": 0.5461361897475134,
+      "grad_norm": 2.3743457059092252,
+      "learning_rate": 1.705922962908228e-05,
+      "loss": 0.6938,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5462892119357307,
+      "grad_norm": 2.2343627337316208,
+      "learning_rate": 1.7057474143477453e-05,
+      "loss": 0.7406,
+      "step": 3570
+    },
+    {
+      "epoch": 0.546442234123948,
+      "grad_norm": 2.0820331103238745,
+      "learning_rate": 1.7055718224445536e-05,
+      "loss": 0.7538,
+      "step": 3571
+    },
+    {
+      "epoch": 0.5465952563121652,
+      "grad_norm": 2.45795322798732,
+      "learning_rate": 1.705396187209435e-05,
+      "loss": 0.8631,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5467482785003825,
+      "grad_norm": 2.0978156116712694,
+      "learning_rate": 1.7052205086531777e-05,
+      "loss": 0.7448,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5469013006885999,
+      "grad_norm": 2.8515911974508024,
+      "learning_rate": 1.7050447867865703e-05,
+      "loss": 0.7892,
+      "step": 3574
+    },
+    {
+      "epoch": 0.5470543228768171,
+      "grad_norm": 2.4076632991660487,
+      "learning_rate": 1.704869021620404e-05,
+      "loss": 0.8385,
+      "step": 3575
+    },
+    {
+      "epoch": 0.5472073450650344,
+      "grad_norm": 2.0957414053029826,
+      "learning_rate": 1.7046932131654736e-05,
+      "loss": 0.739,
+      "step": 3576
+    },
+    {
+      "epoch": 0.5473603672532518,
+      "grad_norm": 2.5671839888269194,
+      "learning_rate": 1.704517361432576e-05,
+      "loss": 0.8024,
+      "step": 3577
+    },
+    {
+      "epoch": 0.547513389441469,
+      "grad_norm": 2.3047558703626185,
+      "learning_rate": 1.7043414664325117e-05,
+      "loss": 0.7954,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5476664116296863,
+      "grad_norm": 2.3600811561008017,
+      "learning_rate": 1.7041655281760824e-05,
+      "loss": 0.7037,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5478194338179035,
+      "grad_norm": 2.3556484582162147,
+      "learning_rate": 1.703989546674093e-05,
+      "loss": 0.8624,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5479724560061209,
+      "grad_norm": 2.275299767646912,
+      "learning_rate": 1.7038135219373518e-05,
+      "loss": 0.8452,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5481254781943382,
+      "grad_norm": 2.4187029345883224,
+      "learning_rate": 1.7036374539766688e-05,
+      "loss": 0.7792,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5482785003825554,
+      "grad_norm": 2.206175619502996,
+      "learning_rate": 1.703461342802857e-05,
+      "loss": 0.8213,
+      "step": 3583
+    },
+    {
+      "epoch": 0.5484315225707728,
+      "grad_norm": 2.6504692224635553,
+      "learning_rate": 1.7032851884267323e-05,
+      "loss": 0.8608,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5485845447589901,
+      "grad_norm": 4.131748832937652,
+      "learning_rate": 1.703108990859113e-05,
+      "loss": 0.8366,
+      "step": 3585
+    },
+    {
+      "epoch": 0.5487375669472073,
+      "grad_norm": 2.2101905827124373,
+      "learning_rate": 1.70293275011082e-05,
+      "loss": 0.7445,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5488905891354247,
+      "grad_norm": 2.4726466589537517,
+      "learning_rate": 1.7027564661926766e-05,
+      "loss": 0.7431,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5490436113236419,
+      "grad_norm": 2.4049526878760585,
+      "learning_rate": 1.7025801391155097e-05,
+      "loss": 0.6889,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5491966335118592,
+      "grad_norm": 2.364932204717252,
+      "learning_rate": 1.7024037688901477e-05,
+      "loss": 0.7787,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5493496557000765,
+      "grad_norm": 2.1140087888729324,
+      "learning_rate": 1.7022273555274228e-05,
+      "loss": 0.7424,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5495026778882938,
+      "grad_norm": 2.3098634667372884,
+      "learning_rate": 1.7020508990381685e-05,
+      "loss": 0.7408,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5496557000765111,
+      "grad_norm": 2.4846746652950555,
+      "learning_rate": 1.7018743994332222e-05,
+      "loss": 0.9127,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5498087222647284,
+      "grad_norm": 2.475667889661541,
+      "learning_rate": 1.7016978567234234e-05,
+      "loss": 0.855,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5499617444529457,
+      "grad_norm": 2.391009217771575,
+      "learning_rate": 1.7015212709196142e-05,
+      "loss": 0.7626,
+      "step": 3594
+    },
+    {
+      "epoch": 0.550114766641163,
+      "grad_norm": 2.7017637632302116,
+      "learning_rate": 1.7013446420326393e-05,
+      "loss": 0.7821,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5502677888293802,
+      "grad_norm": 2.302889942316845,
+      "learning_rate": 1.701167970073346e-05,
+      "loss": 0.7432,
+      "step": 3596
+    },
+    {
+      "epoch": 0.5504208110175975,
+      "grad_norm": 2.0950840579929983,
+      "learning_rate": 1.700991255052585e-05,
+      "loss": 0.6559,
+      "step": 3597
+    },
+    {
+      "epoch": 0.5505738332058149,
+      "grad_norm": 2.0143471965497652,
+      "learning_rate": 1.7008144969812085e-05,
+      "loss": 0.6854,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5507268553940321,
+      "grad_norm": 2.153960943563205,
+      "learning_rate": 1.7006376958700724e-05,
+      "loss": 0.722,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5508798775822494,
+      "grad_norm": 2.252566815376236,
+      "learning_rate": 1.7004608517300343e-05,
+      "loss": 0.7874,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5510328997704668,
+      "grad_norm": 2.306690084716881,
+      "learning_rate": 1.700283964571955e-05,
+      "loss": 0.6778,
+      "step": 3601
+    },
+    {
+      "epoch": 0.551185921958684,
+      "grad_norm": 2.0446451450569083,
+      "learning_rate": 1.7001070344066982e-05,
+      "loss": 0.729,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5513389441469013,
+      "grad_norm": 2.249420963070831,
+      "learning_rate": 1.6999300612451294e-05,
+      "loss": 0.6717,
+      "step": 3603
+    },
+    {
+      "epoch": 0.5514919663351185,
+      "grad_norm": 2.3610035001536853,
+      "learning_rate": 1.6997530450981175e-05,
+      "loss": 0.8365,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5516449885233359,
+      "grad_norm": 2.301984232252648,
+      "learning_rate": 1.6995759859765332e-05,
+      "loss": 0.828,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5517980107115532,
+      "grad_norm": 2.242175879504416,
+      "learning_rate": 1.6993988838912514e-05,
+      "loss": 0.7475,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5519510328997704,
+      "grad_norm": 2.3666687002372737,
+      "learning_rate": 1.699221738853148e-05,
+      "loss": 0.5838,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5521040550879878,
+      "grad_norm": 2.3339878325157417,
+      "learning_rate": 1.6990445508731023e-05,
+      "loss": 0.8548,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5522570772762051,
+      "grad_norm": 2.2915212295829672,
+      "learning_rate": 1.6988673199619955e-05,
+      "loss": 0.7516,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5524100994644223,
+      "grad_norm": 2.6222980931777418,
+      "learning_rate": 1.698690046130713e-05,
+      "loss": 0.7324,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5525631216526397,
+      "grad_norm": 2.591361000180004,
+      "learning_rate": 1.6985127293901417e-05,
+      "loss": 0.8047,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5527161438408569,
+      "grad_norm": 2.5309834401100066,
+      "learning_rate": 1.698335369751171e-05,
+      "loss": 0.7602,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5528691660290742,
+      "grad_norm": 2.332044775123289,
+      "learning_rate": 1.6981579672246932e-05,
+      "loss": 0.8728,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5530221882172915,
+      "grad_norm": 2.3418254681031527,
+      "learning_rate": 1.6979805218216033e-05,
+      "loss": 0.7037,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5531752104055088,
+      "grad_norm": 2.2946913600869783,
+      "learning_rate": 1.697803033552799e-05,
+      "loss": 0.6968,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5533282325937261,
+      "grad_norm": 2.345597481651466,
+      "learning_rate": 1.697625502429181e-05,
+      "loss": 0.8231,
+      "step": 3616
+    },
+    {
+      "epoch": 0.5534812547819434,
+      "grad_norm": 2.158189225934225,
+      "learning_rate": 1.6974479284616512e-05,
+      "loss": 0.6546,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5536342769701607,
+      "grad_norm": 2.1114929224062187,
+      "learning_rate": 1.697270311661116e-05,
+      "loss": 0.7127,
+      "step": 3618
+    },
+    {
+      "epoch": 0.553787299158378,
+      "grad_norm": 2.1596841291083453,
+      "learning_rate": 1.6970926520384833e-05,
+      "loss": 0.7419,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5539403213465952,
+      "grad_norm": 2.400828622459582,
+      "learning_rate": 1.696914949604664e-05,
+      "loss": 0.7161,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5540933435348125,
+      "grad_norm": 2.323163384739257,
+      "learning_rate": 1.6967372043705707e-05,
+      "loss": 0.8356,
+      "step": 3621
+    },
+    {
+      "epoch": 0.5542463657230299,
+      "grad_norm": 2.4295012870436152,
+      "learning_rate": 1.6965594163471202e-05,
+      "loss": 0.8251,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5543993879112471,
+      "grad_norm": 2.2234601452756864,
+      "learning_rate": 1.6963815855452312e-05,
+      "loss": 0.6826,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5545524100994644,
+      "grad_norm": 1.9956514045453755,
+      "learning_rate": 1.6962037119758247e-05,
+      "loss": 0.7141,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5547054322876818,
+      "grad_norm": 2.3909418043326545,
+      "learning_rate": 1.696025795649825e-05,
+      "loss": 0.8332,
+      "step": 3625
+    },
+    {
+      "epoch": 0.554858454475899,
+      "grad_norm": 2.453103266939776,
+      "learning_rate": 1.6958478365781576e-05,
+      "loss": 0.8726,
+      "step": 3626
+    },
+    {
+      "epoch": 0.5550114766641163,
+      "grad_norm": 2.2746636360271855,
+      "learning_rate": 1.695669834771753e-05,
+      "loss": 0.8042,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5551644988523335,
+      "grad_norm": 2.2324321290351796,
+      "learning_rate": 1.6954917902415423e-05,
+      "loss": 0.7195,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5553175210405509,
+      "grad_norm": 2.2940189638814728,
+      "learning_rate": 1.6953137029984597e-05,
+      "loss": 0.7609,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5554705432287682,
+      "grad_norm": 2.3775879354384677,
+      "learning_rate": 1.6951355730534426e-05,
+      "loss": 0.6981,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5556235654169854,
+      "grad_norm": 2.201686255606368,
+      "learning_rate": 1.6949574004174304e-05,
+      "loss": 0.6593,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5557765876052028,
+      "grad_norm": 2.1482634948848554,
+      "learning_rate": 1.694779185101366e-05,
+      "loss": 0.7824,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5559296097934201,
+      "grad_norm": 2.173950406005488,
+      "learning_rate": 1.6946009271161936e-05,
+      "loss": 0.5722,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5560826319816373,
+      "grad_norm": 2.1168563259507946,
+      "learning_rate": 1.6944226264728612e-05,
+      "loss": 0.703,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5562356541698547,
+      "grad_norm": 2.267913969094195,
+      "learning_rate": 1.6942442831823184e-05,
+      "loss": 0.8432,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5563886763580719,
+      "grad_norm": 2.6212417618753983,
+      "learning_rate": 1.6940658972555184e-05,
+      "loss": 0.8025,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5565416985462892,
+      "grad_norm": 2.5846678141954467,
+      "learning_rate": 1.6938874687034163e-05,
+      "loss": 0.8314,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5566947207345065,
+      "grad_norm": 2.334021760479528,
+      "learning_rate": 1.6937089975369705e-05,
+      "loss": 0.6797,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5568477429227238,
+      "grad_norm": 2.107816213370517,
+      "learning_rate": 1.693530483767141e-05,
+      "loss": 0.6848,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5570007651109411,
+      "grad_norm": 2.356072011721238,
+      "learning_rate": 1.693351927404892e-05,
+      "loss": 0.7297,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5571537872991584,
+      "grad_norm": 2.169257056229893,
+      "learning_rate": 1.6931733284611882e-05,
+      "loss": 0.8036,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5573068094873757,
+      "grad_norm": 2.2003860483288302,
+      "learning_rate": 1.6929946869469987e-05,
+      "loss": 0.6347,
+      "step": 3642
+    },
+    {
+      "epoch": 0.557459831675593,
+      "grad_norm": 2.170921678736145,
+      "learning_rate": 1.6928160028732942e-05,
+      "loss": 0.6982,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5576128538638102,
+      "grad_norm": 2.29010650862317,
+      "learning_rate": 1.6926372762510492e-05,
+      "loss": 0.8291,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5577658760520275,
+      "grad_norm": 2.4638778340052565,
+      "learning_rate": 1.692458507091239e-05,
+      "loss": 0.7613,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5579188982402449,
+      "grad_norm": 2.054167992139388,
+      "learning_rate": 1.6922796954048434e-05,
+      "loss": 0.8133,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5580719204284621,
+      "grad_norm": 2.7024998915374643,
+      "learning_rate": 1.6921008412028435e-05,
+      "loss": 0.8554,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5582249426166794,
+      "grad_norm": 2.797028845762761,
+      "learning_rate": 1.691921944496223e-05,
+      "loss": 0.8489,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5583779648048968,
+      "grad_norm": 2.1668006488701166,
+      "learning_rate": 1.6917430052959692e-05,
+      "loss": 0.7914,
+      "step": 3649
+    },
+    {
+      "epoch": 0.558530986993114,
+      "grad_norm": 2.17202935119517,
+      "learning_rate": 1.6915640236130716e-05,
+      "loss": 0.6803,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5586840091813313,
+      "grad_norm": 2.2835949452400257,
+      "learning_rate": 1.6913849994585217e-05,
+      "loss": 0.7817,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5588370313695485,
+      "grad_norm": 2.3830929379834656,
+      "learning_rate": 1.6912059328433144e-05,
+      "loss": 0.7796,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5589900535577659,
+      "grad_norm": 2.3925932080902808,
+      "learning_rate": 1.6910268237784465e-05,
+      "loss": 0.8567,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5591430757459832,
+      "grad_norm": 2.0860786207614894,
+      "learning_rate": 1.6908476722749182e-05,
+      "loss": 0.7333,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5592960979342004,
+      "grad_norm": 2.307073599086852,
+      "learning_rate": 1.6906684783437315e-05,
+      "loss": 0.8835,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5594491201224178,
+      "grad_norm": 2.552393727596661,
+      "learning_rate": 1.6904892419958918e-05,
+      "loss": 0.832,
+      "step": 3656
+    },
+    {
+      "epoch": 0.559602142310635,
+      "grad_norm": 2.1716406728706112,
+      "learning_rate": 1.6903099632424067e-05,
+      "loss": 0.655,
+      "step": 3657
+    },
+    {
+      "epoch": 0.5597551644988523,
+      "grad_norm": 2.431452976880131,
+      "learning_rate": 1.6901306420942862e-05,
+      "loss": 0.812,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5599081866870697,
+      "grad_norm": 2.2599738041832333,
+      "learning_rate": 1.689951278562543e-05,
+      "loss": 0.7497,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5600612088752869,
+      "grad_norm": 2.3432851680364095,
+      "learning_rate": 1.6897718726581926e-05,
+      "loss": 0.7247,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5602142310635042,
+      "grad_norm": 2.385858830011283,
+      "learning_rate": 1.6895924243922535e-05,
+      "loss": 0.8256,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5603672532517215,
+      "grad_norm": 2.2732882617133487,
+      "learning_rate": 1.6894129337757458e-05,
+      "loss": 0.8056,
+      "step": 3662
+    },
+    {
+      "epoch": 0.5605202754399388,
+      "grad_norm": 2.3518883584912156,
+      "learning_rate": 1.6892334008196925e-05,
+      "loss": 0.7344,
+      "step": 3663
+    },
+    {
+      "epoch": 0.5606732976281561,
+      "grad_norm": 2.3340522250983273,
+      "learning_rate": 1.68905382553512e-05,
+      "loss": 0.8086,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5608263198163733,
+      "grad_norm": 2.256805069227207,
+      "learning_rate": 1.6888742079330566e-05,
+      "loss": 0.7789,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5609793420045907,
+      "grad_norm": 2.3041015030287415,
+      "learning_rate": 1.688694548024533e-05,
+      "loss": 0.749,
+      "step": 3666
+    },
+    {
+      "epoch": 0.561132364192808,
+      "grad_norm": 2.601621539475915,
+      "learning_rate": 1.688514845820583e-05,
+      "loss": 0.8205,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5612853863810252,
+      "grad_norm": 2.4104490877496945,
+      "learning_rate": 1.6883351013322427e-05,
+      "loss": 0.9014,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5614384085692425,
+      "grad_norm": 2.1468517678220076,
+      "learning_rate": 1.6881553145705512e-05,
+      "loss": 0.7408,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5615914307574599,
+      "grad_norm": 2.0830615950634783,
+      "learning_rate": 1.68797548554655e-05,
+      "loss": 0.6995,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5617444529456771,
+      "grad_norm": 2.6874796790277204,
+      "learning_rate": 1.6877956142712827e-05,
+      "loss": 0.7714,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5618974751338944,
+      "grad_norm": 2.4850035657522955,
+      "learning_rate": 1.6876157007557958e-05,
+      "loss": 0.9281,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5620504973221117,
+      "grad_norm": 2.273469641065428,
+      "learning_rate": 1.6874357450111392e-05,
+      "loss": 0.7668,
+      "step": 3673
+    },
+    {
+      "epoch": 0.562203519510329,
+      "grad_norm": 2.2483399413604674,
+      "learning_rate": 1.6872557470483638e-05,
+      "loss": 0.689,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5623565416985463,
+      "grad_norm": 2.1512441247663845,
+      "learning_rate": 1.687075706878525e-05,
+      "loss": 0.7961,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5625095638867635,
+      "grad_norm": 2.348647850963959,
+      "learning_rate": 1.6868956245126785e-05,
+      "loss": 0.8583,
+      "step": 3676
+    },
+    {
+      "epoch": 0.5626625860749809,
+      "grad_norm": 2.4585945763130486,
+      "learning_rate": 1.686715499961885e-05,
+      "loss": 0.7551,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5628156082631982,
+      "grad_norm": 2.227822096524331,
+      "learning_rate": 1.686535333237206e-05,
+      "loss": 0.7732,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5629686304514154,
+      "grad_norm": 2.241153131196082,
+      "learning_rate": 1.6863551243497064e-05,
+      "loss": 0.7789,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5631216526396328,
+      "grad_norm": 2.493229819494121,
+      "learning_rate": 1.686174873310454e-05,
+      "loss": 0.7737,
+      "step": 3680
+    },
+    {
+      "epoch": 0.56327467482785,
+      "grad_norm": 2.204665508561256,
+      "learning_rate": 1.6859945801305176e-05,
+      "loss": 0.7376,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5634276970160673,
+      "grad_norm": 2.46220972958079,
+      "learning_rate": 1.685814244820971e-05,
+      "loss": 0.7946,
+      "step": 3682
+    },
+    {
+      "epoch": 0.5635807192042847,
+      "grad_norm": 2.145178057791112,
+      "learning_rate": 1.6856338673928883e-05,
+      "loss": 0.686,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5637337413925019,
+      "grad_norm": 2.4499506242368474,
+      "learning_rate": 1.6854534478573475e-05,
+      "loss": 0.842,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5638867635807192,
+      "grad_norm": 2.056117094582968,
+      "learning_rate": 1.6852729862254297e-05,
+      "loss": 0.7311,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5640397857689365,
+      "grad_norm": 2.6838442750214786,
+      "learning_rate": 1.6850924825082164e-05,
+      "loss": 0.9012,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5641928079571538,
+      "grad_norm": 2.3160834319934858,
+      "learning_rate": 1.6849119367167935e-05,
+      "loss": 0.758,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5643458301453711,
+      "grad_norm": 2.3640101619510423,
+      "learning_rate": 1.6847313488622495e-05,
+      "loss": 0.7606,
+      "step": 3688
+    },
+    {
+      "epoch": 0.5644988523335883,
+      "grad_norm": 2.121683684858718,
+      "learning_rate": 1.684550718955675e-05,
+      "loss": 0.6647,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5646518745218057,
+      "grad_norm": 2.2062012781293823,
+      "learning_rate": 1.6843700470081625e-05,
+      "loss": 0.8113,
+      "step": 3690
+    },
+    {
+      "epoch": 0.564804896710023,
+      "grad_norm": 2.3360486325838505,
+      "learning_rate": 1.684189333030808e-05,
+      "loss": 0.6079,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5649579188982402,
+      "grad_norm": 2.197760459122702,
+      "learning_rate": 1.68400857703471e-05,
+      "loss": 0.6825,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5651109410864575,
+      "grad_norm": 2.1093971294742326,
+      "learning_rate": 1.6838277790309697e-05,
+      "loss": 0.7355,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5652639632746749,
+      "grad_norm": 2.3687841166774524,
+      "learning_rate": 1.6836469390306903e-05,
+      "loss": 0.7743,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5654169854628921,
+      "grad_norm": 3.161887730066366,
+      "learning_rate": 1.683466057044978e-05,
+      "loss": 0.827,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5655700076511094,
+      "grad_norm": 2.216397568711454,
+      "learning_rate": 1.683285133084941e-05,
+      "loss": 0.8549,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5657230298393267,
+      "grad_norm": 1.9890486097667865,
+      "learning_rate": 1.6831041671616914e-05,
+      "loss": 0.7672,
+      "step": 3697
+    },
+    {
+      "epoch": 0.565876052027544,
+      "grad_norm": 2.493183723503674,
+      "learning_rate": 1.6829231592863425e-05,
+      "loss": 0.8644,
+      "step": 3698
+    },
+    {
+      "epoch": 0.5660290742157613,
+      "grad_norm": 2.2253445860436685,
+      "learning_rate": 1.6827421094700108e-05,
+      "loss": 0.7744,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5661820964039785,
+      "grad_norm": 2.5572294147324848,
+      "learning_rate": 1.682561017723815e-05,
+      "loss": 0.7337,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5663351185921959,
+      "grad_norm": 2.3673349447586554,
+      "learning_rate": 1.682379884058877e-05,
+      "loss": 0.7592,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5664881407804132,
+      "grad_norm": 2.231170609581366,
+      "learning_rate": 1.6821987084863208e-05,
+      "loss": 0.7259,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5666411629686304,
+      "grad_norm": 2.312622201395522,
+      "learning_rate": 1.6820174910172732e-05,
+      "loss": 0.7672,
+      "step": 3703
+    },
+    {
+      "epoch": 0.5667941851568478,
+      "grad_norm": 2.5794283462785215,
+      "learning_rate": 1.6818362316628635e-05,
+      "loss": 0.7342,
+      "step": 3704
+    },
+    {
+      "epoch": 0.566947207345065,
+      "grad_norm": 2.2887391825650525,
+      "learning_rate": 1.6816549304342233e-05,
+      "loss": 0.7277,
+      "step": 3705
+    },
+    {
+      "epoch": 0.5671002295332823,
+      "grad_norm": 2.3369358739869197,
+      "learning_rate": 1.6814735873424874e-05,
+      "loss": 0.7064,
+      "step": 3706
+    },
+    {
+      "epoch": 0.5672532517214997,
+      "grad_norm": 2.1653393567600103,
+      "learning_rate": 1.6812922023987922e-05,
+      "loss": 0.837,
+      "step": 3707
+    },
+    {
+      "epoch": 0.5674062739097169,
+      "grad_norm": 2.3845278093061384,
+      "learning_rate": 1.6811107756142776e-05,
+      "loss": 0.8588,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5675592960979342,
+      "grad_norm": 2.495972598753815,
+      "learning_rate": 1.6809293070000862e-05,
+      "loss": 0.7833,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5677123182861515,
+      "grad_norm": 2.247267803617609,
+      "learning_rate": 1.680747796567362e-05,
+      "loss": 0.9386,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5678653404743688,
+      "grad_norm": 2.5184558160832107,
+      "learning_rate": 1.6805662443272525e-05,
+      "loss": 0.8083,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5680183626625861,
+      "grad_norm": 2.165389680357376,
+      "learning_rate": 1.6803846502909074e-05,
+      "loss": 0.7876,
+      "step": 3712
+    },
+    {
+      "epoch": 0.5681713848508033,
+      "grad_norm": 2.370201233814579,
+      "learning_rate": 1.6802030144694798e-05,
+      "loss": 0.7525,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5683244070390207,
+      "grad_norm": 2.4514170401052175,
+      "learning_rate": 1.6800213368741236e-05,
+      "loss": 0.816,
+      "step": 3714
+    },
+    {
+      "epoch": 0.568477429227238,
+      "grad_norm": 2.2577834479368963,
+      "learning_rate": 1.679839617515997e-05,
+      "loss": 0.7959,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5686304514154552,
+      "grad_norm": 2.5236928956082605,
+      "learning_rate": 1.67965785640626e-05,
+      "loss": 0.7833,
+      "step": 3716
+    },
+    {
+      "epoch": 0.5687834736036725,
+      "grad_norm": 2.188326961484387,
+      "learning_rate": 1.679476053556075e-05,
+      "loss": 0.7064,
+      "step": 3717
+    },
+    {
+      "epoch": 0.5689364957918899,
+      "grad_norm": 2.3822901414350306,
+      "learning_rate": 1.6792942089766076e-05,
+      "loss": 0.8851,
+      "step": 3718
+    },
+    {
+      "epoch": 0.5690895179801071,
+      "grad_norm": 2.1457955667343134,
+      "learning_rate": 1.6791123226790255e-05,
+      "loss": 0.7607,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5692425401683244,
+      "grad_norm": 2.323996375369015,
+      "learning_rate": 1.6789303946744985e-05,
+      "loss": 0.7703,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5693955623565417,
+      "grad_norm": 2.3656552591801523,
+      "learning_rate": 1.6787484249742004e-05,
+      "loss": 0.768,
+      "step": 3721
+    },
+    {
+      "epoch": 0.569548584544759,
+      "grad_norm": 2.3223483884094143,
+      "learning_rate": 1.678566413589306e-05,
+      "loss": 0.7593,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5697016067329763,
+      "grad_norm": 2.2639870672318296,
+      "learning_rate": 1.678384360530994e-05,
+      "loss": 0.733,
+      "step": 3723
+    },
+    {
+      "epoch": 0.5698546289211935,
+      "grad_norm": 2.3355673048906254,
+      "learning_rate": 1.6782022658104444e-05,
+      "loss": 0.7424,
+      "step": 3724
+    },
+    {
+      "epoch": 0.5700076511094109,
+      "grad_norm": 2.218856200972419,
+      "learning_rate": 1.67802012943884e-05,
+      "loss": 0.769,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5701606732976282,
+      "grad_norm": 2.6124220750851506,
+      "learning_rate": 1.6778379514273677e-05,
+      "loss": 0.8804,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5703136954858454,
+      "grad_norm": 2.3903567269634087,
+      "learning_rate": 1.6776557317872146e-05,
+      "loss": 0.7813,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5704667176740628,
+      "grad_norm": 2.5179833679939247,
+      "learning_rate": 1.6774734705295723e-05,
+      "loss": 0.7096,
+      "step": 3728
+    },
+    {
+      "epoch": 0.57061973986228,
+      "grad_norm": 2.1922598523160928,
+      "learning_rate": 1.677291167665634e-05,
+      "loss": 0.7195,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5707727620504973,
+      "grad_norm": 2.4942133453405333,
+      "learning_rate": 1.677108823206595e-05,
+      "loss": 0.7428,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5709257842387147,
+      "grad_norm": 2.2384937500275104,
+      "learning_rate": 1.6769264371636546e-05,
+      "loss": 0.8616,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5710788064269319,
+      "grad_norm": 2.2191384349159002,
+      "learning_rate": 1.6767440095480136e-05,
+      "loss": 0.7403,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5712318286151492,
+      "grad_norm": 2.380505958844727,
+      "learning_rate": 1.6765615403708756e-05,
+      "loss": 0.748,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5713848508033665,
+      "grad_norm": 2.2650779038130793,
+      "learning_rate": 1.6763790296434463e-05,
+      "loss": 0.7306,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5715378729915838,
+      "grad_norm": 2.6913536745347866,
+      "learning_rate": 1.6761964773769352e-05,
+      "loss": 0.8774,
+      "step": 3735
+    },
+    {
+      "epoch": 0.5716908951798011,
+      "grad_norm": 2.079473361927895,
+      "learning_rate": 1.676013883582553e-05,
+      "loss": 0.6435,
+      "step": 3736
+    },
+    {
+      "epoch": 0.5718439173680183,
+      "grad_norm": 2.236787333970397,
+      "learning_rate": 1.6758312482715137e-05,
+      "loss": 0.8045,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5719969395562357,
+      "grad_norm": 2.1951274280494206,
+      "learning_rate": 1.6756485714550333e-05,
+      "loss": 0.7215,
+      "step": 3738
+    },
+    {
+      "epoch": 0.572149961744453,
+      "grad_norm": 2.2622052474326733,
+      "learning_rate": 1.6754658531443314e-05,
+      "loss": 0.703,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5723029839326702,
+      "grad_norm": 2.3735375982121707,
+      "learning_rate": 1.6752830933506285e-05,
+      "loss": 0.7512,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5724560061208875,
+      "grad_norm": 2.0811855879934225,
+      "learning_rate": 1.6751002920851494e-05,
+      "loss": 0.6298,
+      "step": 3741
+    },
+    {
+      "epoch": 0.5726090283091049,
+      "grad_norm": 2.5859503146517446,
+      "learning_rate": 1.6749174493591205e-05,
+      "loss": 0.7602,
+      "step": 3742
+    },
+    {
+      "epoch": 0.5727620504973221,
+      "grad_norm": 2.1294854580052487,
+      "learning_rate": 1.6747345651837706e-05,
+      "loss": 0.799,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5729150726855394,
+      "grad_norm": 2.2886074445311517,
+      "learning_rate": 1.6745516395703315e-05,
+      "loss": 0.8139,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5730680948737567,
+      "grad_norm": 2.6893542581228806,
+      "learning_rate": 1.6743686725300373e-05,
+      "loss": 0.7434,
+      "step": 3745
+    },
+    {
+      "epoch": 0.573221117061974,
+      "grad_norm": 2.3216860488738424,
+      "learning_rate": 1.6741856640741247e-05,
+      "loss": 0.6948,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5733741392501913,
+      "grad_norm": 2.2001897317732615,
+      "learning_rate": 1.6740026142138332e-05,
+      "loss": 0.6766,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5735271614384085,
+      "grad_norm": 2.2336227878098693,
+      "learning_rate": 1.6738195229604043e-05,
+      "loss": 0.8209,
+      "step": 3748
+    },
+    {
+      "epoch": 0.5736801836266259,
+      "grad_norm": 2.240603429890706,
+      "learning_rate": 1.673636390325083e-05,
+      "loss": 0.7174,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5738332058148432,
+      "grad_norm": 2.3651344356855035,
+      "learning_rate": 1.673453216319115e-05,
+      "loss": 0.791,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5739862280030604,
+      "grad_norm": 2.557825145409085,
+      "learning_rate": 1.6732700009537506e-05,
+      "loss": 0.8281,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5741392501912778,
+      "grad_norm": 2.2939591266531587,
+      "learning_rate": 1.6730867442402417e-05,
+      "loss": 0.7533,
+      "step": 3752
+    },
+    {
+      "epoch": 0.574292272379495,
+      "grad_norm": 2.4369272889263467,
+      "learning_rate": 1.6729034461898428e-05,
+      "loss": 0.7564,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5744452945677123,
+      "grad_norm": 2.383742765723668,
+      "learning_rate": 1.6727201068138106e-05,
+      "loss": 0.8455,
+      "step": 3754
+    },
+    {
+      "epoch": 0.5745983167559297,
+      "grad_norm": 2.1235204798335325,
+      "learning_rate": 1.6725367261234052e-05,
+      "loss": 0.7219,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5747513389441469,
+      "grad_norm": 2.0759737406177035,
+      "learning_rate": 1.6723533041298884e-05,
+      "loss": 0.7086,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5749043611323642,
+      "grad_norm": 2.1343424495745587,
+      "learning_rate": 1.6721698408445248e-05,
+      "loss": 0.8257,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5750573833205815,
+      "grad_norm": 2.0830999607795135,
+      "learning_rate": 1.671986336278582e-05,
+      "loss": 0.7539,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5752104055087988,
+      "grad_norm": 2.3324609080288066,
+      "learning_rate": 1.6718027904433292e-05,
+      "loss": 0.8393,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5753634276970161,
+      "grad_norm": 2.9857624065747403,
+      "learning_rate": 1.671619203350039e-05,
+      "loss": 0.8469,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5755164498852333,
+      "grad_norm": 2.3340876823270857,
+      "learning_rate": 1.6714355750099863e-05,
+      "loss": 0.7146,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5756694720734506,
+      "grad_norm": 2.296826476694342,
+      "learning_rate": 1.6712519054344477e-05,
+      "loss": 0.8085,
+      "step": 3762
+    },
+    {
+      "epoch": 0.575822494261668,
+      "grad_norm": 2.398681606627602,
+      "learning_rate": 1.6710681946347046e-05,
+      "loss": 0.8752,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5759755164498852,
+      "grad_norm": 2.1659581982539042,
+      "learning_rate": 1.6708844426220378e-05,
+      "loss": 0.7433,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5761285386381025,
+      "grad_norm": 2.2458337659004317,
+      "learning_rate": 1.6707006494077328e-05,
+      "loss": 0.7098,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5762815608263198,
+      "grad_norm": 2.2737668614456337,
+      "learning_rate": 1.670516815003077e-05,
+      "loss": 0.7792,
+      "step": 3766
+    },
+    {
+      "epoch": 0.5764345830145371,
+      "grad_norm": 2.305720566758425,
+      "learning_rate": 1.6703329394193612e-05,
+      "loss": 0.6744,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5765876052027544,
+      "grad_norm": 2.557917734239461,
+      "learning_rate": 1.6701490226678768e-05,
+      "loss": 0.6602,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5767406273909716,
+      "grad_norm": 2.22313464861298,
+      "learning_rate": 1.6699650647599194e-05,
+      "loss": 0.7318,
+      "step": 3769
+    },
+    {
+      "epoch": 0.576893649579189,
+      "grad_norm": 2.3570935722006143,
+      "learning_rate": 1.669781065706786e-05,
+      "loss": 0.6865,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5770466717674063,
+      "grad_norm": 2.328764118860341,
+      "learning_rate": 1.6695970255197775e-05,
+      "loss": 0.8174,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5771996939556235,
+      "grad_norm": 2.206086498235264,
+      "learning_rate": 1.669412944210196e-05,
+      "loss": 0.7195,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5773527161438409,
+      "grad_norm": 2.4574715151597206,
+      "learning_rate": 1.6692288217893473e-05,
+      "loss": 0.6601,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5775057383320581,
+      "grad_norm": 2.5554333693809337,
+      "learning_rate": 1.6690446582685384e-05,
+      "loss": 0.9636,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5776587605202754,
+      "grad_norm": 2.0830553360852653,
+      "learning_rate": 1.6688604536590797e-05,
+      "loss": 0.688,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5778117827084928,
+      "grad_norm": 2.1131326036234825,
+      "learning_rate": 1.668676207972284e-05,
+      "loss": 0.7221,
+      "step": 3776
+    },
+    {
+      "epoch": 0.57796480489671,
+      "grad_norm": 2.055677955743591,
+      "learning_rate": 1.6684919212194664e-05,
+      "loss": 0.6651,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5781178270849273,
+      "grad_norm": 2.2701708775009135,
+      "learning_rate": 1.6683075934119448e-05,
+      "loss": 0.7376,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5782708492731446,
+      "grad_norm": 2.0336407250984827,
+      "learning_rate": 1.6681232245610393e-05,
+      "loss": 0.76,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5784238714613619,
+      "grad_norm": 2.1902709950009633,
+      "learning_rate": 1.6679388146780732e-05,
+      "loss": 0.7113,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5785768936495792,
+      "grad_norm": 2.283177953498535,
+      "learning_rate": 1.667754363774371e-05,
+      "loss": 0.8212,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5787299158377964,
+      "grad_norm": 2.5358371176198005,
+      "learning_rate": 1.6675698718612613e-05,
+      "loss": 0.7745,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5788829380260138,
+      "grad_norm": 2.4035073241578404,
+      "learning_rate": 1.6673853389500746e-05,
+      "loss": 0.7967,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5790359602142311,
+      "grad_norm": 1.9008232222332522,
+      "learning_rate": 1.667200765052143e-05,
+      "loss": 0.5186,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5791889824024483,
+      "grad_norm": 2.660572345490886,
+      "learning_rate": 1.6670161501788025e-05,
+      "loss": 0.7472,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5793420045906656,
+      "grad_norm": 2.3119643292610608,
+      "learning_rate": 1.666831494341391e-05,
+      "loss": 0.7326,
+      "step": 3786
+    },
+    {
+      "epoch": 0.579495026778883,
+      "grad_norm": 2.2646005923282346,
+      "learning_rate": 1.666646797551248e-05,
+      "loss": 0.7679,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5796480489671002,
+      "grad_norm": 2.158895112775971,
+      "learning_rate": 1.666462059819718e-05,
+      "loss": 0.7638,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5798010711553175,
+      "grad_norm": 2.155066404690859,
+      "learning_rate": 1.6662772811581453e-05,
+      "loss": 0.6579,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5799540933435348,
+      "grad_norm": 2.524080120158428,
+      "learning_rate": 1.6660924615778784e-05,
+      "loss": 0.7391,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5801071155317521,
+      "grad_norm": 2.3536130897830856,
+      "learning_rate": 1.6659076010902678e-05,
+      "loss": 0.7922,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5802601377199694,
+      "grad_norm": 2.139282136745696,
+      "learning_rate": 1.665722699706666e-05,
+      "loss": 0.7086,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5804131599081866,
+      "grad_norm": 2.0993206873081407,
+      "learning_rate": 1.6655377574384294e-05,
+      "loss": 0.6831,
+      "step": 3793
+    },
+    {
+      "epoch": 0.580566182096404,
+      "grad_norm": 2.0100115117297492,
+      "learning_rate": 1.6653527742969152e-05,
+      "loss": 0.7001,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5807192042846213,
+      "grad_norm": 2.3782783324435783,
+      "learning_rate": 1.665167750293484e-05,
+      "loss": 0.7298,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5808722264728385,
+      "grad_norm": 2.040780849933868,
+      "learning_rate": 1.6649826854394997e-05,
+      "loss": 0.7327,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5810252486610559,
+      "grad_norm": 2.5334375804404625,
+      "learning_rate": 1.664797579746327e-05,
+      "loss": 0.6432,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5811782708492731,
+      "grad_norm": 2.2545821441478933,
+      "learning_rate": 1.664612433225334e-05,
+      "loss": 0.7446,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5813312930374904,
+      "grad_norm": 2.8266389872037845,
+      "learning_rate": 1.664427245887892e-05,
+      "loss": 0.8615,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5814843152257078,
+      "grad_norm": 2.357933150397747,
+      "learning_rate": 1.6642420177453728e-05,
+      "loss": 0.7502,
+      "step": 3800
+    },
+    {
+      "epoch": 0.581637337413925,
+      "grad_norm": 2.4860458336444364,
+      "learning_rate": 1.6640567488091536e-05,
+      "loss": 0.7914,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5817903596021423,
+      "grad_norm": 2.5167304694736052,
+      "learning_rate": 1.663871439090611e-05,
+      "loss": 0.7699,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5819433817903596,
+      "grad_norm": 2.336138556851833,
+      "learning_rate": 1.6636860886011266e-05,
+      "loss": 0.7217,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5820964039785769,
+      "grad_norm": 2.3654993443453485,
+      "learning_rate": 1.663500697352083e-05,
+      "loss": 0.738,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5822494261667942,
+      "grad_norm": 2.4112699878549546,
+      "learning_rate": 1.663315265354866e-05,
+      "loss": 0.7535,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5824024483550114,
+      "grad_norm": 2.381433431863148,
+      "learning_rate": 1.6631297926208637e-05,
+      "loss": 0.8315,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5825554705432288,
+      "grad_norm": 2.540843627525302,
+      "learning_rate": 1.6629442791614667e-05,
+      "loss": 0.8621,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5827084927314461,
+      "grad_norm": 2.301553310120931,
+      "learning_rate": 1.6627587249880677e-05,
+      "loss": 0.721,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5828615149196633,
+      "grad_norm": 2.3190169251043353,
+      "learning_rate": 1.662573130112063e-05,
+      "loss": 0.6618,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5830145371078806,
+      "grad_norm": 2.3010873402470033,
+      "learning_rate": 1.6623874945448503e-05,
+      "loss": 0.8682,
+      "step": 3810
+    },
+    {
+      "epoch": 0.583167559296098,
+      "grad_norm": 2.1375246030186035,
+      "learning_rate": 1.6622018182978304e-05,
+      "loss": 0.7486,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5833205814843152,
+      "grad_norm": 2.293568821408462,
+      "learning_rate": 1.662016101382406e-05,
+      "loss": 0.7397,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5834736036725325,
+      "grad_norm": 2.417223850861851,
+      "learning_rate": 1.6618303438099834e-05,
+      "loss": 0.7969,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5836266258607498,
+      "grad_norm": 2.419752913361654,
+      "learning_rate": 1.6616445455919704e-05,
+      "loss": 0.7874,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5837796480489671,
+      "grad_norm": 2.297683203684462,
+      "learning_rate": 1.6614587067397767e-05,
+      "loss": 0.7571,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5839326702371844,
+      "grad_norm": 2.1676060389805043,
+      "learning_rate": 1.661272827264817e-05,
+      "loss": 0.6983,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5840856924254016,
+      "grad_norm": 2.2039655542368712,
+      "learning_rate": 1.6610869071785055e-05,
+      "loss": 0.7334,
+      "step": 3817
+    },
+    {
+      "epoch": 0.584238714613619,
+      "grad_norm": 2.0194472601786146,
+      "learning_rate": 1.6609009464922613e-05,
+      "loss": 0.6393,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5843917368018363,
+      "grad_norm": 2.515878768807154,
+      "learning_rate": 1.660714945217504e-05,
+      "loss": 0.7962,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5845447589900535,
+      "grad_norm": 2.5345715323872806,
+      "learning_rate": 1.6605289033656575e-05,
+      "loss": 0.7854,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5846977811782709,
+      "grad_norm": 2.2418644854452334,
+      "learning_rate": 1.660342820948147e-05,
+      "loss": 0.8298,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5848508033664881,
+      "grad_norm": 2.4365303706928603,
+      "learning_rate": 1.6601566979764007e-05,
+      "loss": 0.7993,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5850038255547054,
+      "grad_norm": 2.4005844284928415,
+      "learning_rate": 1.659970534461849e-05,
+      "loss": 0.7509,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5851568477429228,
+      "grad_norm": 2.406705848873317,
+      "learning_rate": 1.6597843304159248e-05,
+      "loss": 0.7272,
+      "step": 3824
+    },
+    {
+      "epoch": 0.58530986993114,
+      "grad_norm": 2.3041140030698615,
+      "learning_rate": 1.659598085850064e-05,
+      "loss": 0.7891,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5854628921193573,
+      "grad_norm": 2.379151533908481,
+      "learning_rate": 1.659411800775704e-05,
+      "loss": 0.7599,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5856159143075746,
+      "grad_norm": 2.287716081596106,
+      "learning_rate": 1.659225475204286e-05,
+      "loss": 0.8061,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5857689364957919,
+      "grad_norm": 2.2414433200668227,
+      "learning_rate": 1.6590391091472526e-05,
+      "loss": 0.7691,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5859219586840092,
+      "grad_norm": 2.5086014363960554,
+      "learning_rate": 1.6588527026160493e-05,
+      "loss": 0.8479,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5860749808722264,
+      "grad_norm": 2.419448089243427,
+      "learning_rate": 1.658666255622124e-05,
+      "loss": 0.8236,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5862280030604438,
+      "grad_norm": 2.3163306355498157,
+      "learning_rate": 1.6584797681769273e-05,
+      "loss": 0.8364,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5863810252486611,
+      "grad_norm": 2.1990484963992585,
+      "learning_rate": 1.6582932402919124e-05,
+      "loss": 0.7264,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5865340474368783,
+      "grad_norm": 2.257065590223458,
+      "learning_rate": 1.658106671978534e-05,
+      "loss": 0.7227,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5866870696250956,
+      "grad_norm": 2.3568110415125805,
+      "learning_rate": 1.6579200632482502e-05,
+      "loss": 0.8643,
+      "step": 3834
+    },
+    {
+      "epoch": 0.586840091813313,
+      "grad_norm": 2.2997320714448537,
+      "learning_rate": 1.6577334141125217e-05,
+      "loss": 0.804,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5869931140015302,
+      "grad_norm": 2.4492067663950596,
+      "learning_rate": 1.657546724582811e-05,
+      "loss": 0.7471,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5871461361897475,
+      "grad_norm": 2.2367454813232346,
+      "learning_rate": 1.657359994670584e-05,
+      "loss": 0.7728,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5872991583779648,
+      "grad_norm": 2.387170074669952,
+      "learning_rate": 1.657173224387308e-05,
+      "loss": 0.7294,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5874521805661821,
+      "grad_norm": 2.6611783518744367,
+      "learning_rate": 1.6569864137444533e-05,
+      "loss": 0.8531,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5876052027543994,
+      "grad_norm": 2.5795969282876507,
+      "learning_rate": 1.6567995627534927e-05,
+      "loss": 0.7192,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5877582249426166,
+      "grad_norm": 2.228751239656882,
+      "learning_rate": 1.6566126714259017e-05,
+      "loss": 0.7878,
+      "step": 3841
+    },
+    {
+      "epoch": 0.587911247130834,
+      "grad_norm": 2.418610511457032,
+      "learning_rate": 1.6564257397731577e-05,
+      "loss": 0.8162,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5880642693190513,
+      "grad_norm": 2.1098189523920357,
+      "learning_rate": 1.656238767806741e-05,
+      "loss": 0.7597,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5882172915072685,
+      "grad_norm": 2.4049855489791803,
+      "learning_rate": 1.6560517555381348e-05,
+      "loss": 0.8416,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5883703136954859,
+      "grad_norm": 2.1864495709233913,
+      "learning_rate": 1.6558647029788234e-05,
+      "loss": 0.6129,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5885233358837031,
+      "grad_norm": 2.571107206508435,
+      "learning_rate": 1.655677610140295e-05,
+      "loss": 0.8715,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5886763580719204,
+      "grad_norm": 2.2997919979778016,
+      "learning_rate": 1.6554904770340393e-05,
+      "loss": 0.8151,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5888293802601378,
+      "grad_norm": 2.2281125845467495,
+      "learning_rate": 1.6553033036715493e-05,
+      "loss": 0.8202,
+      "step": 3848
+    },
+    {
+      "epoch": 0.588982402448355,
+      "grad_norm": 2.1321131423457773,
+      "learning_rate": 1.6551160900643203e-05,
+      "loss": 0.6831,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5891354246365723,
+      "grad_norm": 2.527648900196719,
+      "learning_rate": 1.654928836223849e-05,
+      "loss": 0.7904,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5892884468247896,
+      "grad_norm": 2.28923012533992,
+      "learning_rate": 1.6547415421616353e-05,
+      "loss": 0.7622,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5894414690130069,
+      "grad_norm": 2.7923038554466846,
+      "learning_rate": 1.6545542078891826e-05,
+      "loss": 0.8117,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5895944912012242,
+      "grad_norm": 2.4396804219014423,
+      "learning_rate": 1.654366833417995e-05,
+      "loss": 0.8149,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5897475133894414,
+      "grad_norm": 2.2650099970410795,
+      "learning_rate": 1.654179418759581e-05,
+      "loss": 0.7232,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5899005355776588,
+      "grad_norm": 2.2479396320602865,
+      "learning_rate": 1.6539919639254494e-05,
+      "loss": 0.7672,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5900535577658761,
+      "grad_norm": 2.345314819643059,
+      "learning_rate": 1.6538044689271126e-05,
+      "loss": 0.8204,
+      "step": 3856
+    },
+    {
+      "epoch": 0.5902065799540933,
+      "grad_norm": 2.0268666678732443,
+      "learning_rate": 1.653616933776086e-05,
+      "loss": 0.7722,
+      "step": 3857
+    },
+    {
+      "epoch": 0.5903596021423106,
+      "grad_norm": 2.3791503614863108,
+      "learning_rate": 1.653429358483886e-05,
+      "loss": 0.8463,
+      "step": 3858
+    },
+    {
+      "epoch": 0.590512624330528,
+      "grad_norm": 2.342414633790742,
+      "learning_rate": 1.6532417430620337e-05,
+      "loss": 0.77,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5906656465187452,
+      "grad_norm": 2.327695041795651,
+      "learning_rate": 1.65305408752205e-05,
+      "loss": 0.8732,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5908186687069625,
+      "grad_norm": 2.437108756237025,
+      "learning_rate": 1.6528663918754597e-05,
+      "loss": 0.8092,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5909716908951798,
+      "grad_norm": 2.2774300160190046,
+      "learning_rate": 1.652678656133791e-05,
+      "loss": 0.7207,
+      "step": 3862
+    },
+    {
+      "epoch": 0.5911247130833971,
+      "grad_norm": 2.4901626265698886,
+      "learning_rate": 1.652490880308572e-05,
+      "loss": 0.8836,
+      "step": 3863
+    },
+    {
+      "epoch": 0.5912777352716144,
+      "grad_norm": 2.49244263349783,
+      "learning_rate": 1.6523030644113357e-05,
+      "loss": 0.7631,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5914307574598316,
+      "grad_norm": 2.166963403745061,
+      "learning_rate": 1.6521152084536164e-05,
+      "loss": 0.7394,
+      "step": 3865
+    },
+    {
+      "epoch": 0.591583779648049,
+      "grad_norm": 2.3942979926831196,
+      "learning_rate": 1.6519273124469512e-05,
+      "loss": 0.8902,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5917368018362663,
+      "grad_norm": 2.1880020195653764,
+      "learning_rate": 1.6517393764028793e-05,
+      "loss": 0.7427,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5918898240244835,
+      "grad_norm": 2.106975274430694,
+      "learning_rate": 1.651551400332943e-05,
+      "loss": 0.7053,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5920428462127009,
+      "grad_norm": 2.377254747191573,
+      "learning_rate": 1.6513633842486858e-05,
+      "loss": 0.8433,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5921958684009181,
+      "grad_norm": 2.1159960070343753,
+      "learning_rate": 1.6511753281616552e-05,
+      "loss": 0.6735,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5923488905891354,
+      "grad_norm": 1.9438413653670783,
+      "learning_rate": 1.6509872320834003e-05,
+      "loss": 0.6788,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5925019127773528,
+      "grad_norm": 2.0955062116382943,
+      "learning_rate": 1.6507990960254728e-05,
+      "loss": 0.7541,
+      "step": 3872
+    },
+    {
+      "epoch": 0.59265493496557,
+      "grad_norm": 2.3820312322134187,
+      "learning_rate": 1.650610919999427e-05,
+      "loss": 0.8307,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5928079571537873,
+      "grad_norm": 2.2354983100993446,
+      "learning_rate": 1.6504227040168194e-05,
+      "loss": 0.772,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5929609793420046,
+      "grad_norm": 2.2484229287109625,
+      "learning_rate": 1.650234448089209e-05,
+      "loss": 0.8463,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5931140015302219,
+      "grad_norm": 2.1798032783600783,
+      "learning_rate": 1.6500461522281575e-05,
+      "loss": 0.6458,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5932670237184392,
+      "grad_norm": 2.2753556732841402,
+      "learning_rate": 1.6498578164452285e-05,
+      "loss": 0.7703,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5934200459066564,
+      "grad_norm": 2.4188609922833737,
+      "learning_rate": 1.6496694407519888e-05,
+      "loss": 0.8333,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5935730680948738,
+      "grad_norm": 2.3421062788752307,
+      "learning_rate": 1.6494810251600075e-05,
+      "loss": 0.7059,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5937260902830911,
+      "grad_norm": 2.1227249655673717,
+      "learning_rate": 1.6492925696808555e-05,
+      "loss": 0.8143,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5938791124713083,
+      "grad_norm": 2.308489799932691,
+      "learning_rate": 1.6491040743261065e-05,
+      "loss": 0.8233,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5940321346595256,
+      "grad_norm": 2.4998079895811687,
+      "learning_rate": 1.6489155391073375e-05,
+      "loss": 0.7824,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5941851568477429,
+      "grad_norm": 2.135136759805042,
+      "learning_rate": 1.6487269640361264e-05,
+      "loss": 0.789,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5943381790359602,
+      "grad_norm": 2.0920550383410474,
+      "learning_rate": 1.6485383491240546e-05,
+      "loss": 0.743,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5944912012241775,
+      "grad_norm": 2.355458998837855,
+      "learning_rate": 1.6483496943827056e-05,
+      "loss": 0.728,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5946442234123948,
+      "grad_norm": 2.1107113168909004,
+      "learning_rate": 1.6481609998236656e-05,
+      "loss": 0.6521,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5947972456006121,
+      "grad_norm": 2.116378932065588,
+      "learning_rate": 1.647972265458523e-05,
+      "loss": 0.765,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5949502677888294,
+      "grad_norm": 2.195681235916977,
+      "learning_rate": 1.647783491298869e-05,
+      "loss": 0.7999,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5951032899770466,
+      "grad_norm": 2.65848496667216,
+      "learning_rate": 1.647594677356296e-05,
+      "loss": 0.7752,
+      "step": 3889
+    },
+    {
+      "epoch": 0.595256312165264,
+      "grad_norm": 2.3804867928581412,
+      "learning_rate": 1.647405823642401e-05,
+      "loss": 0.7683,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5954093343534812,
+      "grad_norm": 2.2256140777105893,
+      "learning_rate": 1.6472169301687816e-05,
+      "loss": 0.6946,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5955623565416985,
+      "grad_norm": 2.259429617574836,
+      "learning_rate": 1.6470279969470384e-05,
+      "loss": 0.8007,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5957153787299159,
+      "grad_norm": 2.4244045113345667,
+      "learning_rate": 1.646839023988775e-05,
+      "loss": 0.8071,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5958684009181331,
+      "grad_norm": 2.21180016617028,
+      "learning_rate": 1.6466500113055963e-05,
+      "loss": 0.6792,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5960214231063504,
+      "grad_norm": 2.1941067525426727,
+      "learning_rate": 1.646460958909111e-05,
+      "loss": 0.8207,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5961744452945678,
+      "grad_norm": 2.1232749753013667,
+      "learning_rate": 1.646271866810929e-05,
+      "loss": 0.7,
+      "step": 3896
+    },
+    {
+      "epoch": 0.596327467482785,
+      "grad_norm": 2.190229920133476,
+      "learning_rate": 1.6460827350226637e-05,
+      "loss": 0.6335,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5964804896710023,
+      "grad_norm": 2.2639688761323944,
+      "learning_rate": 1.64589356355593e-05,
+      "loss": 0.7466,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5966335118592195,
+      "grad_norm": 2.460362228450079,
+      "learning_rate": 1.645704352422346e-05,
+      "loss": 0.8105,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5967865340474369,
+      "grad_norm": 2.2461116435022377,
+      "learning_rate": 1.6455151016335312e-05,
+      "loss": 0.697,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5969395562356542,
+      "grad_norm": 2.3594462673289947,
+      "learning_rate": 1.6453258112011094e-05,
+      "loss": 0.7736,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5970925784238714,
+      "grad_norm": 2.106693165284362,
+      "learning_rate": 1.6451364811367044e-05,
+      "loss": 0.5927,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5972456006120888,
+      "grad_norm": 2.247234785821779,
+      "learning_rate": 1.6449471114519447e-05,
+      "loss": 0.7286,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5973986228003061,
+      "grad_norm": 2.0774276621075454,
+      "learning_rate": 1.6447577021584597e-05,
+      "loss": 0.6991,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5975516449885233,
+      "grad_norm": 2.430421209143389,
+      "learning_rate": 1.644568253267882e-05,
+      "loss": 0.7457,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5977046671767406,
+      "grad_norm": 2.330147887370278,
+      "learning_rate": 1.6443787647918464e-05,
+      "loss": 0.7581,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5978576893649579,
+      "grad_norm": 2.0030331146685665,
+      "learning_rate": 1.6441892367419895e-05,
+      "loss": 0.5928,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5980107115531752,
+      "grad_norm": 2.3916619016698166,
+      "learning_rate": 1.643999669129952e-05,
+      "loss": 0.781,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5981637337413925,
+      "grad_norm": 2.399757751453464,
+      "learning_rate": 1.6438100619673757e-05,
+      "loss": 0.7236,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5983167559296098,
+      "grad_norm": 2.5738835936498603,
+      "learning_rate": 1.6436204152659042e-05,
+      "loss": 0.7948,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5984697781178271,
+      "grad_norm": 2.3044877765883913,
+      "learning_rate": 1.6434307290371855e-05,
+      "loss": 0.6954,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5986228003060444,
+      "grad_norm": 2.2008661487646046,
+      "learning_rate": 1.6432410032928686e-05,
+      "loss": 0.7159,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5987758224942616,
+      "grad_norm": 2.3895382666991876,
+      "learning_rate": 1.6430512380446052e-05,
+      "loss": 0.8146,
+      "step": 3913
+    },
+    {
+      "epoch": 0.598928844682479,
+      "grad_norm": 2.065309607267929,
+      "learning_rate": 1.64286143330405e-05,
+      "loss": 0.8357,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5990818668706962,
+      "grad_norm": 2.3359874366462914,
+      "learning_rate": 1.642671589082859e-05,
+      "loss": 0.7237,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5992348890589135,
+      "grad_norm": 2.4962956403966934,
+      "learning_rate": 1.6424817053926917e-05,
+      "loss": 0.8363,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5993879112471309,
+      "grad_norm": 2.468183246991395,
+      "learning_rate": 1.6422917822452093e-05,
+      "loss": 0.8228,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5995409334353481,
+      "grad_norm": 2.5397453332829016,
+      "learning_rate": 1.6421018196520758e-05,
+      "loss": 0.9307,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5996939556235654,
+      "grad_norm": 2.225012213725225,
+      "learning_rate": 1.641911817624958e-05,
+      "loss": 0.76,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5998469778117828,
+      "grad_norm": 2.3966394488811016,
+      "learning_rate": 1.641721776175524e-05,
+      "loss": 0.7146,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.303920410231337,
+      "learning_rate": 1.6415316953154455e-05,
+      "loss": 0.8073,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6001530221882173,
+      "grad_norm": 2.211302259446293,
+      "learning_rate": 1.6413415750563957e-05,
+      "loss": 0.7277,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6003060443764345,
+      "grad_norm": 2.2345826594193166,
+      "learning_rate": 1.6411514154100513e-05,
+      "loss": 0.6917,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6004590665646519,
+      "grad_norm": 2.201839243236848,
+      "learning_rate": 1.6409612163880898e-05,
+      "loss": 0.704,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6006120887528692,
+      "grad_norm": 2.2928406532733194,
+      "learning_rate": 1.6407709780021925e-05,
+      "loss": 0.754,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6007651109410864,
+      "grad_norm": 2.3920891472308736,
+      "learning_rate": 1.640580700264043e-05,
+      "loss": 0.8704,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6009181331293038,
+      "grad_norm": 2.1292889428240294,
+      "learning_rate": 1.6403903831853265e-05,
+      "loss": 0.7045,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6010711553175211,
+      "grad_norm": 2.379304273256747,
+      "learning_rate": 1.640200026777732e-05,
+      "loss": 0.8551,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6012241775057383,
+      "grad_norm": 2.282693014805505,
+      "learning_rate": 1.6400096310529487e-05,
+      "loss": 0.7696,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6013771996939556,
+      "grad_norm": 2.206776247521118,
+      "learning_rate": 1.63981919602267e-05,
+      "loss": 0.6743,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6015302218821729,
+      "grad_norm": 2.608385419308765,
+      "learning_rate": 1.639628721698592e-05,
+      "loss": 0.8967,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6016832440703902,
+      "grad_norm": 2.455187279779116,
+      "learning_rate": 1.639438208092412e-05,
+      "loss": 0.7461,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6018362662586075,
+      "grad_norm": 2.2605743335848274,
+      "learning_rate": 1.63924765521583e-05,
+      "loss": 0.7199,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6019892884468248,
+      "grad_norm": 2.445481051817927,
+      "learning_rate": 1.6390570630805487e-05,
+      "loss": 0.767,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6021423106350421,
+      "grad_norm": 2.583444641012235,
+      "learning_rate": 1.638866431698273e-05,
+      "loss": 0.8196,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6022953328232594,
+      "grad_norm": 2.5086801152333646,
+      "learning_rate": 1.6386757610807106e-05,
+      "loss": 0.816,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6024483550114766,
+      "grad_norm": 2.3705704487061348,
+      "learning_rate": 1.6384850512395715e-05,
+      "loss": 0.7676,
+      "step": 3937
+    },
+    {
+      "epoch": 0.602601377199694,
+      "grad_norm": 2.1626859005056236,
+      "learning_rate": 1.638294302186567e-05,
+      "loss": 0.7177,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6027543993879112,
+      "grad_norm": 2.46703763158888,
+      "learning_rate": 1.6381035139334128e-05,
+      "loss": 0.7297,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6029074215761285,
+      "grad_norm": 2.5716263838061666,
+      "learning_rate": 1.6379126864918256e-05,
+      "loss": 0.9171,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6030604437643459,
+      "grad_norm": 2.193123379792748,
+      "learning_rate": 1.6377218198735246e-05,
+      "loss": 0.7157,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6032134659525631,
+      "grad_norm": 2.3421215315975807,
+      "learning_rate": 1.637530914090232e-05,
+      "loss": 0.748,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6033664881407804,
+      "grad_norm": 2.20680704151605,
+      "learning_rate": 1.637339969153672e-05,
+      "loss": 0.7783,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6035195103289978,
+      "grad_norm": 2.5368724067005823,
+      "learning_rate": 1.6371489850755712e-05,
+      "loss": 0.7137,
+      "step": 3944
+    },
+    {
+      "epoch": 0.603672532517215,
+      "grad_norm": 2.347964186524102,
+      "learning_rate": 1.6369579618676584e-05,
+      "loss": 0.7715,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6038255547054323,
+      "grad_norm": 2.4869195064086487,
+      "learning_rate": 1.636766899541666e-05,
+      "loss": 0.8303,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6039785768936495,
+      "grad_norm": 2.2848097041733326,
+      "learning_rate": 1.6365757981093266e-05,
+      "loss": 0.8186,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6041315990818669,
+      "grad_norm": 2.4440133764573035,
+      "learning_rate": 1.6363846575823772e-05,
+      "loss": 0.8781,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6042846212700842,
+      "grad_norm": 2.527286580207838,
+      "learning_rate": 1.6361934779725564e-05,
+      "loss": 0.7026,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6044376434583014,
+      "grad_norm": 2.6768839444382713,
+      "learning_rate": 1.6360022592916056e-05,
+      "loss": 0.7902,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6045906656465188,
+      "grad_norm": 2.563244839367599,
+      "learning_rate": 1.635811001551268e-05,
+      "loss": 0.7985,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6047436878347361,
+      "grad_norm": 2.435160345656961,
+      "learning_rate": 1.6356197047632894e-05,
+      "loss": 0.8541,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6048967100229533,
+      "grad_norm": 2.660440652538607,
+      "learning_rate": 1.635428368939418e-05,
+      "loss": 0.7915,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6050497322111706,
+      "grad_norm": 2.302355575360249,
+      "learning_rate": 1.635236994091405e-05,
+      "loss": 0.7916,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6052027543993879,
+      "grad_norm": 2.5678687348149127,
+      "learning_rate": 1.6350455802310027e-05,
+      "loss": 0.8768,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6053557765876052,
+      "grad_norm": 2.4133307260880015,
+      "learning_rate": 1.6348541273699672e-05,
+      "loss": 0.8734,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6055087987758225,
+      "grad_norm": 2.1442604005186716,
+      "learning_rate": 1.6346626355200564e-05,
+      "loss": 0.8057,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6056618209640398,
+      "grad_norm": 2.2777122530462344,
+      "learning_rate": 1.63447110469303e-05,
+      "loss": 0.6944,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6058148431522571,
+      "grad_norm": 2.2702458426898118,
+      "learning_rate": 1.6342795349006514e-05,
+      "loss": 0.7787,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6059678653404744,
+      "grad_norm": 2.4138860942835665,
+      "learning_rate": 1.6340879261546848e-05,
+      "loss": 0.7319,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6061208875286916,
+      "grad_norm": 2.406986355064948,
+      "learning_rate": 1.6338962784668984e-05,
+      "loss": 0.8305,
+      "step": 3961
+    },
+    {
+      "epoch": 0.606273909716909,
+      "grad_norm": 2.1320662468950053,
+      "learning_rate": 1.633704591849061e-05,
+      "loss": 0.7107,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6064269319051262,
+      "grad_norm": 2.170118161412104,
+      "learning_rate": 1.6335128663129466e-05,
+      "loss": 0.6988,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6065799540933435,
+      "grad_norm": 2.288231162241274,
+      "learning_rate": 1.6333211018703282e-05,
+      "loss": 0.7682,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6067329762815609,
+      "grad_norm": 2.0736456947567863,
+      "learning_rate": 1.6331292985329835e-05,
+      "loss": 0.7372,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6068859984697781,
+      "grad_norm": 2.107363700837136,
+      "learning_rate": 1.632937456312692e-05,
+      "loss": 0.7282,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6070390206579954,
+      "grad_norm": 2.0881937408877524,
+      "learning_rate": 1.6327455752212348e-05,
+      "loss": 0.7685,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6071920428462128,
+      "grad_norm": 2.1645877937641425,
+      "learning_rate": 1.6325536552703963e-05,
+      "loss": 0.834,
+      "step": 3968
+    },
+    {
+      "epoch": 0.60734506503443,
+      "grad_norm": 2.377608992372084,
+      "learning_rate": 1.6323616964719642e-05,
+      "loss": 0.7594,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6074980872226473,
+      "grad_norm": 2.028436733602878,
+      "learning_rate": 1.632169698837726e-05,
+      "loss": 0.7597,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6076511094108645,
+      "grad_norm": 2.113181958173014,
+      "learning_rate": 1.631977662379473e-05,
+      "loss": 0.8032,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6078041315990819,
+      "grad_norm": 2.5583012059184402,
+      "learning_rate": 1.6317855871090003e-05,
+      "loss": 0.7125,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6079571537872992,
+      "grad_norm": 2.3758675047618008,
+      "learning_rate": 1.6315934730381027e-05,
+      "loss": 0.81,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6081101759755164,
+      "grad_norm": 2.2335382453442225,
+      "learning_rate": 1.631401320178579e-05,
+      "loss": 0.7042,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6082631981637338,
+      "grad_norm": 2.388399327130747,
+      "learning_rate": 1.6312091285422305e-05,
+      "loss": 0.8333,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6084162203519511,
+      "grad_norm": 2.545928594962447,
+      "learning_rate": 1.63101689814086e-05,
+      "loss": 0.8372,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6085692425401683,
+      "grad_norm": 2.0449415964423037,
+      "learning_rate": 1.630824628986273e-05,
+      "loss": 0.7397,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6087222647283856,
+      "grad_norm": 2.567039032923848,
+      "learning_rate": 1.6306323210902784e-05,
+      "loss": 0.7466,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6088752869166029,
+      "grad_norm": 2.4160435851987905,
+      "learning_rate": 1.6304399744646854e-05,
+      "loss": 0.7566,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6090283091048202,
+      "grad_norm": 2.536239714103674,
+      "learning_rate": 1.630247589121307e-05,
+      "loss": 0.9314,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6091813312930375,
+      "grad_norm": 2.278829958141022,
+      "learning_rate": 1.630055165071959e-05,
+      "loss": 0.7122,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6093343534812548,
+      "grad_norm": 2.387173473026114,
+      "learning_rate": 1.6298627023284584e-05,
+      "loss": 0.7642,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6094873756694721,
+      "grad_norm": 2.3274832439959856,
+      "learning_rate": 1.6296702009026256e-05,
+      "loss": 0.7805,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6096403978576894,
+      "grad_norm": 2.413713126301562,
+      "learning_rate": 1.6294776608062818e-05,
+      "loss": 0.8906,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6097934200459066,
+      "grad_norm": 2.2657003983898134,
+      "learning_rate": 1.629285082051253e-05,
+      "loss": 0.759,
+      "step": 3985
+    },
+    {
+      "epoch": 0.609946442234124,
+      "grad_norm": 2.059377996694824,
+      "learning_rate": 1.6290924646493654e-05,
+      "loss": 0.6946,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6100994644223412,
+      "grad_norm": 2.240469903234145,
+      "learning_rate": 1.6288998086124478e-05,
+      "loss": 0.7731,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6102524866105585,
+      "grad_norm": 2.1045040115900506,
+      "learning_rate": 1.6287071139523334e-05,
+      "loss": 0.7427,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6104055087987759,
+      "grad_norm": 2.2817203881792567,
+      "learning_rate": 1.6285143806808554e-05,
+      "loss": 0.7773,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6105585309869931,
+      "grad_norm": 2.445802561201833,
+      "learning_rate": 1.628321608809851e-05,
+      "loss": 0.838,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6107115531752104,
+      "grad_norm": 2.5828642223393214,
+      "learning_rate": 1.628128798351158e-05,
+      "loss": 0.6483,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6108645753634276,
+      "grad_norm": 2.4672233905559477,
+      "learning_rate": 1.6279359493166183e-05,
+      "loss": 0.8611,
+      "step": 3992
+    },
+    {
+      "epoch": 0.611017597551645,
+      "grad_norm": 2.219671457282571,
+      "learning_rate": 1.6277430617180755e-05,
+      "loss": 0.7172,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6111706197398623,
+      "grad_norm": 2.4789990648031153,
+      "learning_rate": 1.6275501355673756e-05,
+      "loss": 0.8063,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6113236419280795,
+      "grad_norm": 2.3551450398894467,
+      "learning_rate": 1.6273571708763665e-05,
+      "loss": 0.7361,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6114766641162969,
+      "grad_norm": 2.529516672420973,
+      "learning_rate": 1.6271641676569e-05,
+      "loss": 0.8252,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6116296863045142,
+      "grad_norm": 2.1194774939225143,
+      "learning_rate": 1.626971125920828e-05,
+      "loss": 0.7994,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6117827084927314,
+      "grad_norm": 2.130797951705835,
+      "learning_rate": 1.6267780456800066e-05,
+      "loss": 0.7897,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6119357306809488,
+      "grad_norm": 2.3216160730004636,
+      "learning_rate": 1.6265849269462936e-05,
+      "loss": 0.8488,
+      "step": 3999
+    },
+    {
+      "epoch": 0.612088752869166,
+      "grad_norm": 2.7626139147850224,
+      "learning_rate": 1.6263917697315488e-05,
+      "loss": 0.7936,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6122417750573833,
+      "grad_norm": 2.082553450963893,
+      "learning_rate": 1.6261985740476348e-05,
+      "loss": 0.7721,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6123947972456006,
+      "grad_norm": 2.3936912056745965,
+      "learning_rate": 1.626005339906417e-05,
+      "loss": 0.8584,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6125478194338179,
+      "grad_norm": 2.2087315724400036,
+      "learning_rate": 1.6258120673197623e-05,
+      "loss": 0.7804,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6127008416220352,
+      "grad_norm": 2.354583524321165,
+      "learning_rate": 1.6256187562995403e-05,
+      "loss": 0.7383,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6128538638102525,
+      "grad_norm": 2.098041581723552,
+      "learning_rate": 1.6254254068576227e-05,
+      "loss": 0.7599,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6130068859984698,
+      "grad_norm": 2.0327806604154754,
+      "learning_rate": 1.6252320190058847e-05,
+      "loss": 0.7705,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6131599081866871,
+      "grad_norm": 2.496718004055835,
+      "learning_rate": 1.625038592756202e-05,
+      "loss": 0.7543,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6133129303749043,
+      "grad_norm": 2.4936165790493554,
+      "learning_rate": 1.6248451281204545e-05,
+      "loss": 0.8754,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6134659525631216,
+      "grad_norm": 2.2205096614417075,
+      "learning_rate": 1.624651625110523e-05,
+      "loss": 0.7938,
+      "step": 4009
+    },
+    {
+      "epoch": 0.613618974751339,
+      "grad_norm": 2.1301811356198046,
+      "learning_rate": 1.624458083738292e-05,
+      "loss": 0.7472,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6137719969395562,
+      "grad_norm": 2.3619766841954326,
+      "learning_rate": 1.624264504015647e-05,
+      "loss": 0.7185,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6139250191277735,
+      "grad_norm": 2.6369274563223524,
+      "learning_rate": 1.6240708859544766e-05,
+      "loss": 0.8762,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6140780413159909,
+      "grad_norm": 2.3391139454523877,
+      "learning_rate": 1.6238772295666718e-05,
+      "loss": 0.8134,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6142310635042081,
+      "grad_norm": 2.472559033054485,
+      "learning_rate": 1.6236835348641254e-05,
+      "loss": 0.8442,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6143840856924254,
+      "grad_norm": 2.30552269826598,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.7564,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6145371078806426,
+      "grad_norm": 2.2725183178366644,
+      "learning_rate": 1.623296030562394e-05,
+      "loss": 0.8004,
+      "step": 4016
+    },
+    {
+      "epoch": 0.61469013006886,
+      "grad_norm": 2.1530496291189736,
+      "learning_rate": 1.6231022209870063e-05,
+      "loss": 0.7653,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6148431522570773,
+      "grad_norm": 2.1620242897683797,
+      "learning_rate": 1.6229083731444745e-05,
+      "loss": 0.7129,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6149961744452945,
+      "grad_norm": 2.25085442826096,
+      "learning_rate": 1.622714487046702e-05,
+      "loss": 0.7554,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6151491966335119,
+      "grad_norm": 2.169103357595413,
+      "learning_rate": 1.622520562705597e-05,
+      "loss": 0.7607,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6153022188217292,
+      "grad_norm": 2.3292061531576955,
+      "learning_rate": 1.622326600133069e-05,
+      "loss": 0.8104,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6154552410099464,
+      "grad_norm": 2.170753862893607,
+      "learning_rate": 1.6221325993410295e-05,
+      "loss": 0.745,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6156082631981638,
+      "grad_norm": 2.4632365927830207,
+      "learning_rate": 1.6219385603413937e-05,
+      "loss": 0.68,
+      "step": 4023
+    },
+    {
+      "epoch": 0.615761285386381,
+      "grad_norm": 2.085184298715942,
+      "learning_rate": 1.6217444831460777e-05,
+      "loss": 0.7114,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6159143075745983,
+      "grad_norm": 2.7017932076046374,
+      "learning_rate": 1.621550367767001e-05,
+      "loss": 0.8658,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6160673297628156,
+      "grad_norm": 2.66819564271173,
+      "learning_rate": 1.6213562142160842e-05,
+      "loss": 0.8669,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6162203519510329,
+      "grad_norm": 2.137016223484698,
+      "learning_rate": 1.621162022505252e-05,
+      "loss": 0.7017,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6163733741392502,
+      "grad_norm": 2.3297061769686174,
+      "learning_rate": 1.6209677926464297e-05,
+      "loss": 0.7102,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6165263963274675,
+      "grad_norm": 2.2418631040587287,
+      "learning_rate": 1.620773524651546e-05,
+      "loss": 0.8347,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6166794185156848,
+      "grad_norm": 2.2617571464792423,
+      "learning_rate": 1.6205792185325318e-05,
+      "loss": 0.8687,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6168324407039021,
+      "grad_norm": 2.5724992577181705,
+      "learning_rate": 1.6203848743013202e-05,
+      "loss": 0.724,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6169854628921193,
+      "grad_norm": 2.227267112590208,
+      "learning_rate": 1.6201904919698463e-05,
+      "loss": 0.6963,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6171384850803366,
+      "grad_norm": 2.172006678534527,
+      "learning_rate": 1.6199960715500482e-05,
+      "loss": 0.7051,
+      "step": 4033
+    },
+    {
+      "epoch": 0.617291507268554,
+      "grad_norm": 2.1373746947180012,
+      "learning_rate": 1.619801613053866e-05,
+      "loss": 0.7327,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6174445294567712,
+      "grad_norm": 2.11476592544764,
+      "learning_rate": 1.6196071164932415e-05,
+      "loss": 0.7032,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6175975516449885,
+      "grad_norm": 2.300625622423079,
+      "learning_rate": 1.6194125818801207e-05,
+      "loss": 0.7691,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6177505738332059,
+      "grad_norm": 2.428594165574761,
+      "learning_rate": 1.61921800922645e-05,
+      "loss": 0.723,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6179035960214231,
+      "grad_norm": 2.169473308137043,
+      "learning_rate": 1.6190233985441786e-05,
+      "loss": 0.8408,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6180566182096404,
+      "grad_norm": 2.5360559820737376,
+      "learning_rate": 1.618828749845259e-05,
+      "loss": 0.8811,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6182096403978576,
+      "grad_norm": 2.39178836667541,
+      "learning_rate": 1.6186340631416452e-05,
+      "loss": 0.788,
+      "step": 4040
+    },
+    {
+      "epoch": 0.618362662586075,
+      "grad_norm": 2.3918450316007376,
+      "learning_rate": 1.6184393384452928e-05,
+      "loss": 0.7039,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6185156847742923,
+      "grad_norm": 2.338128319850209,
+      "learning_rate": 1.6182445757681616e-05,
+      "loss": 0.8499,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6186687069625095,
+      "grad_norm": 2.3173269279379136,
+      "learning_rate": 1.6180497751222127e-05,
+      "loss": 0.7303,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6188217291507269,
+      "grad_norm": 2.438141149323398,
+      "learning_rate": 1.6178549365194086e-05,
+      "loss": 0.8503,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6189747513389442,
+      "grad_norm": 2.2321932250097354,
+      "learning_rate": 1.6176600599717165e-05,
+      "loss": 0.7228,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6191277735271614,
+      "grad_norm": 2.3226574409768825,
+      "learning_rate": 1.6174651454911034e-05,
+      "loss": 0.7527,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6192807957153788,
+      "grad_norm": 2.0741031426015915,
+      "learning_rate": 1.6172701930895404e-05,
+      "loss": 0.7485,
+      "step": 4047
+    },
+    {
+      "epoch": 0.619433817903596,
+      "grad_norm": 2.4837597123114494,
+      "learning_rate": 1.6170752027790002e-05,
+      "loss": 0.7971,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6195868400918133,
+      "grad_norm": 2.349665609451401,
+      "learning_rate": 1.6168801745714576e-05,
+      "loss": 0.7738,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6197398622800306,
+      "grad_norm": 2.3536660431739045,
+      "learning_rate": 1.6166851084788898e-05,
+      "loss": 0.7474,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6198928844682479,
+      "grad_norm": 2.516752609598269,
+      "learning_rate": 1.6164900045132777e-05,
+      "loss": 0.7539,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6200459066564652,
+      "grad_norm": 2.5523705841883824,
+      "learning_rate": 1.6162948626866023e-05,
+      "loss": 0.8509,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6201989288446825,
+      "grad_norm": 2.2026496418341495,
+      "learning_rate": 1.6160996830108486e-05,
+      "loss": 0.8403,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6203519510328998,
+      "grad_norm": 2.017536909380819,
+      "learning_rate": 1.615904465498003e-05,
+      "loss": 0.6502,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6205049732211171,
+      "grad_norm": 2.174952113668595,
+      "learning_rate": 1.6157092101600548e-05,
+      "loss": 0.7717,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6206579954093343,
+      "grad_norm": 1.8763628970458017,
+      "learning_rate": 1.6155139170089956e-05,
+      "loss": 0.757,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6208110175975516,
+      "grad_norm": 2.322317658519997,
+      "learning_rate": 1.6153185860568187e-05,
+      "loss": 0.7672,
+      "step": 4057
+    },
+    {
+      "epoch": 0.620964039785769,
+      "grad_norm": 2.3623718327673764,
+      "learning_rate": 1.61512321731552e-05,
+      "loss": 0.7172,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6211170619739862,
+      "grad_norm": 2.1276117730331805,
+      "learning_rate": 1.6149278107970983e-05,
+      "loss": 0.7645,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6212700841622035,
+      "grad_norm": 2.564749256711884,
+      "learning_rate": 1.6147323665135542e-05,
+      "loss": 0.7936,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6214231063504209,
+      "grad_norm": 2.413120918684202,
+      "learning_rate": 1.6145368844768908e-05,
+      "loss": 0.6705,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6215761285386381,
+      "grad_norm": 2.445889193129111,
+      "learning_rate": 1.6143413646991132e-05,
+      "loss": 0.8038,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6217291507268554,
+      "grad_norm": 2.117825343139216,
+      "learning_rate": 1.6141458071922285e-05,
+      "loss": 0.7411,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6218821729150726,
+      "grad_norm": 2.344357755958016,
+      "learning_rate": 1.6139502119682478e-05,
+      "loss": 0.855,
+      "step": 4064
+    },
+    {
+      "epoch": 0.62203519510329,
+      "grad_norm": 2.275018128282366,
+      "learning_rate": 1.613754579039183e-05,
+      "loss": 0.7594,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6221882172915073,
+      "grad_norm": 2.123207758811535,
+      "learning_rate": 1.613558908417048e-05,
+      "loss": 0.831,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6223412394797245,
+      "grad_norm": 2.318206412634962,
+      "learning_rate": 1.6133632001138604e-05,
+      "loss": 0.7474,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6224942616679419,
+      "grad_norm": 2.3643768786552535,
+      "learning_rate": 1.613167454141639e-05,
+      "loss": 0.728,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6226472838561592,
+      "grad_norm": 2.1021830426618084,
+      "learning_rate": 1.6129716705124055e-05,
+      "loss": 0.7078,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6228003060443764,
+      "grad_norm": 2.1384768150213858,
+      "learning_rate": 1.612775849238184e-05,
+      "loss": 0.7496,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6229533282325938,
+      "grad_norm": 2.0607954889554434,
+      "learning_rate": 1.612579990331e-05,
+      "loss": 0.6618,
+      "step": 4071
+    },
+    {
+      "epoch": 0.623106350420811,
+      "grad_norm": 2.370009379278904,
+      "learning_rate": 1.612384093802883e-05,
+      "loss": 0.8368,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6232593726090283,
+      "grad_norm": 2.0723916821807427,
+      "learning_rate": 1.612188159665863e-05,
+      "loss": 0.6796,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6234123947972456,
+      "grad_norm": 2.7584617297969425,
+      "learning_rate": 1.6119921879319728e-05,
+      "loss": 0.8882,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6235654169854629,
+      "grad_norm": 2.436001429058616,
+      "learning_rate": 1.6117961786132485e-05,
+      "loss": 0.7565,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6237184391736802,
+      "grad_norm": 2.4908275513686626,
+      "learning_rate": 1.6116001317217277e-05,
+      "loss": 0.8403,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6238714613618975,
+      "grad_norm": 2.1720916230556515,
+      "learning_rate": 1.6114040472694496e-05,
+      "loss": 0.7549,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6240244835501148,
+      "grad_norm": 2.1638352336847517,
+      "learning_rate": 1.611207925268458e-05,
+      "loss": 0.6232,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6241775057383321,
+      "grad_norm": 2.254748294563211,
+      "learning_rate": 1.611011765730796e-05,
+      "loss": 0.8494,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6243305279265493,
+      "grad_norm": 2.051169583162102,
+      "learning_rate": 1.610815568668511e-05,
+      "loss": 0.7465,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6244835501147666,
+      "grad_norm": 2.480860331299139,
+      "learning_rate": 1.610619334093653e-05,
+      "loss": 0.9505,
+      "step": 4081
+    },
+    {
+      "epoch": 0.624636572302984,
+      "grad_norm": 2.581081394347132,
+      "learning_rate": 1.6104230620182724e-05,
+      "loss": 0.8689,
+      "step": 4082
+    },
+    {
+      "epoch": 0.6247895944912012,
+      "grad_norm": 2.4647455923728416,
+      "learning_rate": 1.610226752454424e-05,
+      "loss": 0.7986,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6249426166794185,
+      "grad_norm": 2.3084776587149376,
+      "learning_rate": 1.6100304054141634e-05,
+      "loss": 0.8543,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6250956388676359,
+      "grad_norm": 2.278351087731257,
+      "learning_rate": 1.609834020909549e-05,
+      "loss": 0.8122,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6252486610558531,
+      "grad_norm": 2.276115015814235,
+      "learning_rate": 1.609637598952642e-05,
+      "loss": 0.647,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6254016832440704,
+      "grad_norm": 2.4313627499110235,
+      "learning_rate": 1.609441139555505e-05,
+      "loss": 0.7891,
+      "step": 4087
+    },
+    {
+      "epoch": 0.6255547054322876,
+      "grad_norm": 2.4151414352242297,
+      "learning_rate": 1.6092446427302028e-05,
+      "loss": 0.7811,
+      "step": 4088
+    },
+    {
+      "epoch": 0.625707727620505,
+      "grad_norm": 2.1618931472619822,
+      "learning_rate": 1.609048108488805e-05,
+      "loss": 0.8202,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6258607498087223,
+      "grad_norm": 2.2892118398805485,
+      "learning_rate": 1.6088515368433794e-05,
+      "loss": 0.7341,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6260137719969395,
+      "grad_norm": 2.319986784666118,
+      "learning_rate": 1.6086549278059988e-05,
+      "loss": 0.7378,
+      "step": 4091
+    },
+    {
+      "epoch": 0.6261667941851569,
+      "grad_norm": 2.388332513852767,
+      "learning_rate": 1.6084582813887385e-05,
+      "loss": 0.8177,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6263198163733742,
+      "grad_norm": 2.4953449296501113,
+      "learning_rate": 1.608261597603675e-05,
+      "loss": 0.7417,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6264728385615914,
+      "grad_norm": 2.2534640175752516,
+      "learning_rate": 1.608064876462887e-05,
+      "loss": 0.6792,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6266258607498087,
+      "grad_norm": 2.1813003582366566,
+      "learning_rate": 1.6078681179784563e-05,
+      "loss": 0.7858,
+      "step": 4095
+    },
+    {
+      "epoch": 0.626778882938026,
+      "grad_norm": 2.265021488746793,
+      "learning_rate": 1.6076713221624663e-05,
+      "loss": 0.7861,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6269319051262433,
+      "grad_norm": 2.394134574972118,
+      "learning_rate": 1.6074744890270033e-05,
+      "loss": 0.7879,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6270849273144606,
+      "grad_norm": 2.324105771909182,
+      "learning_rate": 1.6072776185841553e-05,
+      "loss": 0.8359,
+      "step": 4098
+    },
+    {
+      "epoch": 0.6272379495026779,
+      "grad_norm": 2.238004787754105,
+      "learning_rate": 1.6070807108460134e-05,
+      "loss": 0.7473,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6273909716908952,
+      "grad_norm": 2.4442647854498007,
+      "learning_rate": 1.60688376582467e-05,
+      "loss": 0.7507,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6275439938791124,
+      "grad_norm": 2.2586968036292965,
+      "learning_rate": 1.6066867835322205e-05,
+      "loss": 0.7854,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6276970160673297,
+      "grad_norm": 2.1196358650293647,
+      "learning_rate": 1.606489763980762e-05,
+      "loss": 0.6373,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6278500382555471,
+      "grad_norm": 2.211548528141079,
+      "learning_rate": 1.6062927071823944e-05,
+      "loss": 0.6638,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6280030604437643,
+      "grad_norm": 2.4010811990195124,
+      "learning_rate": 1.6060956131492203e-05,
+      "loss": 0.7585,
+      "step": 4104
+    },
+    {
+      "epoch": 0.6281560826319816,
+      "grad_norm": 2.5356718330834416,
+      "learning_rate": 1.605898481893343e-05,
+      "loss": 0.8063,
+      "step": 4105
+    },
+    {
+      "epoch": 0.628309104820199,
+      "grad_norm": 2.205202390214095,
+      "learning_rate": 1.6057013134268703e-05,
+      "loss": 0.7701,
+      "step": 4106
+    },
+    {
+      "epoch": 0.6284621270084162,
+      "grad_norm": 2.2819514096743228,
+      "learning_rate": 1.6055041077619094e-05,
+      "loss": 0.7002,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6286151491966335,
+      "grad_norm": 2.3022349264292745,
+      "learning_rate": 1.6053068649105735e-05,
+      "loss": 0.9141,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6287681713848507,
+      "grad_norm": 2.277035553809249,
+      "learning_rate": 1.6051095848849747e-05,
+      "loss": 0.8087,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6289211935730681,
+      "grad_norm": 2.4443791729171997,
+      "learning_rate": 1.604912267697229e-05,
+      "loss": 0.7876,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6290742157612854,
+      "grad_norm": 2.1725975257265784,
+      "learning_rate": 1.6047149133594546e-05,
+      "loss": 0.7552,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6292272379495026,
+      "grad_norm": 2.328562803605211,
+      "learning_rate": 1.6045175218837716e-05,
+      "loss": 0.7316,
+      "step": 4112
+    },
+    {
+      "epoch": 0.62938026013772,
+      "grad_norm": 2.247396465495871,
+      "learning_rate": 1.604320093282303e-05,
+      "loss": 0.6445,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6295332823259373,
+      "grad_norm": 2.173139920348392,
+      "learning_rate": 1.6041226275671727e-05,
+      "loss": 0.7427,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6296863045141545,
+      "grad_norm": 2.099285404139486,
+      "learning_rate": 1.603925124750509e-05,
+      "loss": 0.7111,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6298393267023719,
+      "grad_norm": 1.9917731708573834,
+      "learning_rate": 1.603727584844441e-05,
+      "loss": 0.6858,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6299923488905891,
+      "grad_norm": 2.215400393635466,
+      "learning_rate": 1.6035300078611e-05,
+      "loss": 0.7705,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6301453710788064,
+      "grad_norm": 2.0245212493288447,
+      "learning_rate": 1.6033323938126198e-05,
+      "loss": 0.6323,
+      "step": 4118
+    },
+    {
+      "epoch": 0.6302983932670237,
+      "grad_norm": 2.3197733532871885,
+      "learning_rate": 1.603134742711138e-05,
+      "loss": 0.7391,
+      "step": 4119
+    },
+    {
+      "epoch": 0.630451415455241,
+      "grad_norm": 2.4936860390189213,
+      "learning_rate": 1.6029370545687912e-05,
+      "loss": 0.8139,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6306044376434583,
+      "grad_norm": 2.181252151265491,
+      "learning_rate": 1.6027393293977217e-05,
+      "loss": 0.7221,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6307574598316756,
+      "grad_norm": 2.1530316724351755,
+      "learning_rate": 1.602541567210072e-05,
+      "loss": 0.7455,
+      "step": 4122
+    },
+    {
+      "epoch": 0.6309104820198929,
+      "grad_norm": 2.215998435857359,
+      "learning_rate": 1.6023437680179875e-05,
+      "loss": 0.7335,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6310635042081102,
+      "grad_norm": 2.4792874976399717,
+      "learning_rate": 1.6021459318336154e-05,
+      "loss": 0.7476,
+      "step": 4124
+    },
+    {
+      "epoch": 0.6312165263963274,
+      "grad_norm": 2.026954969477471,
+      "learning_rate": 1.6019480586691062e-05,
+      "loss": 0.5765,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6313695485845447,
+      "grad_norm": 2.595157669941229,
+      "learning_rate": 1.6017501485366125e-05,
+      "loss": 0.8033,
+      "step": 4126
+    },
+    {
+      "epoch": 0.6315225707727621,
+      "grad_norm": 2.3899715610542094,
+      "learning_rate": 1.6015522014482877e-05,
+      "loss": 0.7496,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6316755929609793,
+      "grad_norm": 2.234234590699503,
+      "learning_rate": 1.601354217416289e-05,
+      "loss": 0.7119,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6318286151491966,
+      "grad_norm": 2.359124165929824,
+      "learning_rate": 1.6011561964527748e-05,
+      "loss": 0.8503,
+      "step": 4129
+    },
+    {
+      "epoch": 0.631981637337414,
+      "grad_norm": 2.3720387035217527,
+      "learning_rate": 1.6009581385699076e-05,
+      "loss": 0.7733,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6321346595256312,
+      "grad_norm": 2.134605187970971,
+      "learning_rate": 1.6007600437798495e-05,
+      "loss": 0.7318,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6322876817138485,
+      "grad_norm": 2.129685324766079,
+      "learning_rate": 1.6005619120947672e-05,
+      "loss": 0.7498,
+      "step": 4132
+    },
+    {
+      "epoch": 0.6324407039020657,
+      "grad_norm": 2.3785700225702224,
+      "learning_rate": 1.6003637435268285e-05,
+      "loss": 0.7669,
+      "step": 4133
+    },
+    {
+      "epoch": 0.6325937260902831,
+      "grad_norm": 2.2794316845639355,
+      "learning_rate": 1.6001655380882036e-05,
+      "loss": 0.8118,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6327467482785004,
+      "grad_norm": 2.7039235062628566,
+      "learning_rate": 1.599967295791065e-05,
+      "loss": 0.8262,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6328997704667176,
+      "grad_norm": 2.3953376457651268,
+      "learning_rate": 1.599769016647588e-05,
+      "loss": 0.6636,
+      "step": 4136
+    },
+    {
+      "epoch": 0.633052792654935,
+      "grad_norm": 2.4461597091487954,
+      "learning_rate": 1.599570700669949e-05,
+      "loss": 0.8773,
+      "step": 4137
+    },
+    {
+      "epoch": 0.6332058148431523,
+      "grad_norm": 2.2968863879054484,
+      "learning_rate": 1.599372347870328e-05,
+      "loss": 0.7605,
+      "step": 4138
+    },
+    {
+      "epoch": 0.6333588370313695,
+      "grad_norm": 2.319391750065703,
+      "learning_rate": 1.5991739582609066e-05,
+      "loss": 0.6621,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6335118592195869,
+      "grad_norm": 2.0426803522722903,
+      "learning_rate": 1.598975531853868e-05,
+      "loss": 0.7313,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6336648814078041,
+      "grad_norm": 2.2009394418994286,
+      "learning_rate": 1.598777068661399e-05,
+      "loss": 0.8457,
+      "step": 4141
+    },
+    {
+      "epoch": 0.6338179035960214,
+      "grad_norm": 2.265228617457404,
+      "learning_rate": 1.5985785686956877e-05,
+      "loss": 0.7932,
+      "step": 4142
+    },
+    {
+      "epoch": 0.6339709257842387,
+      "grad_norm": 2.2768626036052595,
+      "learning_rate": 1.598380031968925e-05,
+      "loss": 0.6946,
+      "step": 4143
+    },
+    {
+      "epoch": 0.634123947972456,
+      "grad_norm": 2.2367807099521615,
+      "learning_rate": 1.5981814584933037e-05,
+      "loss": 0.7776,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6342769701606733,
+      "grad_norm": 2.174040440407306,
+      "learning_rate": 1.597982848281019e-05,
+      "loss": 0.6848,
+      "step": 4145
+    },
+    {
+      "epoch": 0.6344299923488906,
+      "grad_norm": 2.2403010646934223,
+      "learning_rate": 1.597784201344268e-05,
+      "loss": 0.7351,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6345830145371079,
+      "grad_norm": 2.313387309452502,
+      "learning_rate": 1.5975855176952505e-05,
+      "loss": 0.8039,
+      "step": 4147
+    },
+    {
+      "epoch": 0.6347360367253252,
+      "grad_norm": 2.2112838415716225,
+      "learning_rate": 1.5973867973461687e-05,
+      "loss": 0.7345,
+      "step": 4148
+    },
+    {
+      "epoch": 0.6348890589135424,
+      "grad_norm": 2.1155329059767105,
+      "learning_rate": 1.5971880403092267e-05,
+      "loss": 0.6863,
+      "step": 4149
+    },
+    {
+      "epoch": 0.6350420811017597,
+      "grad_norm": 2.3988266632172373,
+      "learning_rate": 1.5969892465966315e-05,
+      "loss": 0.7015,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6351951032899771,
+      "grad_norm": 2.5735735190206968,
+      "learning_rate": 1.5967904162205906e-05,
+      "loss": 0.7143,
+      "step": 4151
+    },
+    {
+      "epoch": 0.6353481254781943,
+      "grad_norm": 2.276882522860152,
+      "learning_rate": 1.5965915491933154e-05,
+      "loss": 0.7921,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6355011476664116,
+      "grad_norm": 2.3309317873483577,
+      "learning_rate": 1.5963926455270196e-05,
+      "loss": 0.7485,
+      "step": 4153
+    },
+    {
+      "epoch": 0.635654169854629,
+      "grad_norm": 2.1790982554114797,
+      "learning_rate": 1.5961937052339184e-05,
+      "loss": 0.5373,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6358071920428462,
+      "grad_norm": 2.666278766782337,
+      "learning_rate": 1.5959947283262296e-05,
+      "loss": 0.7133,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6359602142310635,
+      "grad_norm": 2.2551622266417404,
+      "learning_rate": 1.5957957148161727e-05,
+      "loss": 0.7513,
+      "step": 4156
+    },
+    {
+      "epoch": 0.6361132364192807,
+      "grad_norm": 2.6383466391094563,
+      "learning_rate": 1.5955966647159706e-05,
+      "loss": 0.8115,
+      "step": 4157
+    },
+    {
+      "epoch": 0.6362662586074981,
+      "grad_norm": 2.1419449882341124,
+      "learning_rate": 1.5953975780378466e-05,
+      "loss": 0.8361,
+      "step": 4158
+    },
+    {
+      "epoch": 0.6364192807957154,
+      "grad_norm": 2.574103577501881,
+      "learning_rate": 1.5951984547940286e-05,
+      "loss": 0.8792,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6365723029839326,
+      "grad_norm": 2.1063506557932348,
+      "learning_rate": 1.594999294996745e-05,
+      "loss": 0.7475,
+      "step": 4160
+    },
+    {
+      "epoch": 0.63672532517215,
+      "grad_norm": 2.321127303574338,
+      "learning_rate": 1.594800098658227e-05,
+      "loss": 0.7782,
+      "step": 4161
+    },
+    {
+      "epoch": 0.6368783473603673,
+      "grad_norm": 2.1374323400738633,
+      "learning_rate": 1.594600865790708e-05,
+      "loss": 0.6834,
+      "step": 4162
+    },
+    {
+      "epoch": 0.6370313695485845,
+      "grad_norm": 2.260939834196185,
+      "learning_rate": 1.594401596406424e-05,
+      "loss": 0.7039,
+      "step": 4163
+    },
+    {
+      "epoch": 0.6371843917368019,
+      "grad_norm": 2.0757294361561622,
+      "learning_rate": 1.5942022905176126e-05,
+      "loss": 0.6579,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6373374139250191,
+      "grad_norm": 2.437640815822106,
+      "learning_rate": 1.5940029481365137e-05,
+      "loss": 0.7543,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6374904361132364,
+      "grad_norm": 2.1674678167376484,
+      "learning_rate": 1.59380356927537e-05,
+      "loss": 0.6833,
+      "step": 4166
+    },
+    {
+      "epoch": 0.6376434583014537,
+      "grad_norm": 2.3065774043485363,
+      "learning_rate": 1.5936041539464266e-05,
+      "loss": 0.8161,
+      "step": 4167
+    },
+    {
+      "epoch": 0.637796480489671,
+      "grad_norm": 2.3020071837379663,
+      "learning_rate": 1.5934047021619295e-05,
+      "loss": 0.7168,
+      "step": 4168
+    },
+    {
+      "epoch": 0.6379495026778883,
+      "grad_norm": 2.379831436116885,
+      "learning_rate": 1.5932052139341285e-05,
+      "loss": 0.8322,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6381025248661056,
+      "grad_norm": 2.215752580256027,
+      "learning_rate": 1.5930056892752745e-05,
+      "loss": 0.789,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6382555470543229,
+      "grad_norm": 2.0602789500986307,
+      "learning_rate": 1.5928061281976215e-05,
+      "loss": 0.7685,
+      "step": 4171
+    },
+    {
+      "epoch": 0.6384085692425402,
+      "grad_norm": 2.249026928393553,
+      "learning_rate": 1.592606530713425e-05,
+      "loss": 0.6997,
+      "step": 4172
+    },
+    {
+      "epoch": 0.6385615914307574,
+      "grad_norm": 2.2625016939238534,
+      "learning_rate": 1.592406896834943e-05,
+      "loss": 0.8841,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6387146136189747,
+      "grad_norm": 2.7844803793506427,
+      "learning_rate": 1.592207226574436e-05,
+      "loss": 0.8634,
+      "step": 4174
+    },
+    {
+      "epoch": 0.6388676358071921,
+      "grad_norm": 2.2100873097986526,
+      "learning_rate": 1.5920075199441665e-05,
+      "loss": 0.7085,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6390206579954093,
+      "grad_norm": 2.339271594267563,
+      "learning_rate": 1.591807776956399e-05,
+      "loss": 0.6688,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6391736801836266,
+      "grad_norm": 2.1828890019441634,
+      "learning_rate": 1.591607997623401e-05,
+      "loss": 0.7627,
+      "step": 4177
+    },
+    {
+      "epoch": 0.639326702371844,
+      "grad_norm": 2.4192884203876304,
+      "learning_rate": 1.5914081819574415e-05,
+      "loss": 0.8232,
+      "step": 4178
+    },
+    {
+      "epoch": 0.6394797245600612,
+      "grad_norm": 2.235385282434219,
+      "learning_rate": 1.5912083299707924e-05,
+      "loss": 0.6711,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6396327467482785,
+      "grad_norm": 2.4216983966310823,
+      "learning_rate": 1.5910084416757264e-05,
+      "loss": 0.8284,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6397857689364957,
+      "grad_norm": 2.326093648518836,
+      "learning_rate": 1.59080851708452e-05,
+      "loss": 0.8516,
+      "step": 4181
+    },
+    {
+      "epoch": 0.6399387911247131,
+      "grad_norm": 2.4974683641820157,
+      "learning_rate": 1.5906085562094512e-05,
+      "loss": 0.7125,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6400918133129304,
+      "grad_norm": 2.2514434183700467,
+      "learning_rate": 1.590408559062801e-05,
+      "loss": 0.7538,
+      "step": 4183
+    },
+    {
+      "epoch": 0.6402448355011476,
+      "grad_norm": 2.2360998818579705,
+      "learning_rate": 1.5902085256568513e-05,
+      "loss": 0.7095,
+      "step": 4184
+    },
+    {
+      "epoch": 0.640397857689365,
+      "grad_norm": 2.4038858226459743,
+      "learning_rate": 1.5900084560038866e-05,
+      "loss": 0.8299,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6405508798775823,
+      "grad_norm": 2.1972831229466845,
+      "learning_rate": 1.589808350116195e-05,
+      "loss": 0.7232,
+      "step": 4186
+    },
+    {
+      "epoch": 0.6407039020657995,
+      "grad_norm": 2.482921431242963,
+      "learning_rate": 1.5896082080060652e-05,
+      "loss": 0.7147,
+      "step": 4187
+    },
+    {
+      "epoch": 0.6408569242540169,
+      "grad_norm": 2.2355154693159904,
+      "learning_rate": 1.5894080296857888e-05,
+      "loss": 0.7198,
+      "step": 4188
+    },
+    {
+      "epoch": 0.6410099464422341,
+      "grad_norm": 2.2241296107205524,
+      "learning_rate": 1.5892078151676594e-05,
+      "loss": 0.7909,
+      "step": 4189
+    },
+    {
+      "epoch": 0.6411629686304514,
+      "grad_norm": 2.605633584782082,
+      "learning_rate": 1.589007564463973e-05,
+      "loss": 0.8084,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6413159908186687,
+      "grad_norm": 2.454508772411002,
+      "learning_rate": 1.588807277587028e-05,
+      "loss": 0.8879,
+      "step": 4191
+    },
+    {
+      "epoch": 0.641469013006886,
+      "grad_norm": 2.4298870166317212,
+      "learning_rate": 1.5886069545491246e-05,
+      "loss": 0.6854,
+      "step": 4192
+    },
+    {
+      "epoch": 0.6416220351951033,
+      "grad_norm": 2.4906106059246733,
+      "learning_rate": 1.5884065953625657e-05,
+      "loss": 0.8755,
+      "step": 4193
+    },
+    {
+      "epoch": 0.6417750573833206,
+      "grad_norm": 2.1516796449874516,
+      "learning_rate": 1.5882062000396558e-05,
+      "loss": 0.7066,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6419280795715379,
+      "grad_norm": 2.493491767448356,
+      "learning_rate": 1.588005768592702e-05,
+      "loss": 0.7756,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6420811017597552,
+      "grad_norm": 2.6459928697516086,
+      "learning_rate": 1.587805301034014e-05,
+      "loss": 0.8202,
+      "step": 4196
+    },
+    {
+      "epoch": 0.6422341239479724,
+      "grad_norm": 2.4018200071295177,
+      "learning_rate": 1.587604797375902e-05,
+      "loss": 0.6751,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6423871461361897,
+      "grad_norm": 2.367904569625447,
+      "learning_rate": 1.5874042576306817e-05,
+      "loss": 0.7709,
+      "step": 4198
+    },
+    {
+      "epoch": 0.6425401683244071,
+      "grad_norm": 2.3896747175232327,
+      "learning_rate": 1.5872036818106672e-05,
+      "loss": 0.8643,
+      "step": 4199
+    },
+    {
+      "epoch": 0.6426931905126243,
+      "grad_norm": 2.7365450842405306,
+      "learning_rate": 1.5870030699281783e-05,
+      "loss": 0.709,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6428462127008416,
+      "grad_norm": 1.9490071277493157,
+      "learning_rate": 1.5868024219955337e-05,
+      "loss": 0.61,
+      "step": 4201
+    },
+    {
+      "epoch": 0.642999234889059,
+      "grad_norm": 2.2493824782464,
+      "learning_rate": 1.5866017380250573e-05,
+      "loss": 0.6998,
+      "step": 4202
+    },
+    {
+      "epoch": 0.6431522570772762,
+      "grad_norm": 2.078835252832914,
+      "learning_rate": 1.5864010180290732e-05,
+      "loss": 0.6524,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6433052792654935,
+      "grad_norm": 2.295853031190946,
+      "learning_rate": 1.5862002620199086e-05,
+      "loss": 0.7786,
+      "step": 4204
+    },
+    {
+      "epoch": 0.6434583014537107,
+      "grad_norm": 2.5791704393175507,
+      "learning_rate": 1.5859994700098925e-05,
+      "loss": 0.7799,
+      "step": 4205
+    },
+    {
+      "epoch": 0.6436113236419281,
+      "grad_norm": 2.288695724711449,
+      "learning_rate": 1.5857986420113568e-05,
+      "loss": 0.7069,
+      "step": 4206
+    },
+    {
+      "epoch": 0.6437643458301454,
+      "grad_norm": 2.0640414343256586,
+      "learning_rate": 1.5855977780366347e-05,
+      "loss": 0.6614,
+      "step": 4207
+    },
+    {
+      "epoch": 0.6439173680183626,
+      "grad_norm": 2.687001513306309,
+      "learning_rate": 1.5853968780980624e-05,
+      "loss": 0.6861,
+      "step": 4208
+    },
+    {
+      "epoch": 0.64407039020658,
+      "grad_norm": 2.3769771682509444,
+      "learning_rate": 1.585195942207977e-05,
+      "loss": 0.7767,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6442234123947972,
+      "grad_norm": 2.5169919724774026,
+      "learning_rate": 1.58499497037872e-05,
+      "loss": 0.7743,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6443764345830145,
+      "grad_norm": 2.3993993027725664,
+      "learning_rate": 1.5847939626226336e-05,
+      "loss": 0.7773,
+      "step": 4211
+    },
+    {
+      "epoch": 0.6445294567712319,
+      "grad_norm": 2.37377422024334,
+      "learning_rate": 1.5845929189520623e-05,
+      "loss": 0.8059,
+      "step": 4212
+    },
+    {
+      "epoch": 0.6446824789594491,
+      "grad_norm": 2.4955691858759743,
+      "learning_rate": 1.5843918393793523e-05,
+      "loss": 0.6907,
+      "step": 4213
+    },
+    {
+      "epoch": 0.6448355011476664,
+      "grad_norm": 2.4733738421996296,
+      "learning_rate": 1.5841907239168535e-05,
+      "loss": 0.909,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6449885233358837,
+      "grad_norm": 2.1763820948064976,
+      "learning_rate": 1.5839895725769173e-05,
+      "loss": 0.7843,
+      "step": 4215
+    },
+    {
+      "epoch": 0.645141545524101,
+      "grad_norm": 2.1398197268300114,
+      "learning_rate": 1.5837883853718964e-05,
+      "loss": 0.6951,
+      "step": 4216
+    },
+    {
+      "epoch": 0.6452945677123183,
+      "grad_norm": 2.5145027467966528,
+      "learning_rate": 1.583587162314147e-05,
+      "loss": 0.8311,
+      "step": 4217
+    },
+    {
+      "epoch": 0.6454475899005355,
+      "grad_norm": 2.4465579575147656,
+      "learning_rate": 1.5833859034160275e-05,
+      "loss": 0.6852,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6456006120887529,
+      "grad_norm": 2.3378074319427453,
+      "learning_rate": 1.583184608689897e-05,
+      "loss": 0.7895,
+      "step": 4219
+    },
+    {
+      "epoch": 0.6457536342769702,
+      "grad_norm": 1.9247135984095447,
+      "learning_rate": 1.5829832781481186e-05,
+      "loss": 0.6891,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6459066564651874,
+      "grad_norm": 2.2282892978052735,
+      "learning_rate": 1.582781911803056e-05,
+      "loss": 0.736,
+      "step": 4221
+    },
+    {
+      "epoch": 0.6460596786534047,
+      "grad_norm": 2.285669724890519,
+      "learning_rate": 1.5825805096670766e-05,
+      "loss": 0.7469,
+      "step": 4222
+    },
+    {
+      "epoch": 0.6462127008416221,
+      "grad_norm": 2.581987099916605,
+      "learning_rate": 1.5823790717525487e-05,
+      "loss": 0.8131,
+      "step": 4223
+    },
+    {
+      "epoch": 0.6463657230298393,
+      "grad_norm": 2.00862726818773,
+      "learning_rate": 1.5821775980718443e-05,
+      "loss": 0.7074,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6465187452180566,
+      "grad_norm": 2.3151539318728314,
+      "learning_rate": 1.581976088637336e-05,
+      "loss": 0.711,
+      "step": 4225
+    },
+    {
+      "epoch": 0.6466717674062739,
+      "grad_norm": 2.159998848069135,
+      "learning_rate": 1.581774543461399e-05,
+      "loss": 0.7569,
+      "step": 4226
+    },
+    {
+      "epoch": 0.6468247895944912,
+      "grad_norm": 2.2315436312344916,
+      "learning_rate": 1.5815729625564116e-05,
+      "loss": 0.6982,
+      "step": 4227
+    },
+    {
+      "epoch": 0.6469778117827085,
+      "grad_norm": 2.3296082075550983,
+      "learning_rate": 1.5813713459347532e-05,
+      "loss": 0.6977,
+      "step": 4228
+    },
+    {
+      "epoch": 0.6471308339709257,
+      "grad_norm": 2.3650900215329758,
+      "learning_rate": 1.5811696936088066e-05,
+      "loss": 0.7675,
+      "step": 4229
+    },
+    {
+      "epoch": 0.6472838561591431,
+      "grad_norm": 2.0510993157963515,
+      "learning_rate": 1.5809680055909552e-05,
+      "loss": 0.5907,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6474368783473604,
+      "grad_norm": 2.392826682725871,
+      "learning_rate": 1.580766281893586e-05,
+      "loss": 0.756,
+      "step": 4231
+    },
+    {
+      "epoch": 0.6475899005355776,
+      "grad_norm": 2.159486959255893,
+      "learning_rate": 1.5805645225290872e-05,
+      "loss": 0.7136,
+      "step": 4232
+    },
+    {
+      "epoch": 0.647742922723795,
+      "grad_norm": 2.13632556895698,
+      "learning_rate": 1.58036272750985e-05,
+      "loss": 0.7156,
+      "step": 4233
+    },
+    {
+      "epoch": 0.6478959449120122,
+      "grad_norm": 2.2967568833902283,
+      "learning_rate": 1.5801608968482676e-05,
+      "loss": 0.8513,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6480489671002295,
+      "grad_norm": 2.1817154806929016,
+      "learning_rate": 1.5799590305567345e-05,
+      "loss": 0.7536,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6482019892884469,
+      "grad_norm": 2.0809234386044073,
+      "learning_rate": 1.5797571286476484e-05,
+      "loss": 0.7957,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6483550114766641,
+      "grad_norm": 2.6517476323715115,
+      "learning_rate": 1.5795551911334096e-05,
+      "loss": 0.7822,
+      "step": 4237
+    },
+    {
+      "epoch": 0.6485080336648814,
+      "grad_norm": 2.350343075077592,
+      "learning_rate": 1.579353218026419e-05,
+      "loss": 0.7427,
+      "step": 4238
+    },
+    {
+      "epoch": 0.6486610558530987,
+      "grad_norm": 2.1243287340078383,
+      "learning_rate": 1.579151209339081e-05,
+      "loss": 0.8415,
+      "step": 4239
+    },
+    {
+      "epoch": 0.648814078041316,
+      "grad_norm": 1.989695986982748,
+      "learning_rate": 1.5789491650838013e-05,
+      "loss": 0.733,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6489671002295333,
+      "grad_norm": 2.215200133687346,
+      "learning_rate": 1.5787470852729886e-05,
+      "loss": 0.7286,
+      "step": 4241
+    },
+    {
+      "epoch": 0.6491201224177505,
+      "grad_norm": 2.002861466308425,
+      "learning_rate": 1.5785449699190533e-05,
+      "loss": 0.708,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6492731446059679,
+      "grad_norm": 2.198265404670774,
+      "learning_rate": 1.578342819034408e-05,
+      "loss": 0.6877,
+      "step": 4243
+    },
+    {
+      "epoch": 0.6494261667941852,
+      "grad_norm": 2.2444980427910064,
+      "learning_rate": 1.578140632631468e-05,
+      "loss": 0.7127,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6495791889824024,
+      "grad_norm": 2.3743717173706855,
+      "learning_rate": 1.57793841072265e-05,
+      "loss": 0.7277,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6497322111706197,
+      "grad_norm": 2.094820997637052,
+      "learning_rate": 1.5777361533203733e-05,
+      "loss": 0.6453,
+      "step": 4246
+    },
+    {
+      "epoch": 0.6498852333588371,
+      "grad_norm": 2.359769995436231,
+      "learning_rate": 1.577533860437059e-05,
+      "loss": 0.7156,
+      "step": 4247
+    },
+    {
+      "epoch": 0.6500382555470543,
+      "grad_norm": 2.4827444410396127,
+      "learning_rate": 1.577331532085131e-05,
+      "loss": 0.8702,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6501912777352716,
+      "grad_norm": 2.392016169668508,
+      "learning_rate": 1.5771291682770156e-05,
+      "loss": 0.8331,
+      "step": 4249
+    },
+    {
+      "epoch": 0.6503442999234889,
+      "grad_norm": 2.3940592611494895,
+      "learning_rate": 1.57692676902514e-05,
+      "loss": 0.7551,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6504973221117062,
+      "grad_norm": 2.3839301346084025,
+      "learning_rate": 1.5767243343419342e-05,
+      "loss": 0.8463,
+      "step": 4251
+    },
+    {
+      "epoch": 0.6506503442999235,
+      "grad_norm": 2.4202585155038787,
+      "learning_rate": 1.5765218642398314e-05,
+      "loss": 0.7652,
+      "step": 4252
+    },
+    {
+      "epoch": 0.6508033664881407,
+      "grad_norm": 2.1053361807836937,
+      "learning_rate": 1.5763193587312655e-05,
+      "loss": 0.6554,
+      "step": 4253
+    },
+    {
+      "epoch": 0.6509563886763581,
+      "grad_norm": 1.8735355733426593,
+      "learning_rate": 1.5761168178286727e-05,
+      "loss": 0.691,
+      "step": 4254
+    },
+    {
+      "epoch": 0.6511094108645754,
+      "grad_norm": 2.224692423547609,
+      "learning_rate": 1.5759142415444925e-05,
+      "loss": 0.696,
+      "step": 4255
+    },
+    {
+      "epoch": 0.6512624330527926,
+      "grad_norm": 2.0753222236814683,
+      "learning_rate": 1.575711629891166e-05,
+      "loss": 0.7598,
+      "step": 4256
+    },
+    {
+      "epoch": 0.65141545524101,
+      "grad_norm": 2.2697210222802306,
+      "learning_rate": 1.5755089828811362e-05,
+      "loss": 0.6358,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6515684774292272,
+      "grad_norm": 2.3751008200480275,
+      "learning_rate": 1.5753063005268483e-05,
+      "loss": 0.7119,
+      "step": 4258
+    },
+    {
+      "epoch": 0.6517214996174445,
+      "grad_norm": 2.124262512388635,
+      "learning_rate": 1.5751035828407494e-05,
+      "loss": 0.6876,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6518745218056619,
+      "grad_norm": 2.2054070929703276,
+      "learning_rate": 1.57490082983529e-05,
+      "loss": 0.7775,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6520275439938791,
+      "grad_norm": 2.077240839261722,
+      "learning_rate": 1.5746980415229217e-05,
+      "loss": 0.6499,
+      "step": 4261
+    },
+    {
+      "epoch": 0.6521805661820964,
+      "grad_norm": 2.2208562054207053,
+      "learning_rate": 1.5744952179160985e-05,
+      "loss": 0.8069,
+      "step": 4262
+    },
+    {
+      "epoch": 0.6523335883703137,
+      "grad_norm": 2.135417668181956,
+      "learning_rate": 1.5742923590272762e-05,
+      "loss": 0.7635,
+      "step": 4263
+    },
+    {
+      "epoch": 0.652486610558531,
+      "grad_norm": 2.2710496430797926,
+      "learning_rate": 1.5740894648689138e-05,
+      "loss": 0.8612,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6526396327467483,
+      "grad_norm": 2.097980111255635,
+      "learning_rate": 1.5738865354534714e-05,
+      "loss": 0.7313,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6527926549349655,
+      "grad_norm": 2.3366510066682835,
+      "learning_rate": 1.573683570793412e-05,
+      "loss": 0.7146,
+      "step": 4266
+    },
+    {
+      "epoch": 0.6529456771231829,
+      "grad_norm": 2.162653896948852,
+      "learning_rate": 1.5734805709012e-05,
+      "loss": 0.7029,
+      "step": 4267
+    },
+    {
+      "epoch": 0.6530986993114002,
+      "grad_norm": 2.3656466257792323,
+      "learning_rate": 1.5732775357893024e-05,
+      "loss": 0.8957,
+      "step": 4268
+    },
+    {
+      "epoch": 0.6532517214996174,
+      "grad_norm": 2.098111046017186,
+      "learning_rate": 1.573074465470189e-05,
+      "loss": 0.6822,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6534047436878347,
+      "grad_norm": 2.4117741397672408,
+      "learning_rate": 1.5728713599563306e-05,
+      "loss": 0.7866,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6535577658760521,
+      "grad_norm": 2.174338870646714,
+      "learning_rate": 1.5726682192602015e-05,
+      "loss": 0.8173,
+      "step": 4271
+    },
+    {
+      "epoch": 0.6537107880642693,
+      "grad_norm": 2.186513283043854,
+      "learning_rate": 1.572465043394276e-05,
+      "loss": 0.776,
+      "step": 4272
+    },
+    {
+      "epoch": 0.6538638102524866,
+      "grad_norm": 2.4290604461156273,
+      "learning_rate": 1.572261832371033e-05,
+      "loss": 0.8385,
+      "step": 4273
+    },
+    {
+      "epoch": 0.6540168324407039,
+      "grad_norm": 2.204819777026553,
+      "learning_rate": 1.5720585862029522e-05,
+      "loss": 0.7703,
+      "step": 4274
+    },
+    {
+      "epoch": 0.6541698546289212,
+      "grad_norm": 2.3510224710248004,
+      "learning_rate": 1.571855304902516e-05,
+      "loss": 0.7174,
+      "step": 4275
+    },
+    {
+      "epoch": 0.6543228768171385,
+      "grad_norm": 2.5013178502584084,
+      "learning_rate": 1.571651988482208e-05,
+      "loss": 0.7605,
+      "step": 4276
+    },
+    {
+      "epoch": 0.6544758990053557,
+      "grad_norm": 2.4612060334374517,
+      "learning_rate": 1.5714486369545152e-05,
+      "loss": 0.7707,
+      "step": 4277
+    },
+    {
+      "epoch": 0.6546289211935731,
+      "grad_norm": 2.2725366255063655,
+      "learning_rate": 1.571245250331926e-05,
+      "loss": 0.8654,
+      "step": 4278
+    },
+    {
+      "epoch": 0.6547819433817904,
+      "grad_norm": 2.216534816275824,
+      "learning_rate": 1.5710418286269313e-05,
+      "loss": 0.68,
+      "step": 4279
+    },
+    {
+      "epoch": 0.6549349655700076,
+      "grad_norm": 2.2998627589268352,
+      "learning_rate": 1.570838371852024e-05,
+      "loss": 0.7354,
+      "step": 4280
+    },
+    {
+      "epoch": 0.655087987758225,
+      "grad_norm": 2.559430498766767,
+      "learning_rate": 1.5706348800196993e-05,
+      "loss": 0.7744,
+      "step": 4281
+    },
+    {
+      "epoch": 0.6552410099464422,
+      "grad_norm": 1.9862496757382806,
+      "learning_rate": 1.570431353142454e-05,
+      "loss": 0.674,
+      "step": 4282
+    },
+    {
+      "epoch": 0.6553940321346595,
+      "grad_norm": 2.315587355397246,
+      "learning_rate": 1.570227791232788e-05,
+      "loss": 0.7833,
+      "step": 4283
+    },
+    {
+      "epoch": 0.6555470543228769,
+      "grad_norm": 2.175044828127138,
+      "learning_rate": 1.570024194303202e-05,
+      "loss": 0.6864,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6557000765110941,
+      "grad_norm": 2.037261450774724,
+      "learning_rate": 1.5698205623662013e-05,
+      "loss": 0.619,
+      "step": 4285
+    },
+    {
+      "epoch": 0.6558530986993114,
+      "grad_norm": 2.1956095246913505,
+      "learning_rate": 1.56961689543429e-05,
+      "loss": 0.6983,
+      "step": 4286
+    },
+    {
+      "epoch": 0.6560061208875287,
+      "grad_norm": 2.5219192285962375,
+      "learning_rate": 1.569413193519977e-05,
+      "loss": 0.7786,
+      "step": 4287
+    },
+    {
+      "epoch": 0.656159143075746,
+      "grad_norm": 1.9680199150456783,
+      "learning_rate": 1.5692094566357723e-05,
+      "loss": 0.6138,
+      "step": 4288
+    },
+    {
+      "epoch": 0.6563121652639633,
+      "grad_norm": 2.791679205377205,
+      "learning_rate": 1.569005684794188e-05,
+      "loss": 0.7804,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6564651874521805,
+      "grad_norm": 2.441683768664178,
+      "learning_rate": 1.5688018780077387e-05,
+      "loss": 0.7648,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6566182096403979,
+      "grad_norm": 2.3593952030509393,
+      "learning_rate": 1.5685980362889407e-05,
+      "loss": 0.7983,
+      "step": 4291
+    },
+    {
+      "epoch": 0.6567712318286152,
+      "grad_norm": 2.44463161656858,
+      "learning_rate": 1.5683941596503132e-05,
+      "loss": 0.8427,
+      "step": 4292
+    },
+    {
+      "epoch": 0.6569242540168324,
+      "grad_norm": 2.477271920848612,
+      "learning_rate": 1.568190248104377e-05,
+      "loss": 0.709,
+      "step": 4293
+    },
+    {
+      "epoch": 0.6570772762050497,
+      "grad_norm": 2.0726834727509247,
+      "learning_rate": 1.567986301663654e-05,
+      "loss": 0.7644,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6572302983932671,
+      "grad_norm": 2.4549904996769523,
+      "learning_rate": 1.567782320340671e-05,
+      "loss": 0.6868,
+      "step": 4295
+    },
+    {
+      "epoch": 0.6573833205814843,
+      "grad_norm": 2.133291763535228,
+      "learning_rate": 1.5675783041479542e-05,
+      "loss": 0.6548,
+      "step": 4296
+    },
+    {
+      "epoch": 0.6575363427697016,
+      "grad_norm": 2.69944862833919,
+      "learning_rate": 1.5673742530980337e-05,
+      "loss": 0.7521,
+      "step": 4297
+    },
+    {
+      "epoch": 0.6576893649579189,
+      "grad_norm": 2.318466203294317,
+      "learning_rate": 1.5671701672034406e-05,
+      "loss": 0.6864,
+      "step": 4298
+    },
+    {
+      "epoch": 0.6578423871461362,
+      "grad_norm": 2.4267845713840233,
+      "learning_rate": 1.5669660464767087e-05,
+      "loss": 0.8534,
+      "step": 4299
+    },
+    {
+      "epoch": 0.6579954093343535,
+      "grad_norm": 2.382318083873871,
+      "learning_rate": 1.5667618909303738e-05,
+      "loss": 0.7832,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6581484315225707,
+      "grad_norm": 2.2591848691994003,
+      "learning_rate": 1.566557700576974e-05,
+      "loss": 0.7867,
+      "step": 4301
+    },
+    {
+      "epoch": 0.6583014537107881,
+      "grad_norm": 2.1813851643328905,
+      "learning_rate": 1.5663534754290496e-05,
+      "loss": 0.8173,
+      "step": 4302
+    },
+    {
+      "epoch": 0.6584544758990054,
+      "grad_norm": 2.3371976725354635,
+      "learning_rate": 1.5661492154991424e-05,
+      "loss": 0.6822,
+      "step": 4303
+    },
+    {
+      "epoch": 0.6586074980872226,
+      "grad_norm": 2.4281104499333086,
+      "learning_rate": 1.5659449207997975e-05,
+      "loss": 0.7563,
+      "step": 4304
+    },
+    {
+      "epoch": 0.65876052027544,
+      "grad_norm": 2.332443416486321,
+      "learning_rate": 1.5657405913435608e-05,
+      "loss": 0.7429,
+      "step": 4305
+    },
+    {
+      "epoch": 0.6589135424636572,
+      "grad_norm": 2.1794383230938497,
+      "learning_rate": 1.565536227142981e-05,
+      "loss": 0.7894,
+      "step": 4306
+    },
+    {
+      "epoch": 0.6590665646518745,
+      "grad_norm": 2.2246346257704994,
+      "learning_rate": 1.5653318282106096e-05,
+      "loss": 0.7179,
+      "step": 4307
+    },
+    {
+      "epoch": 0.6592195868400919,
+      "grad_norm": 2.278677670858125,
+      "learning_rate": 1.5651273945589984e-05,
+      "loss": 0.8088,
+      "step": 4308
+    },
+    {
+      "epoch": 0.6593726090283091,
+      "grad_norm": 2.3163101708985665,
+      "learning_rate": 1.5649229262007035e-05,
+      "loss": 0.7675,
+      "step": 4309
+    },
+    {
+      "epoch": 0.6595256312165264,
+      "grad_norm": 2.1382456702220924,
+      "learning_rate": 1.5647184231482816e-05,
+      "loss": 0.6105,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6596786534047437,
+      "grad_norm": 2.6339501865403645,
+      "learning_rate": 1.5645138854142926e-05,
+      "loss": 0.7263,
+      "step": 4311
+    },
+    {
+      "epoch": 0.659831675592961,
+      "grad_norm": 2.622754234753976,
+      "learning_rate": 1.564309313011297e-05,
+      "loss": 0.8526,
+      "step": 4312
+    },
+    {
+      "epoch": 0.6599846977811783,
+      "grad_norm": 2.3998143541848145,
+      "learning_rate": 1.564104705951859e-05,
+      "loss": 0.8304,
+      "step": 4313
+    },
+    {
+      "epoch": 0.6601377199693955,
+      "grad_norm": 2.197439826841551,
+      "learning_rate": 1.563900064248544e-05,
+      "loss": 0.7395,
+      "step": 4314
+    },
+    {
+      "epoch": 0.6602907421576129,
+      "grad_norm": 2.3044007948624072,
+      "learning_rate": 1.5636953879139204e-05,
+      "loss": 0.7928,
+      "step": 4315
+    },
+    {
+      "epoch": 0.6604437643458302,
+      "grad_norm": 2.2479485395298666,
+      "learning_rate": 1.563490676960558e-05,
+      "loss": 0.791,
+      "step": 4316
+    },
+    {
+      "epoch": 0.6605967865340474,
+      "grad_norm": 2.3513886944279325,
+      "learning_rate": 1.563285931401028e-05,
+      "loss": 0.7341,
+      "step": 4317
+    },
+    {
+      "epoch": 0.6607498087222647,
+      "grad_norm": 2.177828319997214,
+      "learning_rate": 1.563081151247906e-05,
+      "loss": 0.7671,
+      "step": 4318
+    },
+    {
+      "epoch": 0.6609028309104821,
+      "grad_norm": 2.192816682090379,
+      "learning_rate": 1.562876336513768e-05,
+      "loss": 0.83,
+      "step": 4319
+    },
+    {
+      "epoch": 0.6610558530986993,
+      "grad_norm": 2.0953415645877778,
+      "learning_rate": 1.5626714872111915e-05,
+      "loss": 0.6911,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6612088752869166,
+      "grad_norm": 2.2103399446325276,
+      "learning_rate": 1.5624666033527577e-05,
+      "loss": 0.7532,
+      "step": 4321
+    },
+    {
+      "epoch": 0.6613618974751339,
+      "grad_norm": 2.4237262824412125,
+      "learning_rate": 1.5622616849510497e-05,
+      "loss": 0.7964,
+      "step": 4322
+    },
+    {
+      "epoch": 0.6615149196633512,
+      "grad_norm": 2.100080910063574,
+      "learning_rate": 1.5620567320186522e-05,
+      "loss": 0.5986,
+      "step": 4323
+    },
+    {
+      "epoch": 0.6616679418515685,
+      "grad_norm": 2.2211004196585367,
+      "learning_rate": 1.561851744568152e-05,
+      "loss": 0.7725,
+      "step": 4324
+    },
+    {
+      "epoch": 0.6618209640397857,
+      "grad_norm": 2.2274632487007358,
+      "learning_rate": 1.561646722612138e-05,
+      "loss": 0.7935,
+      "step": 4325
+    },
+    {
+      "epoch": 0.6619739862280031,
+      "grad_norm": 2.0708948012053563,
+      "learning_rate": 1.5614416661632015e-05,
+      "loss": 0.6594,
+      "step": 4326
+    },
+    {
+      "epoch": 0.6621270084162203,
+      "grad_norm": 2.2197858350008843,
+      "learning_rate": 1.5612365752339364e-05,
+      "loss": 0.766,
+      "step": 4327
+    },
+    {
+      "epoch": 0.6622800306044376,
+      "grad_norm": 2.220248640125984,
+      "learning_rate": 1.5610314498369372e-05,
+      "loss": 0.7767,
+      "step": 4328
+    },
+    {
+      "epoch": 0.662433052792655,
+      "grad_norm": 2.380043383610386,
+      "learning_rate": 1.560826289984802e-05,
+      "loss": 0.7342,
+      "step": 4329
+    },
+    {
+      "epoch": 0.6625860749808722,
+      "grad_norm": 2.6236063675363046,
+      "learning_rate": 1.5606210956901303e-05,
+      "loss": 0.7274,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6627390971690895,
+      "grad_norm": 2.148371880434111,
+      "learning_rate": 1.560415866965524e-05,
+      "loss": 0.6713,
+      "step": 4331
+    },
+    {
+      "epoch": 0.6628921193573069,
+      "grad_norm": 2.0025618604021607,
+      "learning_rate": 1.5602106038235874e-05,
+      "loss": 0.6365,
+      "step": 4332
+    },
+    {
+      "epoch": 0.6630451415455241,
+      "grad_norm": 2.21401912073108,
+      "learning_rate": 1.5600053062769253e-05,
+      "loss": 0.6911,
+      "step": 4333
+    },
+    {
+      "epoch": 0.6631981637337414,
+      "grad_norm": 2.228557608606604,
+      "learning_rate": 1.559799974338147e-05,
+      "loss": 0.6889,
+      "step": 4334
+    },
+    {
+      "epoch": 0.6633511859219586,
+      "grad_norm": 2.438672814007364,
+      "learning_rate": 1.5595946080198624e-05,
+      "loss": 0.7678,
+      "step": 4335
+    },
+    {
+      "epoch": 0.663504208110176,
+      "grad_norm": 2.2428273489360913,
+      "learning_rate": 1.5593892073346837e-05,
+      "loss": 0.7243,
+      "step": 4336
+    },
+    {
+      "epoch": 0.6636572302983933,
+      "grad_norm": 2.4099325849538364,
+      "learning_rate": 1.5591837722952253e-05,
+      "loss": 0.8096,
+      "step": 4337
+    },
+    {
+      "epoch": 0.6638102524866105,
+      "grad_norm": 2.19582365303818,
+      "learning_rate": 1.5589783029141038e-05,
+      "loss": 0.7445,
+      "step": 4338
+    },
+    {
+      "epoch": 0.6639632746748279,
+      "grad_norm": 2.2901855214253173,
+      "learning_rate": 1.558772799203938e-05,
+      "loss": 0.7788,
+      "step": 4339
+    },
+    {
+      "epoch": 0.6641162968630452,
+      "grad_norm": 2.093622043015938,
+      "learning_rate": 1.558567261177349e-05,
+      "loss": 0.8354,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6642693190512624,
+      "grad_norm": 2.233333883954918,
+      "learning_rate": 1.5583616888469594e-05,
+      "loss": 0.8035,
+      "step": 4341
+    },
+    {
+      "epoch": 0.6644223412394797,
+      "grad_norm": 2.027649735934898,
+      "learning_rate": 1.5581560822253938e-05,
+      "loss": 0.6271,
+      "step": 4342
+    },
+    {
+      "epoch": 0.664575363427697,
+      "grad_norm": 2.3410794486831534,
+      "learning_rate": 1.55795044132528e-05,
+      "loss": 0.6748,
+      "step": 4343
+    },
+    {
+      "epoch": 0.6647283856159143,
+      "grad_norm": 2.034670891220707,
+      "learning_rate": 1.5577447661592467e-05,
+      "loss": 0.6798,
+      "step": 4344
+    },
+    {
+      "epoch": 0.6648814078041316,
+      "grad_norm": 2.267729163658624,
+      "learning_rate": 1.5575390567399253e-05,
+      "loss": 0.7299,
+      "step": 4345
+    },
+    {
+      "epoch": 0.6650344299923489,
+      "grad_norm": 2.3485035033426493,
+      "learning_rate": 1.5573333130799494e-05,
+      "loss": 0.7796,
+      "step": 4346
+    },
+    {
+      "epoch": 0.6651874521805662,
+      "grad_norm": 2.3608093796256657,
+      "learning_rate": 1.5571275351919543e-05,
+      "loss": 1.0151,
+      "step": 4347
+    },
+    {
+      "epoch": 0.6653404743687835,
+      "grad_norm": 2.258946043310113,
+      "learning_rate": 1.556921723088578e-05,
+      "loss": 0.668,
+      "step": 4348
+    },
+    {
+      "epoch": 0.6654934965570007,
+      "grad_norm": 2.4385006256619626,
+      "learning_rate": 1.5567158767824603e-05,
+      "loss": 0.7649,
+      "step": 4349
+    },
+    {
+      "epoch": 0.6656465187452181,
+      "grad_norm": 2.187731470321015,
+      "learning_rate": 1.556509996286242e-05,
+      "loss": 0.74,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6657995409334353,
+      "grad_norm": 2.136213298646861,
+      "learning_rate": 1.5563040816125683e-05,
+      "loss": 0.756,
+      "step": 4351
+    },
+    {
+      "epoch": 0.6659525631216526,
+      "grad_norm": 2.464238595477538,
+      "learning_rate": 1.5560981327740846e-05,
+      "loss": 0.7432,
+      "step": 4352
+    },
+    {
+      "epoch": 0.66610558530987,
+      "grad_norm": 2.138263083825212,
+      "learning_rate": 1.5558921497834387e-05,
+      "loss": 0.7623,
+      "step": 4353
+    },
+    {
+      "epoch": 0.6662586074980872,
+      "grad_norm": 2.2159031648485157,
+      "learning_rate": 1.5556861326532817e-05,
+      "loss": 0.6481,
+      "step": 4354
+    },
+    {
+      "epoch": 0.6664116296863045,
+      "grad_norm": 2.3079101636175405,
+      "learning_rate": 1.5554800813962652e-05,
+      "loss": 0.7337,
+      "step": 4355
+    },
+    {
+      "epoch": 0.6665646518745219,
+      "grad_norm": 2.3557552648531095,
+      "learning_rate": 1.555273996025044e-05,
+      "loss": 0.7614,
+      "step": 4356
+    },
+    {
+      "epoch": 0.6667176740627391,
+      "grad_norm": 2.3975045855476234,
+      "learning_rate": 1.5550678765522743e-05,
+      "loss": 0.8563,
+      "step": 4357
+    },
+    {
+      "epoch": 0.6668706962509564,
+      "grad_norm": 2.3698126148175263,
+      "learning_rate": 1.554861722990615e-05,
+      "loss": 0.6551,
+      "step": 4358
+    },
+    {
+      "epoch": 0.6670237184391736,
+      "grad_norm": 1.9109474500910666,
+      "learning_rate": 1.5546555353527268e-05,
+      "loss": 0.5765,
+      "step": 4359
+    },
+    {
+      "epoch": 0.667176740627391,
+      "grad_norm": 2.347918132400794,
+      "learning_rate": 1.554449313651272e-05,
+      "loss": 0.8157,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6673297628156083,
+      "grad_norm": 2.1667966924767517,
+      "learning_rate": 1.554243057898916e-05,
+      "loss": 0.7399,
+      "step": 4361
+    },
+    {
+      "epoch": 0.6674827850038255,
+      "grad_norm": 2.086540955838038,
+      "learning_rate": 1.5540367681083256e-05,
+      "loss": 0.6853,
+      "step": 4362
+    },
+    {
+      "epoch": 0.6676358071920429,
+      "grad_norm": 2.132291609887321,
+      "learning_rate": 1.5538304442921694e-05,
+      "loss": 0.7374,
+      "step": 4363
+    },
+    {
+      "epoch": 0.6677888293802602,
+      "grad_norm": 2.363328018543426,
+      "learning_rate": 1.5536240864631197e-05,
+      "loss": 0.824,
+      "step": 4364
+    },
+    {
+      "epoch": 0.6679418515684774,
+      "grad_norm": 2.3075046294927977,
+      "learning_rate": 1.553417694633849e-05,
+      "loss": 0.7507,
+      "step": 4365
+    },
+    {
+      "epoch": 0.6680948737566947,
+      "grad_norm": 2.5578916953080086,
+      "learning_rate": 1.553211268817032e-05,
+      "loss": 0.9281,
+      "step": 4366
+    },
+    {
+      "epoch": 0.668247895944912,
+      "grad_norm": 2.0753350483746233,
+      "learning_rate": 1.5530048090253475e-05,
+      "loss": 0.7547,
+      "step": 4367
+    },
+    {
+      "epoch": 0.6684009181331293,
+      "grad_norm": 2.069627851881168,
+      "learning_rate": 1.552798315271474e-05,
+      "loss": 0.7056,
+      "step": 4368
+    },
+    {
+      "epoch": 0.6685539403213466,
+      "grad_norm": 2.3010025507959573,
+      "learning_rate": 1.5525917875680936e-05,
+      "loss": 0.6966,
+      "step": 4369
+    },
+    {
+      "epoch": 0.6687069625095639,
+      "grad_norm": 2.5111567762780154,
+      "learning_rate": 1.55238522592789e-05,
+      "loss": 0.7832,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6688599846977812,
+      "grad_norm": 1.9618299176729825,
+      "learning_rate": 1.5521786303635483e-05,
+      "loss": 0.6293,
+      "step": 4371
+    },
+    {
+      "epoch": 0.6690130068859985,
+      "grad_norm": 2.225950898436061,
+      "learning_rate": 1.5519720008877567e-05,
+      "loss": 0.7209,
+      "step": 4372
+    },
+    {
+      "epoch": 0.6691660290742157,
+      "grad_norm": 1.992624936802358,
+      "learning_rate": 1.5517653375132055e-05,
+      "loss": 0.7138,
+      "step": 4373
+    },
+    {
+      "epoch": 0.6693190512624331,
+      "grad_norm": 1.8752156297552323,
+      "learning_rate": 1.551558640252586e-05,
+      "loss": 0.6508,
+      "step": 4374
+    },
+    {
+      "epoch": 0.6694720734506503,
+      "grad_norm": 2.3301325690600945,
+      "learning_rate": 1.5513519091185934e-05,
+      "loss": 0.5898,
+      "step": 4375
+    },
+    {
+      "epoch": 0.6696250956388676,
+      "grad_norm": 2.3208751688365576,
+      "learning_rate": 1.5511451441239227e-05,
+      "loss": 0.6817,
+      "step": 4376
+    },
+    {
+      "epoch": 0.669778117827085,
+      "grad_norm": 2.3173083606568396,
+      "learning_rate": 1.5509383452812725e-05,
+      "loss": 0.5693,
+      "step": 4377
+    },
+    {
+      "epoch": 0.6699311400153022,
+      "grad_norm": 2.383092745146042,
+      "learning_rate": 1.5507315126033435e-05,
+      "loss": 0.6256,
+      "step": 4378
+    },
+    {
+      "epoch": 0.6700841622035195,
+      "grad_norm": 1.8645480695198386,
+      "learning_rate": 1.550524646102838e-05,
+      "loss": 0.526,
+      "step": 4379
+    },
+    {
+      "epoch": 0.6702371843917369,
+      "grad_norm": 2.4363367750243805,
+      "learning_rate": 1.5503177457924597e-05,
+      "loss": 0.7734,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6703902065799541,
+      "grad_norm": 2.5537622593414415,
+      "learning_rate": 1.5501108116849162e-05,
+      "loss": 0.7684,
+      "step": 4381
+    },
+    {
+      "epoch": 0.6705432287681714,
+      "grad_norm": 2.226941029837064,
+      "learning_rate": 1.549903843792916e-05,
+      "loss": 0.7214,
+      "step": 4382
+    },
+    {
+      "epoch": 0.6706962509563886,
+      "grad_norm": 2.4219083185269015,
+      "learning_rate": 1.5496968421291687e-05,
+      "loss": 0.8276,
+      "step": 4383
+    },
+    {
+      "epoch": 0.670849273144606,
+      "grad_norm": 2.160286832500492,
+      "learning_rate": 1.5494898067063885e-05,
+      "loss": 0.7445,
+      "step": 4384
+    },
+    {
+      "epoch": 0.6710022953328233,
+      "grad_norm": 2.207589097594382,
+      "learning_rate": 1.5492827375372895e-05,
+      "loss": 0.5768,
+      "step": 4385
+    },
+    {
+      "epoch": 0.6711553175210405,
+      "grad_norm": 2.163120547285754,
+      "learning_rate": 1.5490756346345887e-05,
+      "loss": 0.7213,
+      "step": 4386
+    },
+    {
+      "epoch": 0.6713083397092579,
+      "grad_norm": 2.2884691646975153,
+      "learning_rate": 1.5488684980110052e-05,
+      "loss": 0.7249,
+      "step": 4387
+    },
+    {
+      "epoch": 0.6714613618974752,
+      "grad_norm": 2.302420265586536,
+      "learning_rate": 1.5486613276792604e-05,
+      "loss": 0.7526,
+      "step": 4388
+    },
+    {
+      "epoch": 0.6716143840856924,
+      "grad_norm": 2.2770861895161585,
+      "learning_rate": 1.5484541236520767e-05,
+      "loss": 0.6862,
+      "step": 4389
+    },
+    {
+      "epoch": 0.6717674062739097,
+      "grad_norm": 2.1419356393123854,
+      "learning_rate": 1.54824688594218e-05,
+      "loss": 0.6584,
+      "step": 4390
+    },
+    {
+      "epoch": 0.671920428462127,
+      "grad_norm": 2.29317049827774,
+      "learning_rate": 1.548039614562297e-05,
+      "loss": 0.7182,
+      "step": 4391
+    },
+    {
+      "epoch": 0.6720734506503443,
+      "grad_norm": 2.432437397742068,
+      "learning_rate": 1.5478323095251572e-05,
+      "loss": 0.7444,
+      "step": 4392
+    },
+    {
+      "epoch": 0.6722264728385616,
+      "grad_norm": 2.129156587913408,
+      "learning_rate": 1.5476249708434928e-05,
+      "loss": 0.7134,
+      "step": 4393
+    },
+    {
+      "epoch": 0.6723794950267789,
+      "grad_norm": 2.4526432928617687,
+      "learning_rate": 1.547417598530036e-05,
+      "loss": 0.8522,
+      "step": 4394
+    },
+    {
+      "epoch": 0.6725325172149962,
+      "grad_norm": 1.9197792032530787,
+      "learning_rate": 1.5472101925975232e-05,
+      "loss": 0.6823,
+      "step": 4395
+    },
+    {
+      "epoch": 0.6726855394032135,
+      "grad_norm": 2.276953404476131,
+      "learning_rate": 1.5470027530586917e-05,
+      "loss": 0.7254,
+      "step": 4396
+    },
+    {
+      "epoch": 0.6728385615914307,
+      "grad_norm": 1.8933823182228429,
+      "learning_rate": 1.546795279926281e-05,
+      "loss": 0.5938,
+      "step": 4397
+    },
+    {
+      "epoch": 0.6729915837796481,
+      "grad_norm": 2.573327299964346,
+      "learning_rate": 1.5465877732130334e-05,
+      "loss": 0.7998,
+      "step": 4398
+    },
+    {
+      "epoch": 0.6731446059678653,
+      "grad_norm": 2.0314553890273146,
+      "learning_rate": 1.5463802329316925e-05,
+      "loss": 0.7874,
+      "step": 4399
+    },
+    {
+      "epoch": 0.6732976281560826,
+      "grad_norm": 2.5070632781376814,
+      "learning_rate": 1.5461726590950038e-05,
+      "loss": 0.8208,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6734506503443,
+      "grad_norm": 2.3025849045760953,
+      "learning_rate": 1.5459650517157155e-05,
+      "loss": 0.7712,
+      "step": 4401
+    },
+    {
+      "epoch": 0.6736036725325172,
+      "grad_norm": 2.7271924988858642,
+      "learning_rate": 1.545757410806578e-05,
+      "loss": 0.8297,
+      "step": 4402
+    },
+    {
+      "epoch": 0.6737566947207345,
+      "grad_norm": 2.4592991194582554,
+      "learning_rate": 1.545549736380342e-05,
+      "loss": 0.755,
+      "step": 4403
+    },
+    {
+      "epoch": 0.6739097169089519,
+      "grad_norm": 2.2928675487770684,
+      "learning_rate": 1.545342028449763e-05,
+      "loss": 0.7825,
+      "step": 4404
+    },
+    {
+      "epoch": 0.6740627390971691,
+      "grad_norm": 2.396197246986,
+      "learning_rate": 1.5451342870275966e-05,
+      "loss": 0.7941,
+      "step": 4405
+    },
+    {
+      "epoch": 0.6742157612853864,
+      "grad_norm": 2.388805239861892,
+      "learning_rate": 1.5449265121266013e-05,
+      "loss": 0.7196,
+      "step": 4406
+    },
+    {
+      "epoch": 0.6743687834736036,
+      "grad_norm": 1.9013526614221556,
+      "learning_rate": 1.5447187037595365e-05,
+      "loss": 0.6383,
+      "step": 4407
+    },
+    {
+      "epoch": 0.674521805661821,
+      "grad_norm": 2.2240279181316263,
+      "learning_rate": 1.5445108619391658e-05,
+      "loss": 0.7109,
+      "step": 4408
+    },
+    {
+      "epoch": 0.6746748278500383,
+      "grad_norm": 2.197479041596447,
+      "learning_rate": 1.5443029866782527e-05,
+      "loss": 0.7536,
+      "step": 4409
+    },
+    {
+      "epoch": 0.6748278500382555,
+      "grad_norm": 2.2551420831296647,
+      "learning_rate": 1.544095077989564e-05,
+      "loss": 0.7972,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6749808722264728,
+      "grad_norm": 2.0258059419407903,
+      "learning_rate": 1.5438871358858677e-05,
+      "loss": 0.6498,
+      "step": 4411
+    },
+    {
+      "epoch": 0.6751338944146902,
+      "grad_norm": 2.09550472938151,
+      "learning_rate": 1.543679160379935e-05,
+      "loss": 0.7254,
+      "step": 4412
+    },
+    {
+      "epoch": 0.6752869166029074,
+      "grad_norm": 2.5330437432780735,
+      "learning_rate": 1.543471151484538e-05,
+      "loss": 0.7032,
+      "step": 4413
+    },
+    {
+      "epoch": 0.6754399387911247,
+      "grad_norm": 2.348682051814609,
+      "learning_rate": 1.5432631092124516e-05,
+      "loss": 0.8294,
+      "step": 4414
+    },
+    {
+      "epoch": 0.675592960979342,
+      "grad_norm": 2.1542786517889905,
+      "learning_rate": 1.5430550335764522e-05,
+      "loss": 0.5664,
+      "step": 4415
+    },
+    {
+      "epoch": 0.6757459831675593,
+      "grad_norm": 2.690330238546623,
+      "learning_rate": 1.542846924589319e-05,
+      "loss": 0.7578,
+      "step": 4416
+    },
+    {
+      "epoch": 0.6758990053557766,
+      "grad_norm": 2.398820373779532,
+      "learning_rate": 1.5426387822638323e-05,
+      "loss": 0.7756,
+      "step": 4417
+    },
+    {
+      "epoch": 0.6760520275439938,
+      "grad_norm": 2.583573843262456,
+      "learning_rate": 1.5424306066127754e-05,
+      "loss": 0.6986,
+      "step": 4418
+    },
+    {
+      "epoch": 0.6762050497322112,
+      "grad_norm": 2.935202140832654,
+      "learning_rate": 1.542222397648933e-05,
+      "loss": 0.8748,
+      "step": 4419
+    },
+    {
+      "epoch": 0.6763580719204285,
+      "grad_norm": 2.119216436906614,
+      "learning_rate": 1.5420141553850914e-05,
+      "loss": 0.6931,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6765110941086457,
+      "grad_norm": 2.2188152996527304,
+      "learning_rate": 1.541805879834041e-05,
+      "loss": 0.7145,
+      "step": 4421
+    },
+    {
+      "epoch": 0.6766641162968631,
+      "grad_norm": 2.2188544191352104,
+      "learning_rate": 1.5415975710085715e-05,
+      "loss": 0.6896,
+      "step": 4422
+    },
+    {
+      "epoch": 0.6768171384850803,
+      "grad_norm": 2.1994570146008567,
+      "learning_rate": 1.5413892289214765e-05,
+      "loss": 0.7304,
+      "step": 4423
+    },
+    {
+      "epoch": 0.6769701606732976,
+      "grad_norm": 2.031793893241159,
+      "learning_rate": 1.5411808535855508e-05,
+      "loss": 0.7114,
+      "step": 4424
+    },
+    {
+      "epoch": 0.677123182861515,
+      "grad_norm": 2.2744688009146516,
+      "learning_rate": 1.540972445013592e-05,
+      "loss": 0.7738,
+      "step": 4425
+    },
+    {
+      "epoch": 0.6772762050497322,
+      "grad_norm": 1.9530244694556822,
+      "learning_rate": 1.5407640032183993e-05,
+      "loss": 0.5981,
+      "step": 4426
+    },
+    {
+      "epoch": 0.6774292272379495,
+      "grad_norm": 2.2978859459660987,
+      "learning_rate": 1.5405555282127732e-05,
+      "loss": 0.6649,
+      "step": 4427
+    },
+    {
+      "epoch": 0.6775822494261668,
+      "grad_norm": 2.279440938423363,
+      "learning_rate": 1.5403470200095176e-05,
+      "loss": 0.7574,
+      "step": 4428
+    },
+    {
+      "epoch": 0.6777352716143841,
+      "grad_norm": 2.40984717918197,
+      "learning_rate": 1.5401384786214377e-05,
+      "loss": 0.7694,
+      "step": 4429
+    },
+    {
+      "epoch": 0.6778882938026014,
+      "grad_norm": 2.3415709152188033,
+      "learning_rate": 1.5399299040613408e-05,
+      "loss": 0.7097,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6780413159908186,
+      "grad_norm": 2.270028435261307,
+      "learning_rate": 1.539721296342036e-05,
+      "loss": 0.7187,
+      "step": 4431
+    },
+    {
+      "epoch": 0.678194338179036,
+      "grad_norm": 2.231886593812637,
+      "learning_rate": 1.5395126554763357e-05,
+      "loss": 0.7461,
+      "step": 4432
+    },
+    {
+      "epoch": 0.6783473603672533,
+      "grad_norm": 2.030871074521151,
+      "learning_rate": 1.539303981477052e-05,
+      "loss": 0.6325,
+      "step": 4433
+    },
+    {
+      "epoch": 0.6785003825554705,
+      "grad_norm": 2.3190377138045912,
+      "learning_rate": 1.5390952743570015e-05,
+      "loss": 0.6715,
+      "step": 4434
+    },
+    {
+      "epoch": 0.6786534047436878,
+      "grad_norm": 2.2574521183138123,
+      "learning_rate": 1.538886534129001e-05,
+      "loss": 0.7406,
+      "step": 4435
+    },
+    {
+      "epoch": 0.6788064269319051,
+      "grad_norm": 2.1855521286184527,
+      "learning_rate": 1.5386777608058702e-05,
+      "loss": 0.7811,
+      "step": 4436
+    },
+    {
+      "epoch": 0.6789594491201224,
+      "grad_norm": 2.4386053936890324,
+      "learning_rate": 1.5384689544004307e-05,
+      "loss": 0.8023,
+      "step": 4437
+    },
+    {
+      "epoch": 0.6791124713083397,
+      "grad_norm": 2.424761529872376,
+      "learning_rate": 1.5382601149255063e-05,
+      "loss": 0.7948,
+      "step": 4438
+    },
+    {
+      "epoch": 0.679265493496557,
+      "grad_norm": 2.14136896989551,
+      "learning_rate": 1.5380512423939227e-05,
+      "loss": 0.7419,
+      "step": 4439
+    },
+    {
+      "epoch": 0.6794185156847743,
+      "grad_norm": 2.1760452971611994,
+      "learning_rate": 1.5378423368185074e-05,
+      "loss": 0.7354,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6795715378729916,
+      "grad_norm": 2.4588643253323603,
+      "learning_rate": 1.53763339821209e-05,
+      "loss": 0.7851,
+      "step": 4441
+    },
+    {
+      "epoch": 0.6797245600612088,
+      "grad_norm": 2.3166074058614443,
+      "learning_rate": 1.5374244265875026e-05,
+      "loss": 0.7038,
+      "step": 4442
+    },
+    {
+      "epoch": 0.6798775822494262,
+      "grad_norm": 2.1635072983925836,
+      "learning_rate": 1.5372154219575788e-05,
+      "loss": 0.8344,
+      "step": 4443
+    },
+    {
+      "epoch": 0.6800306044376434,
+      "grad_norm": 2.2860730808637797,
+      "learning_rate": 1.5370063843351538e-05,
+      "loss": 0.7131,
+      "step": 4444
+    },
+    {
+      "epoch": 0.6801836266258607,
+      "grad_norm": 2.2691362821289847,
+      "learning_rate": 1.5367973137330667e-05,
+      "loss": 0.7036,
+      "step": 4445
+    },
+    {
+      "epoch": 0.6803366488140781,
+      "grad_norm": 2.2305641983243927,
+      "learning_rate": 1.536588210164156e-05,
+      "loss": 0.6088,
+      "step": 4446
+    },
+    {
+      "epoch": 0.6804896710022953,
+      "grad_norm": 2.2145056979037303,
+      "learning_rate": 1.5363790736412646e-05,
+      "loss": 0.8274,
+      "step": 4447
+    },
+    {
+      "epoch": 0.6806426931905126,
+      "grad_norm": 2.8007677770733492,
+      "learning_rate": 1.5361699041772358e-05,
+      "loss": 0.8732,
+      "step": 4448
+    },
+    {
+      "epoch": 0.68079571537873,
+      "grad_norm": 2.3126948601147457,
+      "learning_rate": 1.5359607017849156e-05,
+      "loss": 0.8034,
+      "step": 4449
+    },
+    {
+      "epoch": 0.6809487375669472,
+      "grad_norm": 2.2147887174883616,
+      "learning_rate": 1.535751466477152e-05,
+      "loss": 0.7773,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6811017597551645,
+      "grad_norm": 2.4159115687576618,
+      "learning_rate": 1.535542198266795e-05,
+      "loss": 0.7059,
+      "step": 4451
+    },
+    {
+      "epoch": 0.6812547819433817,
+      "grad_norm": 2.0860891506089665,
+      "learning_rate": 1.535332897166697e-05,
+      "loss": 0.7719,
+      "step": 4452
+    },
+    {
+      "epoch": 0.6814078041315991,
+      "grad_norm": 1.9432480744857157,
+      "learning_rate": 1.535123563189711e-05,
+      "loss": 0.681,
+      "step": 4453
+    },
+    {
+      "epoch": 0.6815608263198164,
+      "grad_norm": 2.4927618482640583,
+      "learning_rate": 1.5349141963486934e-05,
+      "loss": 0.7504,
+      "step": 4454
+    },
+    {
+      "epoch": 0.6817138485080336,
+      "grad_norm": 2.5913353045572856,
+      "learning_rate": 1.5347047966565026e-05,
+      "loss": 0.8291,
+      "step": 4455
+    },
+    {
+      "epoch": 0.681866870696251,
+      "grad_norm": 2.083645177191506,
+      "learning_rate": 1.534495364125999e-05,
+      "loss": 0.73,
+      "step": 4456
+    },
+    {
+      "epoch": 0.6820198928844683,
+      "grad_norm": 2.319715812091627,
+      "learning_rate": 1.5342858987700434e-05,
+      "loss": 0.8116,
+      "step": 4457
+    },
+    {
+      "epoch": 0.6821729150726855,
+      "grad_norm": 2.2755235463056542,
+      "learning_rate": 1.534076400601501e-05,
+      "loss": 0.7882,
+      "step": 4458
+    },
+    {
+      "epoch": 0.6823259372609028,
+      "grad_norm": 2.262029634001563,
+      "learning_rate": 1.5338668696332374e-05,
+      "loss": 0.6669,
+      "step": 4459
+    },
+    {
+      "epoch": 0.6824789594491201,
+      "grad_norm": 2.6664918202648633,
+      "learning_rate": 1.533657305878121e-05,
+      "loss": 0.7102,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6826319816373374,
+      "grad_norm": 2.127031830300462,
+      "learning_rate": 1.5334477093490215e-05,
+      "loss": 0.7144,
+      "step": 4461
+    },
+    {
+      "epoch": 0.6827850038255547,
+      "grad_norm": 2.4017809639453396,
+      "learning_rate": 1.5332380800588116e-05,
+      "loss": 0.7166,
+      "step": 4462
+    },
+    {
+      "epoch": 0.682938026013772,
+      "grad_norm": 2.5503010827973833,
+      "learning_rate": 1.5330284180203648e-05,
+      "loss": 0.8424,
+      "step": 4463
+    },
+    {
+      "epoch": 0.6830910482019893,
+      "grad_norm": 2.31586137036302,
+      "learning_rate": 1.532818723246558e-05,
+      "loss": 0.8279,
+      "step": 4464
+    },
+    {
+      "epoch": 0.6832440703902066,
+      "grad_norm": 2.1965797282146737,
+      "learning_rate": 1.5326089957502688e-05,
+      "loss": 0.7076,
+      "step": 4465
+    },
+    {
+      "epoch": 0.6833970925784238,
+      "grad_norm": 2.3924228128071867,
+      "learning_rate": 1.532399235544378e-05,
+      "loss": 0.7191,
+      "step": 4466
+    },
+    {
+      "epoch": 0.6835501147666412,
+      "grad_norm": 2.1377788155877355,
+      "learning_rate": 1.5321894426417667e-05,
+      "loss": 0.6807,
+      "step": 4467
+    },
+    {
+      "epoch": 0.6837031369548584,
+      "grad_norm": 2.2934918162416698,
+      "learning_rate": 1.5319796170553202e-05,
+      "loss": 0.7232,
+      "step": 4468
+    },
+    {
+      "epoch": 0.6838561591430757,
+      "grad_norm": 2.130277739327962,
+      "learning_rate": 1.5317697587979243e-05,
+      "loss": 0.638,
+      "step": 4469
+    },
+    {
+      "epoch": 0.6840091813312931,
+      "grad_norm": 2.312260633732955,
+      "learning_rate": 1.531559867882467e-05,
+      "loss": 0.8274,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6841622035195103,
+      "grad_norm": 2.586845946207191,
+      "learning_rate": 1.5313499443218388e-05,
+      "loss": 0.7508,
+      "step": 4471
+    },
+    {
+      "epoch": 0.6843152257077276,
+      "grad_norm": 2.378424806595788,
+      "learning_rate": 1.5311399881289322e-05,
+      "loss": 0.7575,
+      "step": 4472
+    },
+    {
+      "epoch": 0.684468247895945,
+      "grad_norm": 2.1522295879865427,
+      "learning_rate": 1.5309299993166405e-05,
+      "loss": 0.6589,
+      "step": 4473
+    },
+    {
+      "epoch": 0.6846212700841622,
+      "grad_norm": 2.3097812688810784,
+      "learning_rate": 1.530719977897861e-05,
+      "loss": 0.7054,
+      "step": 4474
+    },
+    {
+      "epoch": 0.6847742922723795,
+      "grad_norm": 2.3172326880772736,
+      "learning_rate": 1.5305099238854913e-05,
+      "loss": 0.6894,
+      "step": 4475
+    },
+    {
+      "epoch": 0.6849273144605967,
+      "grad_norm": 2.1003335405866106,
+      "learning_rate": 1.530299837292432e-05,
+      "loss": 0.7624,
+      "step": 4476
+    },
+    {
+      "epoch": 0.6850803366488141,
+      "grad_norm": 2.599385998272937,
+      "learning_rate": 1.5300897181315852e-05,
+      "loss": 0.7463,
+      "step": 4477
+    },
+    {
+      "epoch": 0.6852333588370314,
+      "grad_norm": 2.3900101225352395,
+      "learning_rate": 1.5298795664158547e-05,
+      "loss": 0.8077,
+      "step": 4478
+    },
+    {
+      "epoch": 0.6853863810252486,
+      "grad_norm": 2.191784966065849,
+      "learning_rate": 1.5296693821581474e-05,
+      "loss": 0.6986,
+      "step": 4479
+    },
+    {
+      "epoch": 0.685539403213466,
+      "grad_norm": 2.6280784977891147,
+      "learning_rate": 1.529459165371371e-05,
+      "loss": 0.7687,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6856924254016833,
+      "grad_norm": 2.471333458938139,
+      "learning_rate": 1.5292489160684362e-05,
+      "loss": 0.8365,
+      "step": 4481
+    },
+    {
+      "epoch": 0.6858454475899005,
+      "grad_norm": 2.3021022470545582,
+      "learning_rate": 1.529038634262255e-05,
+      "loss": 0.8477,
+      "step": 4482
+    },
+    {
+      "epoch": 0.6859984697781178,
+      "grad_norm": 2.3246002041267304,
+      "learning_rate": 1.528828319965742e-05,
+      "loss": 0.8146,
+      "step": 4483
+    },
+    {
+      "epoch": 0.6861514919663351,
+      "grad_norm": 2.1854019951696757,
+      "learning_rate": 1.5286179731918126e-05,
+      "loss": 0.6378,
+      "step": 4484
+    },
+    {
+      "epoch": 0.6863045141545524,
+      "grad_norm": 1.9833151918210372,
+      "learning_rate": 1.5284075939533854e-05,
+      "loss": 0.8069,
+      "step": 4485
+    },
+    {
+      "epoch": 0.6864575363427697,
+      "grad_norm": 2.324187336487013,
+      "learning_rate": 1.5281971822633812e-05,
+      "loss": 0.858,
+      "step": 4486
+    },
+    {
+      "epoch": 0.686610558530987,
+      "grad_norm": 2.151761922724606,
+      "learning_rate": 1.527986738134721e-05,
+      "loss": 0.63,
+      "step": 4487
+    },
+    {
+      "epoch": 0.6867635807192043,
+      "grad_norm": 1.9302796780045488,
+      "learning_rate": 1.5277762615803308e-05,
+      "loss": 0.7935,
+      "step": 4488
+    },
+    {
+      "epoch": 0.6869166029074216,
+      "grad_norm": 2.0743268312554854,
+      "learning_rate": 1.527565752613135e-05,
+      "loss": 0.7011,
+      "step": 4489
+    },
+    {
+      "epoch": 0.6870696250956388,
+      "grad_norm": 2.5266338890936133,
+      "learning_rate": 1.5273552112460626e-05,
+      "loss": 0.8576,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6872226472838562,
+      "grad_norm": 2.0054601689178035,
+      "learning_rate": 1.5271446374920435e-05,
+      "loss": 0.6995,
+      "step": 4491
+    },
+    {
+      "epoch": 0.6873756694720734,
+      "grad_norm": 2.208190332410164,
+      "learning_rate": 1.5269340313640104e-05,
+      "loss": 0.6487,
+      "step": 4492
+    },
+    {
+      "epoch": 0.6875286916602907,
+      "grad_norm": 2.083733044038154,
+      "learning_rate": 1.526723392874897e-05,
+      "loss": 0.782,
+      "step": 4493
+    },
+    {
+      "epoch": 0.6876817138485081,
+      "grad_norm": 2.2792750232808676,
+      "learning_rate": 1.526512722037639e-05,
+      "loss": 0.726,
+      "step": 4494
+    },
+    {
+      "epoch": 0.6878347360367253,
+      "grad_norm": 2.089507560978964,
+      "learning_rate": 1.526302018865175e-05,
+      "loss": 0.7222,
+      "step": 4495
+    },
+    {
+      "epoch": 0.6879877582249426,
+      "grad_norm": 2.2579331803512193,
+      "learning_rate": 1.526091283370446e-05,
+      "loss": 0.7362,
+      "step": 4496
+    },
+    {
+      "epoch": 0.68814078041316,
+      "grad_norm": 2.2197350362031933,
+      "learning_rate": 1.5258805155663924e-05,
+      "loss": 0.7617,
+      "step": 4497
+    },
+    {
+      "epoch": 0.6882938026013772,
+      "grad_norm": 2.4317255785428267,
+      "learning_rate": 1.5256697154659593e-05,
+      "loss": 0.7737,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6884468247895945,
+      "grad_norm": 2.28504107301684,
+      "learning_rate": 1.5254588830820925e-05,
+      "loss": 0.7777,
+      "step": 4499
+    },
+    {
+      "epoch": 0.6885998469778117,
+      "grad_norm": 2.272161786994709,
+      "learning_rate": 1.5252480184277403e-05,
+      "loss": 0.8356,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6887528691660291,
+      "grad_norm": 2.335814379960226,
+      "learning_rate": 1.5250371215158523e-05,
+      "loss": 0.841,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6889058913542464,
+      "grad_norm": 2.168986813520893,
+      "learning_rate": 1.5248261923593805e-05,
+      "loss": 0.7132,
+      "step": 4502
+    },
+    {
+      "epoch": 0.6890589135424636,
+      "grad_norm": 2.1288853465211113,
+      "learning_rate": 1.5246152309712795e-05,
+      "loss": 0.7998,
+      "step": 4503
+    },
+    {
+      "epoch": 0.689211935730681,
+      "grad_norm": 2.319619776523877,
+      "learning_rate": 1.5244042373645047e-05,
+      "loss": 0.802,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6893649579188983,
+      "grad_norm": 2.2306663348214166,
+      "learning_rate": 1.5241932115520142e-05,
+      "loss": 0.6982,
+      "step": 4505
+    },
+    {
+      "epoch": 0.6895179801071155,
+      "grad_norm": 1.9740335208345237,
+      "learning_rate": 1.523982153546768e-05,
+      "loss": 0.6945,
+      "step": 4506
+    },
+    {
+      "epoch": 0.6896710022953328,
+      "grad_norm": 2.3212474725289,
+      "learning_rate": 1.5237710633617278e-05,
+      "loss": 0.7454,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6898240244835501,
+      "grad_norm": 2.181356444869256,
+      "learning_rate": 1.5235599410098576e-05,
+      "loss": 0.6968,
+      "step": 4508
+    },
+    {
+      "epoch": 0.6899770466717674,
+      "grad_norm": 2.4681098588473676,
+      "learning_rate": 1.5233487865041237e-05,
+      "loss": 0.7217,
+      "step": 4509
+    },
+    {
+      "epoch": 0.6901300688599847,
+      "grad_norm": 2.2233164003664068,
+      "learning_rate": 1.5231375998574929e-05,
+      "loss": 0.6686,
+      "step": 4510
+    },
+    {
+      "epoch": 0.690283091048202,
+      "grad_norm": 2.1293830198002177,
+      "learning_rate": 1.5229263810829355e-05,
+      "loss": 0.7067,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6904361132364193,
+      "grad_norm": 2.333776572385844,
+      "learning_rate": 1.5227151301934235e-05,
+      "loss": 0.7336,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6905891354246366,
+      "grad_norm": 2.488107023266574,
+      "learning_rate": 1.522503847201931e-05,
+      "loss": 0.8592,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6907421576128538,
+      "grad_norm": 2.2582773023106526,
+      "learning_rate": 1.5222925321214326e-05,
+      "loss": 0.6915,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6908951798010712,
+      "grad_norm": 2.222243925719928,
+      "learning_rate": 1.5220811849649066e-05,
+      "loss": 0.7331,
+      "step": 4515
+    },
+    {
+      "epoch": 0.6910482019892884,
+      "grad_norm": 2.0850055482377776,
+      "learning_rate": 1.521869805745333e-05,
+      "loss": 0.7538,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6912012241775057,
+      "grad_norm": 2.221052204851177,
+      "learning_rate": 1.5216583944756927e-05,
+      "loss": 0.6374,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6913542463657231,
+      "grad_norm": 2.1796358857455043,
+      "learning_rate": 1.5214469511689698e-05,
+      "loss": 0.6842,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6915072685539403,
+      "grad_norm": 2.405770425488477,
+      "learning_rate": 1.5212354758381498e-05,
+      "loss": 0.7566,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6916602907421576,
+      "grad_norm": 2.5722939852043254,
+      "learning_rate": 1.5210239684962197e-05,
+      "loss": 0.7809,
+      "step": 4520
+    },
+    {
+      "epoch": 0.691813312930375,
+      "grad_norm": 2.465616131274673,
+      "learning_rate": 1.5208124291561702e-05,
+      "loss": 0.78,
+      "step": 4521
+    },
+    {
+      "epoch": 0.6919663351185922,
+      "grad_norm": 2.3138649418442356,
+      "learning_rate": 1.5206008578309917e-05,
+      "loss": 0.73,
+      "step": 4522
+    },
+    {
+      "epoch": 0.6921193573068095,
+      "grad_norm": 2.3167740039306755,
+      "learning_rate": 1.5203892545336776e-05,
+      "loss": 0.703,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6922723794950267,
+      "grad_norm": 2.670744482743038,
+      "learning_rate": 1.5201776192772237e-05,
+      "loss": 0.7674,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6924254016832441,
+      "grad_norm": 2.217729480121748,
+      "learning_rate": 1.5199659520746275e-05,
+      "loss": 0.7415,
+      "step": 4525
+    },
+    {
+      "epoch": 0.6925784238714614,
+      "grad_norm": 2.001448225210971,
+      "learning_rate": 1.5197542529388878e-05,
+      "loss": 0.6448,
+      "step": 4526
+    },
+    {
+      "epoch": 0.6927314460596786,
+      "grad_norm": 2.238292825597769,
+      "learning_rate": 1.5195425218830063e-05,
+      "loss": 0.7611,
+      "step": 4527
+    },
+    {
+      "epoch": 0.692884468247896,
+      "grad_norm": 2.7725948227174784,
+      "learning_rate": 1.5193307589199862e-05,
+      "loss": 0.8477,
+      "step": 4528
+    },
+    {
+      "epoch": 0.6930374904361133,
+      "grad_norm": 2.110571689962769,
+      "learning_rate": 1.519118964062832e-05,
+      "loss": 0.6644,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6931905126243305,
+      "grad_norm": 2.2568105040470257,
+      "learning_rate": 1.5189071373245521e-05,
+      "loss": 0.6753,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6933435348125478,
+      "grad_norm": 2.5626809993301594,
+      "learning_rate": 1.5186952787181546e-05,
+      "loss": 0.8011,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6934965570007651,
+      "grad_norm": 2.2221418431380155,
+      "learning_rate": 1.5184833882566507e-05,
+      "loss": 0.764,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6936495791889824,
+      "grad_norm": 2.4274947741070005,
+      "learning_rate": 1.518271465953054e-05,
+      "loss": 0.7462,
+      "step": 4533
+    },
+    {
+      "epoch": 0.6938026013771997,
+      "grad_norm": 2.270795029988415,
+      "learning_rate": 1.5180595118203791e-05,
+      "loss": 0.6212,
+      "step": 4534
+    },
+    {
+      "epoch": 0.693955623565417,
+      "grad_norm": 2.3794007844331966,
+      "learning_rate": 1.5178475258716426e-05,
+      "loss": 0.7506,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6941086457536343,
+      "grad_norm": 2.404191858909533,
+      "learning_rate": 1.5176355081198638e-05,
+      "loss": 0.7967,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6942616679418516,
+      "grad_norm": 2.272161701559529,
+      "learning_rate": 1.5174234585780633e-05,
+      "loss": 0.8067,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6944146901300688,
+      "grad_norm": 2.16796211094527,
+      "learning_rate": 1.5172113772592645e-05,
+      "loss": 0.8289,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6945677123182862,
+      "grad_norm": 2.7916757266297894,
+      "learning_rate": 1.5169992641764913e-05,
+      "loss": 0.7694,
+      "step": 4539
+    },
+    {
+      "epoch": 0.6947207345065034,
+      "grad_norm": 2.2621760884957345,
+      "learning_rate": 1.5167871193427709e-05,
+      "loss": 0.7442,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6948737566947207,
+      "grad_norm": 2.2540368894976583,
+      "learning_rate": 1.5165749427711318e-05,
+      "loss": 0.8203,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6950267788829381,
+      "grad_norm": 2.2972220573536957,
+      "learning_rate": 1.5163627344746045e-05,
+      "loss": 0.818,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6951798010711553,
+      "grad_norm": 2.482793009436285,
+      "learning_rate": 1.516150494466222e-05,
+      "loss": 0.8353,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6953328232593726,
+      "grad_norm": 2.3515004139625897,
+      "learning_rate": 1.515938222759018e-05,
+      "loss": 0.778,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6954858454475898,
+      "grad_norm": 2.2361132567474615,
+      "learning_rate": 1.5157259193660295e-05,
+      "loss": 0.7126,
+      "step": 4545
+    },
+    {
+      "epoch": 0.6956388676358072,
+      "grad_norm": 2.5681665991598286,
+      "learning_rate": 1.5155135843002948e-05,
+      "loss": 0.8872,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6957918898240245,
+      "grad_norm": 2.3461733455885727,
+      "learning_rate": 1.5153012175748542e-05,
+      "loss": 0.7425,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6959449120122417,
+      "grad_norm": 2.1546909441667124,
+      "learning_rate": 1.5150888192027502e-05,
+      "loss": 0.5938,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6960979342004591,
+      "grad_norm": 2.1532119909978937,
+      "learning_rate": 1.5148763891970264e-05,
+      "loss": 0.6435,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6962509563886764,
+      "grad_norm": 2.141592558646377,
+      "learning_rate": 1.5146639275707298e-05,
+      "loss": 0.6884,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6964039785768936,
+      "grad_norm": 2.5681176533512216,
+      "learning_rate": 1.5144514343369075e-05,
+      "loss": 0.9509,
+      "step": 4551
+    },
+    {
+      "epoch": 0.696557000765111,
+      "grad_norm": 2.3532875826075554,
+      "learning_rate": 1.5142389095086103e-05,
+      "loss": 0.7296,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6967100229533282,
+      "grad_norm": 2.464874640315048,
+      "learning_rate": 1.5140263530988898e-05,
+      "loss": 0.8494,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6968630451415455,
+      "grad_norm": 2.3773010925552773,
+      "learning_rate": 1.5138137651208004e-05,
+      "loss": 0.6996,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6970160673297628,
+      "grad_norm": 1.9930665522309956,
+      "learning_rate": 1.5136011455873973e-05,
+      "loss": 0.7253,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6971690895179801,
+      "grad_norm": 2.287025093599191,
+      "learning_rate": 1.513388494511739e-05,
+      "loss": 0.6763,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6973221117061974,
+      "grad_norm": 2.141087425346586,
+      "learning_rate": 1.5131758119068844e-05,
+      "loss": 0.6891,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6974751338944147,
+      "grad_norm": 1.9646395637835101,
+      "learning_rate": 1.512963097785896e-05,
+      "loss": 0.6165,
+      "step": 4558
+    },
+    {
+      "epoch": 0.697628156082632,
+      "grad_norm": 2.0897594617619903,
+      "learning_rate": 1.5127503521618371e-05,
+      "loss": 0.5849,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6977811782708493,
+      "grad_norm": 2.305818341701876,
+      "learning_rate": 1.5125375750477731e-05,
+      "loss": 0.7084,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6979342004590665,
+      "grad_norm": 2.5024862701767345,
+      "learning_rate": 1.512324766456771e-05,
+      "loss": 0.8031,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6980872226472838,
+      "grad_norm": 2.4597919538294217,
+      "learning_rate": 1.5121119264019015e-05,
+      "loss": 0.8691,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6982402448355012,
+      "grad_norm": 2.2910984541415385,
+      "learning_rate": 1.511899054896235e-05,
+      "loss": 0.6815,
+      "step": 4563
+    },
+    {
+      "epoch": 0.6983932670237184,
+      "grad_norm": 2.364384147976584,
+      "learning_rate": 1.5116861519528447e-05,
+      "loss": 0.7718,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6985462892119357,
+      "grad_norm": 2.4982226125776226,
+      "learning_rate": 1.5114732175848064e-05,
+      "loss": 0.7911,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6986993114001531,
+      "grad_norm": 2.6777365242008986,
+      "learning_rate": 1.5112602518051971e-05,
+      "loss": 0.9079,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6988523335883703,
+      "grad_norm": 2.3841456314000538,
+      "learning_rate": 1.5110472546270952e-05,
+      "loss": 0.6749,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6990053557765876,
+      "grad_norm": 2.150933316605576,
+      "learning_rate": 1.5108342260635821e-05,
+      "loss": 0.576,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6991583779648048,
+      "grad_norm": 2.383057686481416,
+      "learning_rate": 1.5106211661277412e-05,
+      "loss": 0.7526,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6993114001530222,
+      "grad_norm": 2.273681917930905,
+      "learning_rate": 1.5104080748326568e-05,
+      "loss": 0.817,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6994644223412395,
+      "grad_norm": 2.223542529079718,
+      "learning_rate": 1.510194952191416e-05,
+      "loss": 0.7082,
+      "step": 4571
+    },
+    {
+      "epoch": 0.6996174445294567,
+      "grad_norm": 2.3811340188781487,
+      "learning_rate": 1.509981798217107e-05,
+      "loss": 0.7807,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6997704667176741,
+      "grad_norm": 2.278119095067911,
+      "learning_rate": 1.5097686129228208e-05,
+      "loss": 0.6676,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6999234889058914,
+      "grad_norm": 2.2273074977858,
+      "learning_rate": 1.5095553963216496e-05,
+      "loss": 0.6651,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7000765110941086,
+      "grad_norm": 2.483082682130355,
+      "learning_rate": 1.5093421484266885e-05,
+      "loss": 0.6142,
+      "step": 4575
+    },
+    {
+      "epoch": 0.700229533282326,
+      "grad_norm": 2.1945222620117533,
+      "learning_rate": 1.5091288692510334e-05,
+      "loss": 0.6586,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7003825554705432,
+      "grad_norm": 2.3719793672214244,
+      "learning_rate": 1.5089155588077824e-05,
+      "loss": 0.814,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7005355776587605,
+      "grad_norm": 2.4060459274427783,
+      "learning_rate": 1.5087022171100362e-05,
+      "loss": 0.7127,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7006885998469778,
+      "grad_norm": 2.309202456137207,
+      "learning_rate": 1.5084888441708972e-05,
+      "loss": 0.6896,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7008416220351951,
+      "grad_norm": 2.3124390047154764,
+      "learning_rate": 1.5082754400034687e-05,
+      "loss": 0.8279,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7009946442234124,
+      "grad_norm": 2.0904361229422763,
+      "learning_rate": 1.508062004620857e-05,
+      "loss": 0.6954,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7011476664116297,
+      "grad_norm": 2.094692589966736,
+      "learning_rate": 1.5078485380361702e-05,
+      "loss": 0.6429,
+      "step": 4582
+    },
+    {
+      "epoch": 0.701300688599847,
+      "grad_norm": 2.4664976414160535,
+      "learning_rate": 1.507635040262518e-05,
+      "loss": 0.734,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7014537107880643,
+      "grad_norm": 2.230209940959713,
+      "learning_rate": 1.507421511313012e-05,
+      "loss": 0.7579,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7016067329762815,
+      "grad_norm": 2.129917064589851,
+      "learning_rate": 1.5072079512007661e-05,
+      "loss": 0.7132,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7017597551644988,
+      "grad_norm": 2.6035639966486883,
+      "learning_rate": 1.5069943599388952e-05,
+      "loss": 0.8368,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7019127773527162,
+      "grad_norm": 2.0753617962083895,
+      "learning_rate": 1.506780737540518e-05,
+      "loss": 0.6343,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7020657995409334,
+      "grad_norm": 2.2775042401770023,
+      "learning_rate": 1.5065670840187527e-05,
+      "loss": 0.7708,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7022188217291507,
+      "grad_norm": 2.1526579899134552,
+      "learning_rate": 1.5063533993867209e-05,
+      "loss": 0.6628,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7023718439173681,
+      "grad_norm": 2.176150777690883,
+      "learning_rate": 1.5061396836575463e-05,
+      "loss": 0.7387,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7025248661055853,
+      "grad_norm": 2.129854293577214,
+      "learning_rate": 1.5059259368443541e-05,
+      "loss": 0.6796,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7026778882938026,
+      "grad_norm": 2.207031757536129,
+      "learning_rate": 1.5057121589602703e-05,
+      "loss": 0.6714,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7028309104820198,
+      "grad_norm": 2.148029900786362,
+      "learning_rate": 1.5054983500184245e-05,
+      "loss": 0.8089,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7029839326702372,
+      "grad_norm": 2.138588131175154,
+      "learning_rate": 1.5052845100319478e-05,
+      "loss": 0.6533,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7031369548584545,
+      "grad_norm": 2.350243331629356,
+      "learning_rate": 1.5050706390139724e-05,
+      "loss": 0.7045,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7032899770466717,
+      "grad_norm": 2.4869590482805126,
+      "learning_rate": 1.5048567369776334e-05,
+      "loss": 0.8078,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7034429992348891,
+      "grad_norm": 2.367666383463255,
+      "learning_rate": 1.5046428039360673e-05,
+      "loss": 0.6646,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7035960214231064,
+      "grad_norm": 2.06402576755447,
+      "learning_rate": 1.5044288399024121e-05,
+      "loss": 0.8041,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7037490436113236,
+      "grad_norm": 2.15944555184331,
+      "learning_rate": 1.5042148448898086e-05,
+      "loss": 0.6758,
+      "step": 4599
+    },
+    {
+      "epoch": 0.703902065799541,
+      "grad_norm": 2.5051434873442933,
+      "learning_rate": 1.504000818911399e-05,
+      "loss": 0.8164,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7040550879877582,
+      "grad_norm": 2.3479880611298487,
+      "learning_rate": 1.5037867619803276e-05,
+      "loss": 0.8284,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7042081101759755,
+      "grad_norm": 2.1908232486121,
+      "learning_rate": 1.5035726741097398e-05,
+      "loss": 0.7382,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7043611323641928,
+      "grad_norm": 2.3218035445254226,
+      "learning_rate": 1.5033585553127848e-05,
+      "loss": 0.6095,
+      "step": 4603
+    },
+    {
+      "epoch": 0.7045141545524101,
+      "grad_norm": 2.2092916242716907,
+      "learning_rate": 1.5031444056026114e-05,
+      "loss": 0.8147,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7046671767406274,
+      "grad_norm": 2.217187498878475,
+      "learning_rate": 1.5029302249923716e-05,
+      "loss": 0.7882,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7048201989288447,
+      "grad_norm": 2.2389718404752834,
+      "learning_rate": 1.5027160134952197e-05,
+      "loss": 0.6402,
+      "step": 4606
+    },
+    {
+      "epoch": 0.704973221117062,
+      "grad_norm": 2.4569438697426893,
+      "learning_rate": 1.5025017711243104e-05,
+      "loss": 0.8244,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7051262433052793,
+      "grad_norm": 2.1755220170600422,
+      "learning_rate": 1.5022874978928015e-05,
+      "loss": 0.6957,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7052792654934965,
+      "grad_norm": 2.1244164995728982,
+      "learning_rate": 1.5020731938138523e-05,
+      "loss": 0.6932,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7054322876817138,
+      "grad_norm": 2.2433322410932757,
+      "learning_rate": 1.5018588589006247e-05,
+      "loss": 0.7543,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7055853098699312,
+      "grad_norm": 2.118332433021445,
+      "learning_rate": 1.501644493166281e-05,
+      "loss": 0.8065,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7057383320581484,
+      "grad_norm": 2.1458348809316914,
+      "learning_rate": 1.5014300966239863e-05,
+      "loss": 0.741,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7058913542463657,
+      "grad_norm": 2.2114252595349178,
+      "learning_rate": 1.501215669286908e-05,
+      "loss": 0.7222,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7060443764345831,
+      "grad_norm": 2.114569239497734,
+      "learning_rate": 1.5010012111682146e-05,
+      "loss": 0.6805,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7061973986228003,
+      "grad_norm": 2.065733198288359,
+      "learning_rate": 1.5007867222810769e-05,
+      "loss": 0.6061,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7063504208110176,
+      "grad_norm": 2.214125841483104,
+      "learning_rate": 1.5005722026386675e-05,
+      "loss": 0.6689,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7065034429992348,
+      "grad_norm": 2.2882545330622057,
+      "learning_rate": 1.5003576522541609e-05,
+      "loss": 0.6518,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7066564651874522,
+      "grad_norm": 2.3663303914915375,
+      "learning_rate": 1.5001430711407335e-05,
+      "loss": 0.6424,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7068094873756695,
+      "grad_norm": 2.431556481342764,
+      "learning_rate": 1.4999284593115633e-05,
+      "loss": 0.8046,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7069625095638867,
+      "grad_norm": 2.389676632888257,
+      "learning_rate": 1.499713816779831e-05,
+      "loss": 0.7433,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7071155317521041,
+      "grad_norm": 2.2577603438639184,
+      "learning_rate": 1.4994991435587181e-05,
+      "loss": 0.7294,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7072685539403214,
+      "grad_norm": 2.309661103988358,
+      "learning_rate": 1.499284439661409e-05,
+      "loss": 0.769,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7074215761285386,
+      "grad_norm": 2.117249156808084,
+      "learning_rate": 1.499069705101089e-05,
+      "loss": 0.7223,
+      "step": 4623
+    },
+    {
+      "epoch": 0.707574598316756,
+      "grad_norm": 2.2679542217853284,
+      "learning_rate": 1.4988549398909461e-05,
+      "loss": 0.7463,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7077276205049732,
+      "grad_norm": 2.4233667275174873,
+      "learning_rate": 1.4986401440441698e-05,
+      "loss": 0.6929,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7078806426931905,
+      "grad_norm": 2.4387312401102728,
+      "learning_rate": 1.4984253175739516e-05,
+      "loss": 0.7185,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7080336648814078,
+      "grad_norm": 2.1516389780097813,
+      "learning_rate": 1.4982104604934847e-05,
+      "loss": 0.6914,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7081866870696251,
+      "grad_norm": 2.1911271817443976,
+      "learning_rate": 1.4979955728159648e-05,
+      "loss": 0.7076,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7083397092578424,
+      "grad_norm": 2.2764653082340605,
+      "learning_rate": 1.4977806545545882e-05,
+      "loss": 0.7879,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7084927314460597,
+      "grad_norm": 2.258450187406121,
+      "learning_rate": 1.4975657057225541e-05,
+      "loss": 0.8093,
+      "step": 4630
+    },
+    {
+      "epoch": 0.708645753634277,
+      "grad_norm": 2.102515060129002,
+      "learning_rate": 1.497350726333064e-05,
+      "loss": 0.6645,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7087987758224943,
+      "grad_norm": 2.3893729812773725,
+      "learning_rate": 1.4971357163993201e-05,
+      "loss": 0.7671,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7089517980107115,
+      "grad_norm": 1.9505505804431376,
+      "learning_rate": 1.4969206759345268e-05,
+      "loss": 0.6499,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7091048201989288,
+      "grad_norm": 2.2384013451963143,
+      "learning_rate": 1.4967056049518909e-05,
+      "loss": 0.7929,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7092578423871462,
+      "grad_norm": 2.4005614517986893,
+      "learning_rate": 1.4964905034646207e-05,
+      "loss": 0.7999,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7094108645753634,
+      "grad_norm": 2.0678028281202065,
+      "learning_rate": 1.4962753714859265e-05,
+      "loss": 0.7242,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7095638867635807,
+      "grad_norm": 2.462845353036398,
+      "learning_rate": 1.4960602090290201e-05,
+      "loss": 0.8538,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7097169089517981,
+      "grad_norm": 2.3117301226127798,
+      "learning_rate": 1.495845016107116e-05,
+      "loss": 0.7825,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7098699311400153,
+      "grad_norm": 2.2646133688524297,
+      "learning_rate": 1.4956297927334293e-05,
+      "loss": 0.7812,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7100229533282326,
+      "grad_norm": 2.2227803644605286,
+      "learning_rate": 1.4954145389211783e-05,
+      "loss": 0.6609,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7101759755164498,
+      "grad_norm": 2.282965177589414,
+      "learning_rate": 1.4951992546835822e-05,
+      "loss": 0.7945,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7103289977046672,
+      "grad_norm": 2.4988127974075436,
+      "learning_rate": 1.494983940033863e-05,
+      "loss": 0.7835,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7104820198928845,
+      "grad_norm": 2.0580200836857627,
+      "learning_rate": 1.4947685949852433e-05,
+      "loss": 0.7276,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7106350420811017,
+      "grad_norm": 2.115513884581086,
+      "learning_rate": 1.4945532195509489e-05,
+      "loss": 0.8458,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7107880642693191,
+      "grad_norm": 2.1312885230885694,
+      "learning_rate": 1.4943378137442066e-05,
+      "loss": 0.6838,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7109410864575364,
+      "grad_norm": 2.0192826011804788,
+      "learning_rate": 1.494122377578245e-05,
+      "loss": 0.6525,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7110941086457536,
+      "grad_norm": 2.14333559550761,
+      "learning_rate": 1.4939069110662953e-05,
+      "loss": 0.7441,
+      "step": 4647
+    },
+    {
+      "epoch": 0.711247130833971,
+      "grad_norm": 2.735799505033225,
+      "learning_rate": 1.4936914142215901e-05,
+      "loss": 0.8995,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7114001530221882,
+      "grad_norm": 2.3028298562489193,
+      "learning_rate": 1.4934758870573638e-05,
+      "loss": 0.8153,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7115531752104055,
+      "grad_norm": 2.0552740637304434,
+      "learning_rate": 1.4932603295868529e-05,
+      "loss": 0.7292,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7117061973986228,
+      "grad_norm": 2.0362932635684676,
+      "learning_rate": 1.4930447418232954e-05,
+      "loss": 0.7156,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7118592195868401,
+      "grad_norm": 2.3297210965014323,
+      "learning_rate": 1.4928291237799316e-05,
+      "loss": 0.7321,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7120122417750574,
+      "grad_norm": 2.3448837720146587,
+      "learning_rate": 1.4926134754700033e-05,
+      "loss": 0.7765,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7121652639632746,
+      "grad_norm": 2.090763497741587,
+      "learning_rate": 1.4923977969067542e-05,
+      "loss": 0.7825,
+      "step": 4654
+    },
+    {
+      "epoch": 0.712318286151492,
+      "grad_norm": 2.1568103828538137,
+      "learning_rate": 1.4921820881034303e-05,
+      "loss": 0.6317,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7124713083397093,
+      "grad_norm": 2.1517165425688307,
+      "learning_rate": 1.4919663490732788e-05,
+      "loss": 0.7669,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7126243305279265,
+      "grad_norm": 2.348912710534796,
+      "learning_rate": 1.4917505798295496e-05,
+      "loss": 0.6615,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7127773527161438,
+      "grad_norm": 2.3188303433517197,
+      "learning_rate": 1.491534780385493e-05,
+      "loss": 0.6957,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7129303749043612,
+      "grad_norm": 2.31676216581471,
+      "learning_rate": 1.4913189507543629e-05,
+      "loss": 0.7658,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7130833970925784,
+      "grad_norm": 2.3905665163661616,
+      "learning_rate": 1.491103090949414e-05,
+      "loss": 0.8188,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7132364192807957,
+      "grad_norm": 2.331333999855849,
+      "learning_rate": 1.4908872009839032e-05,
+      "loss": 0.7859,
+      "step": 4661
+    },
+    {
+      "epoch": 0.713389441469013,
+      "grad_norm": 2.383095717161027,
+      "learning_rate": 1.4906712808710887e-05,
+      "loss": 0.7431,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7135424636572303,
+      "grad_norm": 2.225256917614341,
+      "learning_rate": 1.4904553306242315e-05,
+      "loss": 0.7684,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7136954858454476,
+      "grad_norm": 2.313231003189659,
+      "learning_rate": 1.4902393502565938e-05,
+      "loss": 0.7094,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7138485080336648,
+      "grad_norm": 2.602060352447604,
+      "learning_rate": 1.49002333978144e-05,
+      "loss": 0.7902,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7140015302218822,
+      "grad_norm": 2.279485041471543,
+      "learning_rate": 1.4898072992120354e-05,
+      "loss": 0.6993,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7141545524100995,
+      "grad_norm": 2.3635218759370136,
+      "learning_rate": 1.4895912285616488e-05,
+      "loss": 0.6515,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7143075745983167,
+      "grad_norm": 2.096166872095998,
+      "learning_rate": 1.489375127843549e-05,
+      "loss": 0.6639,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7144605967865341,
+      "grad_norm": 2.3381009761630818,
+      "learning_rate": 1.489158997071009e-05,
+      "loss": 0.7515,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7146136189747513,
+      "grad_norm": 2.195710905149429,
+      "learning_rate": 1.488942836257301e-05,
+      "loss": 0.6466,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7147666411629686,
+      "grad_norm": 2.1753748778774877,
+      "learning_rate": 1.4887266454157005e-05,
+      "loss": 0.6547,
+      "step": 4671
+    },
+    {
+      "epoch": 0.714919663351186,
+      "grad_norm": 2.122937116378295,
+      "learning_rate": 1.488510424559485e-05,
+      "loss": 0.7748,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7150726855394032,
+      "grad_norm": 2.406407272988174,
+      "learning_rate": 1.4882941737019334e-05,
+      "loss": 0.8305,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7152257077276205,
+      "grad_norm": 2.2145862249870514,
+      "learning_rate": 1.488077892856326e-05,
+      "loss": 0.6078,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7153787299158378,
+      "grad_norm": 2.1791170205115225,
+      "learning_rate": 1.487861582035946e-05,
+      "loss": 0.6832,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7155317521040551,
+      "grad_norm": 2.21390173825772,
+      "learning_rate": 1.4876452412540778e-05,
+      "loss": 0.7003,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7156847742922724,
+      "grad_norm": 2.276605772799139,
+      "learning_rate": 1.4874288705240077e-05,
+      "loss": 0.739,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7158377964804896,
+      "grad_norm": 2.14951981367647,
+      "learning_rate": 1.4872124698590239e-05,
+      "loss": 0.6115,
+      "step": 4678
+    },
+    {
+      "epoch": 0.715990818668707,
+      "grad_norm": 2.3320357024227785,
+      "learning_rate": 1.4869960392724162e-05,
+      "loss": 0.7127,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7161438408569243,
+      "grad_norm": 2.3126591024809047,
+      "learning_rate": 1.4867795787774766e-05,
+      "loss": 0.7032,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7162968630451415,
+      "grad_norm": 2.4048082026718247,
+      "learning_rate": 1.4865630883874983e-05,
+      "loss": 0.7534,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7164498852333588,
+      "grad_norm": 2.451316391581056,
+      "learning_rate": 1.486346568115778e-05,
+      "loss": 0.7529,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7166029074215762,
+      "grad_norm": 2.053071504108291,
+      "learning_rate": 1.4861300179756122e-05,
+      "loss": 0.6541,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7167559296097934,
+      "grad_norm": 2.3417044813298715,
+      "learning_rate": 1.4859134379803e-05,
+      "loss": 0.7603,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7169089517980107,
+      "grad_norm": 2.3473788742841015,
+      "learning_rate": 1.4856968281431428e-05,
+      "loss": 0.6605,
+      "step": 4685
+    },
+    {
+      "epoch": 0.717061973986228,
+      "grad_norm": 2.2490942525781707,
+      "learning_rate": 1.4854801884774435e-05,
+      "loss": 0.6712,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7172149961744453,
+      "grad_norm": 2.231128453832111,
+      "learning_rate": 1.4852635189965063e-05,
+      "loss": 0.6849,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7173680183626626,
+      "grad_norm": 2.0474238341958215,
+      "learning_rate": 1.485046819713638e-05,
+      "loss": 0.7608,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7175210405508798,
+      "grad_norm": 2.178205755401646,
+      "learning_rate": 1.4848300906421473e-05,
+      "loss": 0.8481,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7176740627390972,
+      "grad_norm": 2.205174854802361,
+      "learning_rate": 1.4846133317953441e-05,
+      "loss": 0.7059,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7178270849273145,
+      "grad_norm": 2.3020851967727944,
+      "learning_rate": 1.4843965431865401e-05,
+      "loss": 0.7218,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7179801071155317,
+      "grad_norm": 2.4590759700573046,
+      "learning_rate": 1.4841797248290494e-05,
+      "loss": 0.8537,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7181331293037491,
+      "grad_norm": 2.1277534202106225,
+      "learning_rate": 1.483962876736188e-05,
+      "loss": 0.6644,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7182861514919663,
+      "grad_norm": 2.3080061593741434,
+      "learning_rate": 1.4837459989212728e-05,
+      "loss": 0.6971,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7184391736801836,
+      "grad_norm": 2.014506867651919,
+      "learning_rate": 1.4835290913976237e-05,
+      "loss": 0.7153,
+      "step": 4695
+    },
+    {
+      "epoch": 0.718592195868401,
+      "grad_norm": 2.3493421964319565,
+      "learning_rate": 1.4833121541785612e-05,
+      "loss": 0.8454,
+      "step": 4696
+    },
+    {
+      "epoch": 0.7187452180566182,
+      "grad_norm": 2.5341913901501676,
+      "learning_rate": 1.4830951872774084e-05,
+      "loss": 0.8979,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7188982402448355,
+      "grad_norm": 1.9746285755318667,
+      "learning_rate": 1.4828781907074907e-05,
+      "loss": 0.6924,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7190512624330528,
+      "grad_norm": 2.3123766427327763,
+      "learning_rate": 1.4826611644821342e-05,
+      "loss": 0.7356,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7192042846212701,
+      "grad_norm": 2.079211608687565,
+      "learning_rate": 1.4824441086146673e-05,
+      "loss": 0.6949,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7193573068094874,
+      "grad_norm": 2.0936906022949326,
+      "learning_rate": 1.4822270231184202e-05,
+      "loss": 0.6237,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7195103289977046,
+      "grad_norm": 2.2139872126331377,
+      "learning_rate": 1.4820099080067256e-05,
+      "loss": 0.8631,
+      "step": 4702
+    },
+    {
+      "epoch": 0.719663351185922,
+      "grad_norm": 2.8509554055876634,
+      "learning_rate": 1.4817927632929166e-05,
+      "loss": 0.865,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7198163733741393,
+      "grad_norm": 2.203551816183465,
+      "learning_rate": 1.4815755889903292e-05,
+      "loss": 0.8114,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7199693955623565,
+      "grad_norm": 2.2383720960632463,
+      "learning_rate": 1.481358385112301e-05,
+      "loss": 0.7976,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7201224177505738,
+      "grad_norm": 2.066559424344547,
+      "learning_rate": 1.4811411516721713e-05,
+      "loss": 0.7183,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7202754399387912,
+      "grad_norm": 2.104497394187465,
+      "learning_rate": 1.480923888683281e-05,
+      "loss": 0.7099,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7204284621270084,
+      "grad_norm": 2.111770796173033,
+      "learning_rate": 1.4807065961589737e-05,
+      "loss": 0.6569,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7205814843152257,
+      "grad_norm": 2.147742730612102,
+      "learning_rate": 1.4804892741125934e-05,
+      "loss": 0.7724,
+      "step": 4709
+    },
+    {
+      "epoch": 0.720734506503443,
+      "grad_norm": 2.246631746401689,
+      "learning_rate": 1.4802719225574876e-05,
+      "loss": 0.8219,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7208875286916603,
+      "grad_norm": 2.13309088287406,
+      "learning_rate": 1.4800545415070037e-05,
+      "loss": 0.6694,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7210405508798776,
+      "grad_norm": 2.0313301699605866,
+      "learning_rate": 1.4798371309744925e-05,
+      "loss": 0.6445,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7211935730680948,
+      "grad_norm": 2.530711557247044,
+      "learning_rate": 1.4796196909733063e-05,
+      "loss": 0.7704,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7213465952563122,
+      "grad_norm": 2.47685892445677,
+      "learning_rate": 1.4794022215167983e-05,
+      "loss": 0.8267,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7214996174445295,
+      "grad_norm": 2.0937083712741194,
+      "learning_rate": 1.4791847226183245e-05,
+      "loss": 0.6132,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7216526396327467,
+      "grad_norm": 2.189520413249951,
+      "learning_rate": 1.4789671942912424e-05,
+      "loss": 0.6962,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7218056618209641,
+      "grad_norm": 2.39097362182118,
+      "learning_rate": 1.4787496365489112e-05,
+      "loss": 0.6877,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7219586840091813,
+      "grad_norm": 2.3351690545072827,
+      "learning_rate": 1.4785320494046917e-05,
+      "loss": 0.7894,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7221117061973986,
+      "grad_norm": 2.061697603752519,
+      "learning_rate": 1.4783144328719471e-05,
+      "loss": 0.6861,
+      "step": 4719
+    },
+    {
+      "epoch": 0.722264728385616,
+      "grad_norm": 2.2225612629930938,
+      "learning_rate": 1.4780967869640424e-05,
+      "loss": 0.6203,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7224177505738332,
+      "grad_norm": 2.573723966712967,
+      "learning_rate": 1.4778791116943432e-05,
+      "loss": 0.7715,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7225707727620505,
+      "grad_norm": 2.594210095762241,
+      "learning_rate": 1.4776614070762183e-05,
+      "loss": 0.6993,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7227237949502678,
+      "grad_norm": 2.371055805718302,
+      "learning_rate": 1.477443673123038e-05,
+      "loss": 0.8131,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7228768171384851,
+      "grad_norm": 2.096628347839013,
+      "learning_rate": 1.4772259098481741e-05,
+      "loss": 0.7364,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7230298393267024,
+      "grad_norm": 2.3398331654499103,
+      "learning_rate": 1.4770081172649995e-05,
+      "loss": 0.8091,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7231828615149196,
+      "grad_norm": 2.1132524686931795,
+      "learning_rate": 1.4767902953868908e-05,
+      "loss": 0.8064,
+      "step": 4726
+    },
+    {
+      "epoch": 0.723335883703137,
+      "grad_norm": 2.0520566030415286,
+      "learning_rate": 1.4765724442272252e-05,
+      "loss": 0.6772,
+      "step": 4727
+    },
+    {
+      "epoch": 0.7234889058913543,
+      "grad_norm": 2.30000431071158,
+      "learning_rate": 1.4763545637993808e-05,
+      "loss": 0.7558,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7236419280795715,
+      "grad_norm": 2.632293513300626,
+      "learning_rate": 1.4761366541167394e-05,
+      "loss": 0.746,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7237949502677888,
+      "grad_norm": 2.379979644133256,
+      "learning_rate": 1.4759187151926833e-05,
+      "loss": 0.8553,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7239479724560062,
+      "grad_norm": 2.2346609604101344,
+      "learning_rate": 1.4757007470405973e-05,
+      "loss": 0.6299,
+      "step": 4731
+    },
+    {
+      "epoch": 0.7241009946442234,
+      "grad_norm": 2.27430307026169,
+      "learning_rate": 1.4754827496738672e-05,
+      "loss": 0.6518,
+      "step": 4732
+    },
+    {
+      "epoch": 0.7242540168324407,
+      "grad_norm": 2.1413328730472427,
+      "learning_rate": 1.4752647231058816e-05,
+      "loss": 0.6566,
+      "step": 4733
+    },
+    {
+      "epoch": 0.724407039020658,
+      "grad_norm": 2.3859874983347202,
+      "learning_rate": 1.47504666735003e-05,
+      "loss": 0.7186,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7245600612088753,
+      "grad_norm": 2.3047056874773912,
+      "learning_rate": 1.4748285824197042e-05,
+      "loss": 0.866,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7247130833970926,
+      "grad_norm": 2.1601134195962137,
+      "learning_rate": 1.4746104683282978e-05,
+      "loss": 0.6606,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7248661055853098,
+      "grad_norm": 2.416663587369718,
+      "learning_rate": 1.4743923250892056e-05,
+      "loss": 0.8479,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7250191277735272,
+      "grad_norm": 2.4091001521362276,
+      "learning_rate": 1.4741741527158246e-05,
+      "loss": 0.7452,
+      "step": 4738
+    },
+    {
+      "epoch": 0.7251721499617445,
+      "grad_norm": 2.1203901701257464,
+      "learning_rate": 1.4739559512215546e-05,
+      "loss": 0.6542,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7253251721499617,
+      "grad_norm": 2.132147928958173,
+      "learning_rate": 1.4737377206197951e-05,
+      "loss": 0.7092,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7254781943381791,
+      "grad_norm": 2.208144698192683,
+      "learning_rate": 1.4735194609239487e-05,
+      "loss": 0.7039,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7256312165263963,
+      "grad_norm": 2.505747937898946,
+      "learning_rate": 1.4733011721474202e-05,
+      "loss": 0.8391,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7257842387146136,
+      "grad_norm": 2.04294489654917,
+      "learning_rate": 1.4730828543036151e-05,
+      "loss": 0.6893,
+      "step": 4743
+    },
+    {
+      "epoch": 0.725937260902831,
+      "grad_norm": 2.2859908932549153,
+      "learning_rate": 1.4728645074059409e-05,
+      "loss": 0.7034,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7260902830910482,
+      "grad_norm": 2.0581272816435408,
+      "learning_rate": 1.4726461314678075e-05,
+      "loss": 0.6576,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7262433052792655,
+      "grad_norm": 2.1795007417010406,
+      "learning_rate": 1.4724277265026263e-05,
+      "loss": 0.7988,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7263963274674828,
+      "grad_norm": 2.2960157182927365,
+      "learning_rate": 1.4722092925238106e-05,
+      "loss": 0.7517,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7265493496557001,
+      "grad_norm": 2.623740200196195,
+      "learning_rate": 1.4719908295447745e-05,
+      "loss": 0.8784,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7267023718439174,
+      "grad_norm": 2.360442706550242,
+      "learning_rate": 1.4717723375789353e-05,
+      "loss": 0.8329,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7268553940321346,
+      "grad_norm": 2.4744425533825245,
+      "learning_rate": 1.4715538166397109e-05,
+      "loss": 0.6913,
+      "step": 4750
+    },
+    {
+      "epoch": 0.727008416220352,
+      "grad_norm": 2.2088002606216492,
+      "learning_rate": 1.4713352667405222e-05,
+      "loss": 0.6413,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7271614384085693,
+      "grad_norm": 2.1576929521842128,
+      "learning_rate": 1.4711166878947911e-05,
+      "loss": 0.7331,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7273144605967865,
+      "grad_norm": 2.1404435433709343,
+      "learning_rate": 1.470898080115941e-05,
+      "loss": 0.7478,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7274674827850038,
+      "grad_norm": 2.0884135256005627,
+      "learning_rate": 1.4706794434173974e-05,
+      "loss": 0.6151,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7276205049732212,
+      "grad_norm": 2.296005837056293,
+      "learning_rate": 1.470460777812588e-05,
+      "loss": 0.7234,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7277735271614384,
+      "grad_norm": 2.2279102958195995,
+      "learning_rate": 1.470242083314942e-05,
+      "loss": 0.7359,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7279265493496557,
+      "grad_norm": 2.4351625984189536,
+      "learning_rate": 1.47002335993789e-05,
+      "loss": 0.7718,
+      "step": 4757
+    },
+    {
+      "epoch": 0.728079571537873,
+      "grad_norm": 2.4439542483712335,
+      "learning_rate": 1.4698046076948647e-05,
+      "loss": 0.7665,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7282325937260903,
+      "grad_norm": 2.157585434584162,
+      "learning_rate": 1.4695858265993005e-05,
+      "loss": 0.7102,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7283856159143076,
+      "grad_norm": 2.38408004828983,
+      "learning_rate": 1.4693670166646337e-05,
+      "loss": 0.7396,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7285386381025248,
+      "grad_norm": 2.2861670459333525,
+      "learning_rate": 1.4691481779043022e-05,
+      "loss": 0.7067,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7286916602907422,
+      "grad_norm": 2.4361904893888395,
+      "learning_rate": 1.4689293103317456e-05,
+      "loss": 0.7469,
+      "step": 4762
+    },
+    {
+      "epoch": 0.7288446824789594,
+      "grad_norm": 2.031659778568357,
+      "learning_rate": 1.4687104139604058e-05,
+      "loss": 0.6313,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7289977046671767,
+      "grad_norm": 2.0556376938872942,
+      "learning_rate": 1.468491488803726e-05,
+      "loss": 0.8087,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7291507268553941,
+      "grad_norm": 2.565523844865347,
+      "learning_rate": 1.468272534875151e-05,
+      "loss": 0.8579,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7293037490436113,
+      "grad_norm": 2.363770120524518,
+      "learning_rate": 1.4680535521881277e-05,
+      "loss": 0.7409,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7294567712318286,
+      "grad_norm": 2.318947990020625,
+      "learning_rate": 1.4678345407561046e-05,
+      "loss": 0.7089,
+      "step": 4767
+    },
+    {
+      "epoch": 0.729609793420046,
+      "grad_norm": 2.1665273836897954,
+      "learning_rate": 1.4676155005925323e-05,
+      "loss": 0.6989,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7297628156082632,
+      "grad_norm": 2.282864821091944,
+      "learning_rate": 1.4673964317108627e-05,
+      "loss": 0.6375,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7299158377964805,
+      "grad_norm": 2.1550424071289473,
+      "learning_rate": 1.4671773341245499e-05,
+      "loss": 0.7207,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7300688599846977,
+      "grad_norm": 2.1784007795790505,
+      "learning_rate": 1.4669582078470494e-05,
+      "loss": 0.6942,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7302218821729151,
+      "grad_norm": 2.2050045587066025,
+      "learning_rate": 1.4667390528918186e-05,
+      "loss": 0.7414,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7303749043611324,
+      "grad_norm": 2.107735164826548,
+      "learning_rate": 1.4665198692723166e-05,
+      "loss": 0.7075,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7305279265493496,
+      "grad_norm": 2.145877315751559,
+      "learning_rate": 1.4663006570020044e-05,
+      "loss": 0.7179,
+      "step": 4774
+    },
+    {
+      "epoch": 0.730680948737567,
+      "grad_norm": 2.282915199457753,
+      "learning_rate": 1.4660814160943448e-05,
+      "loss": 0.7671,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7308339709257843,
+      "grad_norm": 2.0724299524788115,
+      "learning_rate": 1.465862146562802e-05,
+      "loss": 0.6496,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7309869931140015,
+      "grad_norm": 2.36617080210269,
+      "learning_rate": 1.4656428484208423e-05,
+      "loss": 0.7133,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7311400153022188,
+      "grad_norm": 2.3154604337397346,
+      "learning_rate": 1.4654235216819337e-05,
+      "loss": 0.6548,
+      "step": 4778
+    },
+    {
+      "epoch": 0.7312930374904361,
+      "grad_norm": 2.284140303142969,
+      "learning_rate": 1.4652041663595457e-05,
+      "loss": 0.6584,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7314460596786534,
+      "grad_norm": 2.549272830669625,
+      "learning_rate": 1.4649847824671503e-05,
+      "loss": 0.7738,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7315990818668707,
+      "grad_norm": 2.3716627589861474,
+      "learning_rate": 1.4647653700182203e-05,
+      "loss": 0.6948,
+      "step": 4781
+    },
+    {
+      "epoch": 0.731752104055088,
+      "grad_norm": 2.254070886426423,
+      "learning_rate": 1.4645459290262304e-05,
+      "loss": 0.7778,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7319051262433053,
+      "grad_norm": 2.4077618806749204,
+      "learning_rate": 1.464326459504658e-05,
+      "loss": 0.7467,
+      "step": 4783
+    },
+    {
+      "epoch": 0.7320581484315226,
+      "grad_norm": 2.4854494866694794,
+      "learning_rate": 1.4641069614669814e-05,
+      "loss": 0.799,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7322111706197398,
+      "grad_norm": 2.097663474137104,
+      "learning_rate": 1.4638874349266806e-05,
+      "loss": 0.7088,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7323641928079572,
+      "grad_norm": 2.2978940941367743,
+      "learning_rate": 1.4636678798972374e-05,
+      "loss": 0.6974,
+      "step": 4786
+    },
+    {
+      "epoch": 0.7325172149961744,
+      "grad_norm": 2.2559176492956676,
+      "learning_rate": 1.4634482963921362e-05,
+      "loss": 0.7634,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7326702371843917,
+      "grad_norm": 2.29409352342819,
+      "learning_rate": 1.4632286844248618e-05,
+      "loss": 0.677,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7328232593726091,
+      "grad_norm": 2.186287939404946,
+      "learning_rate": 1.463009044008902e-05,
+      "loss": 0.7605,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7329762815608263,
+      "grad_norm": 2.487897472414095,
+      "learning_rate": 1.4627893751577454e-05,
+      "loss": 0.7555,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7331293037490436,
+      "grad_norm": 2.1811604114846115,
+      "learning_rate": 1.4625696778848826e-05,
+      "loss": 0.7005,
+      "step": 4791
+    },
+    {
+      "epoch": 0.733282325937261,
+      "grad_norm": 2.344426014732196,
+      "learning_rate": 1.4623499522038064e-05,
+      "loss": 0.7748,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7334353481254782,
+      "grad_norm": 2.344890853831977,
+      "learning_rate": 1.4621301981280112e-05,
+      "loss": 0.8042,
+      "step": 4793
+    },
+    {
+      "epoch": 0.7335883703136955,
+      "grad_norm": 1.9922549418646183,
+      "learning_rate": 1.4619104156709924e-05,
+      "loss": 0.6517,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7337413925019127,
+      "grad_norm": 2.0746510135183365,
+      "learning_rate": 1.461690604846248e-05,
+      "loss": 0.5896,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7338944146901301,
+      "grad_norm": 2.168429526773284,
+      "learning_rate": 1.4614707656672775e-05,
+      "loss": 0.7213,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7340474368783474,
+      "grad_norm": 2.317615047962928,
+      "learning_rate": 1.4612508981475819e-05,
+      "loss": 0.7598,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7342004590665646,
+      "grad_norm": 2.000689298049806,
+      "learning_rate": 1.4610310023006643e-05,
+      "loss": 0.67,
+      "step": 4798
+    },
+    {
+      "epoch": 0.734353481254782,
+      "grad_norm": 2.340228400641039,
+      "learning_rate": 1.4608110781400293e-05,
+      "loss": 0.7711,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7345065034429993,
+      "grad_norm": 2.044964571461997,
+      "learning_rate": 1.4605911256791831e-05,
+      "loss": 0.6814,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7346595256312165,
+      "grad_norm": 2.3241346893015478,
+      "learning_rate": 1.4603711449316342e-05,
+      "loss": 0.7522,
+      "step": 4801
+    },
+    {
+      "epoch": 0.7348125478194338,
+      "grad_norm": 2.3624868394174325,
+      "learning_rate": 1.4601511359108924e-05,
+      "loss": 0.7272,
+      "step": 4802
+    },
+    {
+      "epoch": 0.7349655700076511,
+      "grad_norm": 2.475081517037485,
+      "learning_rate": 1.4599310986304691e-05,
+      "loss": 0.735,
+      "step": 4803
+    },
+    {
+      "epoch": 0.7351185921958684,
+      "grad_norm": 2.2359477222481865,
+      "learning_rate": 1.4597110331038775e-05,
+      "loss": 0.6041,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7352716143840857,
+      "grad_norm": 2.0780513139500374,
+      "learning_rate": 1.4594909393446334e-05,
+      "loss": 0.7102,
+      "step": 4805
+    },
+    {
+      "epoch": 0.735424636572303,
+      "grad_norm": 2.0920809638081943,
+      "learning_rate": 1.459270817366253e-05,
+      "loss": 0.7028,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7355776587605203,
+      "grad_norm": 2.222222916454537,
+      "learning_rate": 1.459050667182255e-05,
+      "loss": 0.5553,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7357306809487376,
+      "grad_norm": 2.2204854583000886,
+      "learning_rate": 1.4588304888061597e-05,
+      "loss": 0.6874,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7358837031369548,
+      "grad_norm": 2.2720313280348208,
+      "learning_rate": 1.4586102822514896e-05,
+      "loss": 0.7076,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7360367253251722,
+      "grad_norm": 2.387822079584415,
+      "learning_rate": 1.458390047531767e-05,
+      "loss": 0.7585,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7361897475133894,
+      "grad_norm": 2.428520327068487,
+      "learning_rate": 1.4581697846605192e-05,
+      "loss": 0.7851,
+      "step": 4811
+    },
+    {
+      "epoch": 0.7363427697016067,
+      "grad_norm": 2.1643869130717475,
+      "learning_rate": 1.4579494936512722e-05,
+      "loss": 0.7187,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7364957918898241,
+      "grad_norm": 2.2809620466631193,
+      "learning_rate": 1.4577291745175555e-05,
+      "loss": 0.6697,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7366488140780413,
+      "grad_norm": 2.015931630069814,
+      "learning_rate": 1.4575088272728992e-05,
+      "loss": 0.7241,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7368018362662586,
+      "grad_norm": 2.48264062580387,
+      "learning_rate": 1.4572884519308363e-05,
+      "loss": 0.7079,
+      "step": 4815
+    },
+    {
+      "epoch": 0.736954858454476,
+      "grad_norm": 2.5497171898501088,
+      "learning_rate": 1.4570680485049007e-05,
+      "loss": 0.7401,
+      "step": 4816
+    },
+    {
+      "epoch": 0.7371078806426932,
+      "grad_norm": 2.2285617237271182,
+      "learning_rate": 1.4568476170086281e-05,
+      "loss": 0.6802,
+      "step": 4817
+    },
+    {
+      "epoch": 0.7372609028309105,
+      "grad_norm": 2.4025875558068925,
+      "learning_rate": 1.4566271574555559e-05,
+      "loss": 0.6301,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7374139250191277,
+      "grad_norm": 2.4112017394461644,
+      "learning_rate": 1.4564066698592238e-05,
+      "loss": 0.6759,
+      "step": 4819
+    },
+    {
+      "epoch": 0.7375669472073451,
+      "grad_norm": 2.195257542503888,
+      "learning_rate": 1.4561861542331725e-05,
+      "loss": 0.6785,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7377199693955624,
+      "grad_norm": 2.134046236215962,
+      "learning_rate": 1.4559656105909449e-05,
+      "loss": 0.66,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7378729915837796,
+      "grad_norm": 1.7848717620286476,
+      "learning_rate": 1.4557450389460859e-05,
+      "loss": 0.592,
+      "step": 4822
+    },
+    {
+      "epoch": 0.738026013771997,
+      "grad_norm": 2.1999375705173656,
+      "learning_rate": 1.4555244393121406e-05,
+      "loss": 0.6376,
+      "step": 4823
+    },
+    {
+      "epoch": 0.7381790359602143,
+      "grad_norm": 2.192595297089319,
+      "learning_rate": 1.4553038117026577e-05,
+      "loss": 0.6949,
+      "step": 4824
+    },
+    {
+      "epoch": 0.7383320581484315,
+      "grad_norm": 2.5214156134793764,
+      "learning_rate": 1.4550831561311864e-05,
+      "loss": 0.8018,
+      "step": 4825
+    },
+    {
+      "epoch": 0.7384850803366488,
+      "grad_norm": 2.8632598195698655,
+      "learning_rate": 1.4548624726112782e-05,
+      "loss": 0.7681,
+      "step": 4826
+    },
+    {
+      "epoch": 0.7386381025248661,
+      "grad_norm": 2.310980789411813,
+      "learning_rate": 1.4546417611564864e-05,
+      "loss": 0.7135,
+      "step": 4827
+    },
+    {
+      "epoch": 0.7387911247130834,
+      "grad_norm": 2.324914918749278,
+      "learning_rate": 1.4544210217803651e-05,
+      "loss": 0.6451,
+      "step": 4828
+    },
+    {
+      "epoch": 0.7389441469013007,
+      "grad_norm": 2.2455011773034212,
+      "learning_rate": 1.4542002544964713e-05,
+      "loss": 0.7251,
+      "step": 4829
+    },
+    {
+      "epoch": 0.739097169089518,
+      "grad_norm": 2.3997081873057353,
+      "learning_rate": 1.4539794593183634e-05,
+      "loss": 0.6291,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7392501912777353,
+      "grad_norm": 2.4566939192312356,
+      "learning_rate": 1.4537586362596005e-05,
+      "loss": 0.7435,
+      "step": 4831
+    },
+    {
+      "epoch": 0.7394032134659526,
+      "grad_norm": 2.4455663976068145,
+      "learning_rate": 1.4535377853337448e-05,
+      "loss": 0.6491,
+      "step": 4832
+    },
+    {
+      "epoch": 0.7395562356541698,
+      "grad_norm": 2.1715569659281044,
+      "learning_rate": 1.4533169065543593e-05,
+      "loss": 0.6749,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7397092578423872,
+      "grad_norm": 2.291412184131578,
+      "learning_rate": 1.4530959999350095e-05,
+      "loss": 0.643,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7398622800306044,
+      "grad_norm": 2.204427629847068,
+      "learning_rate": 1.4528750654892614e-05,
+      "loss": 0.8016,
+      "step": 4835
+    },
+    {
+      "epoch": 0.7400153022188217,
+      "grad_norm": 1.9802730994342306,
+      "learning_rate": 1.452654103230684e-05,
+      "loss": 0.7306,
+      "step": 4836
+    },
+    {
+      "epoch": 0.7401683244070391,
+      "grad_norm": 2.4833975120925977,
+      "learning_rate": 1.4524331131728472e-05,
+      "loss": 0.7693,
+      "step": 4837
+    },
+    {
+      "epoch": 0.7403213465952563,
+      "grad_norm": 2.2973589798555634,
+      "learning_rate": 1.4522120953293233e-05,
+      "loss": 0.6689,
+      "step": 4838
+    },
+    {
+      "epoch": 0.7404743687834736,
+      "grad_norm": 2.025223238613172,
+      "learning_rate": 1.4519910497136856e-05,
+      "loss": 0.6473,
+      "step": 4839
+    },
+    {
+      "epoch": 0.740627390971691,
+      "grad_norm": 2.348915095079573,
+      "learning_rate": 1.451769976339509e-05,
+      "loss": 0.8239,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7407804131599082,
+      "grad_norm": 2.309855719910271,
+      "learning_rate": 1.4515488752203713e-05,
+      "loss": 0.7992,
+      "step": 4841
+    },
+    {
+      "epoch": 0.7409334353481255,
+      "grad_norm": 2.206226950757133,
+      "learning_rate": 1.4513277463698503e-05,
+      "loss": 0.6737,
+      "step": 4842
+    },
+    {
+      "epoch": 0.7410864575363427,
+      "grad_norm": 2.2352741297777756,
+      "learning_rate": 1.4511065898015269e-05,
+      "loss": 0.7212,
+      "step": 4843
+    },
+    {
+      "epoch": 0.7412394797245601,
+      "grad_norm": 2.38240113752834,
+      "learning_rate": 1.450885405528983e-05,
+      "loss": 0.8639,
+      "step": 4844
+    },
+    {
+      "epoch": 0.7413925019127774,
+      "grad_norm": 2.2151203809554496,
+      "learning_rate": 1.4506641935658023e-05,
+      "loss": 0.7709,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7415455241009946,
+      "grad_norm": 2.156453587248878,
+      "learning_rate": 1.4504429539255708e-05,
+      "loss": 0.7863,
+      "step": 4846
+    },
+    {
+      "epoch": 0.741698546289212,
+      "grad_norm": 2.139602715597007,
+      "learning_rate": 1.4502216866218755e-05,
+      "loss": 0.7202,
+      "step": 4847
+    },
+    {
+      "epoch": 0.7418515684774293,
+      "grad_norm": 2.2784471326319156,
+      "learning_rate": 1.4500003916683047e-05,
+      "loss": 0.8374,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7420045906656465,
+      "grad_norm": 2.3563875101559217,
+      "learning_rate": 1.4497790690784498e-05,
+      "loss": 0.6719,
+      "step": 4849
+    },
+    {
+      "epoch": 0.7421576128538638,
+      "grad_norm": 2.2202100939612825,
+      "learning_rate": 1.4495577188659028e-05,
+      "loss": 0.6286,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7423106350420811,
+      "grad_norm": 2.5782599669858586,
+      "learning_rate": 1.449336341044257e-05,
+      "loss": 0.7125,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7424636572302984,
+      "grad_norm": 2.374270686758012,
+      "learning_rate": 1.4491149356271092e-05,
+      "loss": 0.6643,
+      "step": 4852
+    },
+    {
+      "epoch": 0.7426166794185157,
+      "grad_norm": 2.5637372901568094,
+      "learning_rate": 1.4488935026280561e-05,
+      "loss": 0.6725,
+      "step": 4853
+    },
+    {
+      "epoch": 0.742769701606733,
+      "grad_norm": 2.1071033399459136,
+      "learning_rate": 1.448672042060697e-05,
+      "loss": 0.676,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7429227237949503,
+      "grad_norm": 2.4993248270173747,
+      "learning_rate": 1.4484505539386324e-05,
+      "loss": 0.7052,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7430757459831676,
+      "grad_norm": 2.076983925023967,
+      "learning_rate": 1.448229038275465e-05,
+      "loss": 0.6189,
+      "step": 4856
+    },
+    {
+      "epoch": 0.7432287681713848,
+      "grad_norm": 2.433934235940275,
+      "learning_rate": 1.4480074950847991e-05,
+      "loss": 0.8599,
+      "step": 4857
+    },
+    {
+      "epoch": 0.7433817903596022,
+      "grad_norm": 2.2124457746802637,
+      "learning_rate": 1.4477859243802401e-05,
+      "loss": 0.7823,
+      "step": 4858
+    },
+    {
+      "epoch": 0.7435348125478194,
+      "grad_norm": 2.570932508736398,
+      "learning_rate": 1.447564326175396e-05,
+      "loss": 0.8181,
+      "step": 4859
+    },
+    {
+      "epoch": 0.7436878347360367,
+      "grad_norm": 2.4117017713360904,
+      "learning_rate": 1.4473427004838754e-05,
+      "loss": 0.7226,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7438408569242541,
+      "grad_norm": 2.176505200952037,
+      "learning_rate": 1.4471210473192896e-05,
+      "loss": 0.7055,
+      "step": 4861
+    },
+    {
+      "epoch": 0.7439938791124713,
+      "grad_norm": 2.1857445379147755,
+      "learning_rate": 1.4468993666952512e-05,
+      "loss": 0.6834,
+      "step": 4862
+    },
+    {
+      "epoch": 0.7441469013006886,
+      "grad_norm": 2.635693368966689,
+      "learning_rate": 1.4466776586253745e-05,
+      "loss": 0.8412,
+      "step": 4863
+    },
+    {
+      "epoch": 0.744299923488906,
+      "grad_norm": 2.0451933975323193,
+      "learning_rate": 1.446455923123275e-05,
+      "loss": 0.6854,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7444529456771232,
+      "grad_norm": 2.275358040160873,
+      "learning_rate": 1.4462341602025714e-05,
+      "loss": 0.7035,
+      "step": 4865
+    },
+    {
+      "epoch": 0.7446059678653405,
+      "grad_norm": 2.4710038242202956,
+      "learning_rate": 1.4460123698768817e-05,
+      "loss": 0.8401,
+      "step": 4866
+    },
+    {
+      "epoch": 0.7447589900535577,
+      "grad_norm": 2.4289044690905897,
+      "learning_rate": 1.4457905521598279e-05,
+      "loss": 0.7351,
+      "step": 4867
+    },
+    {
+      "epoch": 0.7449120122417751,
+      "grad_norm": 2.31025768005515,
+      "learning_rate": 1.4455687070650324e-05,
+      "loss": 0.782,
+      "step": 4868
+    },
+    {
+      "epoch": 0.7450650344299924,
+      "grad_norm": 2.1544459445175232,
+      "learning_rate": 1.4453468346061193e-05,
+      "loss": 0.7421,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7452180566182096,
+      "grad_norm": 2.3447105605966807,
+      "learning_rate": 1.445124934796715e-05,
+      "loss": 0.7559,
+      "step": 4870
+    },
+    {
+      "epoch": 0.745371078806427,
+      "grad_norm": 2.239958777557601,
+      "learning_rate": 1.444903007650447e-05,
+      "loss": 0.6981,
+      "step": 4871
+    },
+    {
+      "epoch": 0.7455241009946443,
+      "grad_norm": 2.249734548223549,
+      "learning_rate": 1.444681053180945e-05,
+      "loss": 0.751,
+      "step": 4872
+    },
+    {
+      "epoch": 0.7456771231828615,
+      "grad_norm": 2.340819325628896,
+      "learning_rate": 1.4444590714018398e-05,
+      "loss": 0.7303,
+      "step": 4873
+    },
+    {
+      "epoch": 0.7458301453710788,
+      "grad_norm": 2.34414004865152,
+      "learning_rate": 1.4442370623267643e-05,
+      "loss": 0.7181,
+      "step": 4874
+    },
+    {
+      "epoch": 0.7459831675592961,
+      "grad_norm": 3.0286066038526562,
+      "learning_rate": 1.444015025969353e-05,
+      "loss": 0.831,
+      "step": 4875
+    },
+    {
+      "epoch": 0.7461361897475134,
+      "grad_norm": 2.425379330178627,
+      "learning_rate": 1.4437929623432418e-05,
+      "loss": 0.7453,
+      "step": 4876
+    },
+    {
+      "epoch": 0.7462892119357307,
+      "grad_norm": 2.228493718435368,
+      "learning_rate": 1.443570871462069e-05,
+      "loss": 0.6826,
+      "step": 4877
+    },
+    {
+      "epoch": 0.746442234123948,
+      "grad_norm": 2.1177422595743685,
+      "learning_rate": 1.4433487533394734e-05,
+      "loss": 0.6157,
+      "step": 4878
+    },
+    {
+      "epoch": 0.7465952563121653,
+      "grad_norm": 2.0833369615704513,
+      "learning_rate": 1.4431266079890964e-05,
+      "loss": 0.5789,
+      "step": 4879
+    },
+    {
+      "epoch": 0.7467482785003825,
+      "grad_norm": 2.8278316899269873,
+      "learning_rate": 1.4429044354245812e-05,
+      "loss": 0.7478,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7469013006885998,
+      "grad_norm": 2.478404645618318,
+      "learning_rate": 1.4426822356595718e-05,
+      "loss": 0.6872,
+      "step": 4881
+    },
+    {
+      "epoch": 0.7470543228768172,
+      "grad_norm": 2.3329379350497508,
+      "learning_rate": 1.4424600087077148e-05,
+      "loss": 0.7983,
+      "step": 4882
+    },
+    {
+      "epoch": 0.7472073450650344,
+      "grad_norm": 2.2267621885651003,
+      "learning_rate": 1.4422377545826574e-05,
+      "loss": 0.7222,
+      "step": 4883
+    },
+    {
+      "epoch": 0.7473603672532517,
+      "grad_norm": 2.4536454346196424,
+      "learning_rate": 1.4420154732980493e-05,
+      "loss": 0.7625,
+      "step": 4884
+    },
+    {
+      "epoch": 0.7475133894414691,
+      "grad_norm": 2.022708265593278,
+      "learning_rate": 1.4417931648675423e-05,
+      "loss": 0.7153,
+      "step": 4885
+    },
+    {
+      "epoch": 0.7476664116296863,
+      "grad_norm": 2.256460123168098,
+      "learning_rate": 1.4415708293047884e-05,
+      "loss": 0.7022,
+      "step": 4886
+    },
+    {
+      "epoch": 0.7478194338179036,
+      "grad_norm": 2.136078642899449,
+      "learning_rate": 1.4413484666234427e-05,
+      "loss": 0.6409,
+      "step": 4887
+    },
+    {
+      "epoch": 0.7479724560061208,
+      "grad_norm": 2.2325055404407994,
+      "learning_rate": 1.4411260768371609e-05,
+      "loss": 0.729,
+      "step": 4888
+    },
+    {
+      "epoch": 0.7481254781943382,
+      "grad_norm": 2.3585569399038695,
+      "learning_rate": 1.440903659959601e-05,
+      "loss": 0.7476,
+      "step": 4889
+    },
+    {
+      "epoch": 0.7482785003825555,
+      "grad_norm": 2.3183366182918888,
+      "learning_rate": 1.4406812160044226e-05,
+      "loss": 0.7876,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7484315225707727,
+      "grad_norm": 2.4632969026460687,
+      "learning_rate": 1.4404587449852865e-05,
+      "loss": 0.7183,
+      "step": 4891
+    },
+    {
+      "epoch": 0.7485845447589901,
+      "grad_norm": 2.265010236192547,
+      "learning_rate": 1.4402362469158562e-05,
+      "loss": 0.7735,
+      "step": 4892
+    },
+    {
+      "epoch": 0.7487375669472074,
+      "grad_norm": 2.2999201065686266,
+      "learning_rate": 1.4400137218097956e-05,
+      "loss": 0.7304,
+      "step": 4893
+    },
+    {
+      "epoch": 0.7488905891354246,
+      "grad_norm": 2.2799436177017216,
+      "learning_rate": 1.4397911696807708e-05,
+      "loss": 0.7187,
+      "step": 4894
+    },
+    {
+      "epoch": 0.749043611323642,
+      "grad_norm": 2.133201532746618,
+      "learning_rate": 1.43956859054245e-05,
+      "loss": 0.735,
+      "step": 4895
+    },
+    {
+      "epoch": 0.7491966335118592,
+      "grad_norm": 2.3822318079150535,
+      "learning_rate": 1.4393459844085027e-05,
+      "loss": 0.751,
+      "step": 4896
+    },
+    {
+      "epoch": 0.7493496557000765,
+      "grad_norm": 2.276622654679582,
+      "learning_rate": 1.4391233512925994e-05,
+      "loss": 0.6863,
+      "step": 4897
+    },
+    {
+      "epoch": 0.7495026778882938,
+      "grad_norm": 2.4161787604889193,
+      "learning_rate": 1.4389006912084136e-05,
+      "loss": 0.8374,
+      "step": 4898
+    },
+    {
+      "epoch": 0.7496557000765111,
+      "grad_norm": 2.27888042381393,
+      "learning_rate": 1.4386780041696189e-05,
+      "loss": 0.6775,
+      "step": 4899
+    },
+    {
+      "epoch": 0.7498087222647284,
+      "grad_norm": 2.520476013025121,
+      "learning_rate": 1.438455290189892e-05,
+      "loss": 0.7428,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7499617444529457,
+      "grad_norm": 2.232440986785114,
+      "learning_rate": 1.4382325492829108e-05,
+      "loss": 0.6929,
+      "step": 4901
+    },
+    {
+      "epoch": 0.7501147666411629,
+      "grad_norm": 2.2194515852033816,
+      "learning_rate": 1.4380097814623539e-05,
+      "loss": 0.7759,
+      "step": 4902
+    },
+    {
+      "epoch": 0.7502677888293803,
+      "grad_norm": 2.493029827031643,
+      "learning_rate": 1.437786986741903e-05,
+      "loss": 0.7942,
+      "step": 4903
+    },
+    {
+      "epoch": 0.7504208110175975,
+      "grad_norm": 2.532507217587678,
+      "learning_rate": 1.4375641651352408e-05,
+      "loss": 0.7536,
+      "step": 4904
+    },
+    {
+      "epoch": 0.7505738332058148,
+      "grad_norm": 2.1324616550429343,
+      "learning_rate": 1.4373413166560512e-05,
+      "loss": 0.6899,
+      "step": 4905
+    },
+    {
+      "epoch": 0.7507268553940322,
+      "grad_norm": 2.404454010344963,
+      "learning_rate": 1.4371184413180205e-05,
+      "loss": 0.7995,
+      "step": 4906
+    },
+    {
+      "epoch": 0.7508798775822494,
+      "grad_norm": 2.212175857528018,
+      "learning_rate": 1.4368955391348366e-05,
+      "loss": 0.5926,
+      "step": 4907
+    },
+    {
+      "epoch": 0.7510328997704667,
+      "grad_norm": 2.155592680667354,
+      "learning_rate": 1.436672610120188e-05,
+      "loss": 0.6284,
+      "step": 4908
+    },
+    {
+      "epoch": 0.751185921958684,
+      "grad_norm": 2.3741727185360832,
+      "learning_rate": 1.4364496542877663e-05,
+      "loss": 0.7462,
+      "step": 4909
+    },
+    {
+      "epoch": 0.7513389441469013,
+      "grad_norm": 2.1469671874999325,
+      "learning_rate": 1.4362266716512643e-05,
+      "loss": 0.7229,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7514919663351186,
+      "grad_norm": 2.121527275206317,
+      "learning_rate": 1.4360036622243754e-05,
+      "loss": 0.7239,
+      "step": 4911
+    },
+    {
+      "epoch": 0.7516449885233358,
+      "grad_norm": 2.2376730417194812,
+      "learning_rate": 1.4357806260207962e-05,
+      "loss": 0.696,
+      "step": 4912
+    },
+    {
+      "epoch": 0.7517980107115532,
+      "grad_norm": 2.2821150028836574,
+      "learning_rate": 1.435557563054224e-05,
+      "loss": 0.785,
+      "step": 4913
+    },
+    {
+      "epoch": 0.7519510328997705,
+      "grad_norm": 2.0970063152782066,
+      "learning_rate": 1.4353344733383576e-05,
+      "loss": 0.7399,
+      "step": 4914
+    },
+    {
+      "epoch": 0.7521040550879877,
+      "grad_norm": 2.083207329954238,
+      "learning_rate": 1.4351113568868987e-05,
+      "loss": 0.5428,
+      "step": 4915
+    },
+    {
+      "epoch": 0.752257077276205,
+      "grad_norm": 2.245970612654284,
+      "learning_rate": 1.434888213713549e-05,
+      "loss": 0.7392,
+      "step": 4916
+    },
+    {
+      "epoch": 0.7524100994644224,
+      "grad_norm": 2.3609964289459113,
+      "learning_rate": 1.4346650438320124e-05,
+      "loss": 0.6814,
+      "step": 4917
+    },
+    {
+      "epoch": 0.7525631216526396,
+      "grad_norm": 2.1525419459253556,
+      "learning_rate": 1.4344418472559956e-05,
+      "loss": 0.6854,
+      "step": 4918
+    },
+    {
+      "epoch": 0.7527161438408569,
+      "grad_norm": 2.330687789914893,
+      "learning_rate": 1.434218623999205e-05,
+      "loss": 0.6747,
+      "step": 4919
+    },
+    {
+      "epoch": 0.7528691660290742,
+      "grad_norm": 2.337433181523013,
+      "learning_rate": 1.4339953740753502e-05,
+      "loss": 0.6729,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7530221882172915,
+      "grad_norm": 2.165023230727938,
+      "learning_rate": 1.4337720974981417e-05,
+      "loss": 0.6102,
+      "step": 4921
+    },
+    {
+      "epoch": 0.7531752104055088,
+      "grad_norm": 2.095474532830923,
+      "learning_rate": 1.4335487942812919e-05,
+      "loss": 0.7267,
+      "step": 4922
+    },
+    {
+      "epoch": 0.753328232593726,
+      "grad_norm": 2.1802082090993777,
+      "learning_rate": 1.4333254644385144e-05,
+      "loss": 0.6848,
+      "step": 4923
+    },
+    {
+      "epoch": 0.7534812547819434,
+      "grad_norm": 2.6591708663481715,
+      "learning_rate": 1.4331021079835249e-05,
+      "loss": 0.8153,
+      "step": 4924
+    },
+    {
+      "epoch": 0.7536342769701607,
+      "grad_norm": 2.3136679846972665,
+      "learning_rate": 1.432878724930041e-05,
+      "loss": 0.69,
+      "step": 4925
+    },
+    {
+      "epoch": 0.7537872991583779,
+      "grad_norm": 2.2807435166362113,
+      "learning_rate": 1.4326553152917808e-05,
+      "loss": 0.6869,
+      "step": 4926
+    },
+    {
+      "epoch": 0.7539403213465953,
+      "grad_norm": 2.1228252902705402,
+      "learning_rate": 1.432431879082465e-05,
+      "loss": 0.7477,
+      "step": 4927
+    },
+    {
+      "epoch": 0.7540933435348125,
+      "grad_norm": 2.4278219077579473,
+      "learning_rate": 1.432208416315816e-05,
+      "loss": 0.7974,
+      "step": 4928
+    },
+    {
+      "epoch": 0.7542463657230298,
+      "grad_norm": 2.5232419497439396,
+      "learning_rate": 1.4319849270055576e-05,
+      "loss": 0.732,
+      "step": 4929
+    },
+    {
+      "epoch": 0.7543993879112472,
+      "grad_norm": 2.0991061394057406,
+      "learning_rate": 1.4317614111654145e-05,
+      "loss": 0.7423,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7545524100994644,
+      "grad_norm": 2.3815083789734164,
+      "learning_rate": 1.4315378688091143e-05,
+      "loss": 0.9114,
+      "step": 4931
+    },
+    {
+      "epoch": 0.7547054322876817,
+      "grad_norm": 2.185730306642332,
+      "learning_rate": 1.4313142999503855e-05,
+      "loss": 0.7058,
+      "step": 4932
+    },
+    {
+      "epoch": 0.754858454475899,
+      "grad_norm": 2.1376989838870113,
+      "learning_rate": 1.4310907046029581e-05,
+      "loss": 0.702,
+      "step": 4933
+    },
+    {
+      "epoch": 0.7550114766641163,
+      "grad_norm": 2.185315720970579,
+      "learning_rate": 1.4308670827805642e-05,
+      "loss": 0.6436,
+      "step": 4934
+    },
+    {
+      "epoch": 0.7551644988523336,
+      "grad_norm": 2.325819141198931,
+      "learning_rate": 1.430643434496937e-05,
+      "loss": 0.7432,
+      "step": 4935
+    },
+    {
+      "epoch": 0.7553175210405508,
+      "grad_norm": 2.022900910011311,
+      "learning_rate": 1.4304197597658119e-05,
+      "loss": 0.7569,
+      "step": 4936
+    },
+    {
+      "epoch": 0.7554705432287682,
+      "grad_norm": 2.2978717235749917,
+      "learning_rate": 1.4301960586009255e-05,
+      "loss": 0.8398,
+      "step": 4937
+    },
+    {
+      "epoch": 0.7556235654169855,
+      "grad_norm": 2.0323787781258114,
+      "learning_rate": 1.4299723310160163e-05,
+      "loss": 0.8232,
+      "step": 4938
+    },
+    {
+      "epoch": 0.7557765876052027,
+      "grad_norm": 2.4687308360656344,
+      "learning_rate": 1.429748577024824e-05,
+      "loss": 0.7811,
+      "step": 4939
+    },
+    {
+      "epoch": 0.75592960979342,
+      "grad_norm": 2.416728876586808,
+      "learning_rate": 1.4295247966410903e-05,
+      "loss": 0.7862,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7560826319816374,
+      "grad_norm": 2.430436233578812,
+      "learning_rate": 1.429300989878559e-05,
+      "loss": 0.7025,
+      "step": 4941
+    },
+    {
+      "epoch": 0.7562356541698546,
+      "grad_norm": 2.2106978733410556,
+      "learning_rate": 1.4290771567509745e-05,
+      "loss": 0.6891,
+      "step": 4942
+    },
+    {
+      "epoch": 0.7563886763580719,
+      "grad_norm": 2.4306496637568484,
+      "learning_rate": 1.4288532972720825e-05,
+      "loss": 0.8306,
+      "step": 4943
+    },
+    {
+      "epoch": 0.7565416985462892,
+      "grad_norm": 2.469829207837791,
+      "learning_rate": 1.4286294114556325e-05,
+      "loss": 0.692,
+      "step": 4944
+    },
+    {
+      "epoch": 0.7566947207345065,
+      "grad_norm": 2.4309218792187273,
+      "learning_rate": 1.4284054993153735e-05,
+      "loss": 0.7137,
+      "step": 4945
+    },
+    {
+      "epoch": 0.7568477429227238,
+      "grad_norm": 2.579777669684564,
+      "learning_rate": 1.4281815608650565e-05,
+      "loss": 0.7477,
+      "step": 4946
+    },
+    {
+      "epoch": 0.757000765110941,
+      "grad_norm": 2.0971978445223227,
+      "learning_rate": 1.4279575961184348e-05,
+      "loss": 0.6919,
+      "step": 4947
+    },
+    {
+      "epoch": 0.7571537872991584,
+      "grad_norm": 2.2553951720139422,
+      "learning_rate": 1.4277336050892631e-05,
+      "loss": 0.7135,
+      "step": 4948
+    },
+    {
+      "epoch": 0.7573068094873757,
+      "grad_norm": 2.149299088970806,
+      "learning_rate": 1.4275095877912976e-05,
+      "loss": 0.6381,
+      "step": 4949
+    },
+    {
+      "epoch": 0.7574598316755929,
+      "grad_norm": 2.1629746985410523,
+      "learning_rate": 1.4272855442382957e-05,
+      "loss": 0.6806,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7576128538638103,
+      "grad_norm": 2.3261663770030157,
+      "learning_rate": 1.427061474444017e-05,
+      "loss": 0.6916,
+      "step": 4951
+    },
+    {
+      "epoch": 0.7577658760520275,
+      "grad_norm": 2.1499157181080566,
+      "learning_rate": 1.4268373784222225e-05,
+      "loss": 0.676,
+      "step": 4952
+    },
+    {
+      "epoch": 0.7579188982402448,
+      "grad_norm": 2.4744089486881293,
+      "learning_rate": 1.4266132561866747e-05,
+      "loss": 0.8978,
+      "step": 4953
+    },
+    {
+      "epoch": 0.7580719204284622,
+      "grad_norm": 2.168743668047201,
+      "learning_rate": 1.4263891077511383e-05,
+      "loss": 0.6623,
+      "step": 4954
+    },
+    {
+      "epoch": 0.7582249426166794,
+      "grad_norm": 2.3449514837783063,
+      "learning_rate": 1.4261649331293781e-05,
+      "loss": 0.6634,
+      "step": 4955
+    },
+    {
+      "epoch": 0.7583779648048967,
+      "grad_norm": 2.5034072035723023,
+      "learning_rate": 1.4259407323351626e-05,
+      "loss": 0.7541,
+      "step": 4956
+    },
+    {
+      "epoch": 0.758530986993114,
+      "grad_norm": 2.3696350755990485,
+      "learning_rate": 1.4257165053822605e-05,
+      "loss": 0.6219,
+      "step": 4957
+    },
+    {
+      "epoch": 0.7586840091813313,
+      "grad_norm": 2.3382755188679742,
+      "learning_rate": 1.425492252284442e-05,
+      "loss": 0.7696,
+      "step": 4958
+    },
+    {
+      "epoch": 0.7588370313695486,
+      "grad_norm": 2.135467705544331,
+      "learning_rate": 1.4252679730554801e-05,
+      "loss": 0.7307,
+      "step": 4959
+    },
+    {
+      "epoch": 0.7589900535577658,
+      "grad_norm": 2.408284072064693,
+      "learning_rate": 1.4250436677091482e-05,
+      "loss": 0.7607,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7591430757459832,
+      "grad_norm": 2.2487141060628337,
+      "learning_rate": 1.424819336259222e-05,
+      "loss": 0.7047,
+      "step": 4961
+    },
+    {
+      "epoch": 0.7592960979342005,
+      "grad_norm": 2.174625864768563,
+      "learning_rate": 1.4245949787194783e-05,
+      "loss": 0.7431,
+      "step": 4962
+    },
+    {
+      "epoch": 0.7594491201224177,
+      "grad_norm": 2.3367292934749084,
+      "learning_rate": 1.4243705951036961e-05,
+      "loss": 0.7039,
+      "step": 4963
+    },
+    {
+      "epoch": 0.759602142310635,
+      "grad_norm": 2.3698750474828425,
+      "learning_rate": 1.4241461854256553e-05,
+      "loss": 0.692,
+      "step": 4964
+    },
+    {
+      "epoch": 0.7597551644988524,
+      "grad_norm": 2.3837948494422525,
+      "learning_rate": 1.423921749699138e-05,
+      "loss": 0.682,
+      "step": 4965
+    },
+    {
+      "epoch": 0.7599081866870696,
+      "grad_norm": 1.9568026488800232,
+      "learning_rate": 1.423697287937928e-05,
+      "loss": 0.5177,
+      "step": 4966
+    },
+    {
+      "epoch": 0.7600612088752869,
+      "grad_norm": 2.011750007271983,
+      "learning_rate": 1.4234728001558098e-05,
+      "loss": 0.7043,
+      "step": 4967
+    },
+    {
+      "epoch": 0.7602142310635042,
+      "grad_norm": 2.523991868179148,
+      "learning_rate": 1.42324828636657e-05,
+      "loss": 0.8005,
+      "step": 4968
+    },
+    {
+      "epoch": 0.7603672532517215,
+      "grad_norm": 1.9438899634689,
+      "learning_rate": 1.4230237465839975e-05,
+      "loss": 0.6676,
+      "step": 4969
+    },
+    {
+      "epoch": 0.7605202754399388,
+      "grad_norm": 2.224899265987228,
+      "learning_rate": 1.422799180821882e-05,
+      "loss": 0.6246,
+      "step": 4970
+    },
+    {
+      "epoch": 0.760673297628156,
+      "grad_norm": 2.2936026183785567,
+      "learning_rate": 1.4225745890940145e-05,
+      "loss": 0.7018,
+      "step": 4971
+    },
+    {
+      "epoch": 0.7608263198163734,
+      "grad_norm": 2.283389788103775,
+      "learning_rate": 1.4223499714141885e-05,
+      "loss": 0.7678,
+      "step": 4972
+    },
+    {
+      "epoch": 0.7609793420045907,
+      "grad_norm": 2.1905560444534546,
+      "learning_rate": 1.4221253277961987e-05,
+      "loss": 0.696,
+      "step": 4973
+    },
+    {
+      "epoch": 0.7611323641928079,
+      "grad_norm": 2.2357422103556504,
+      "learning_rate": 1.4219006582538409e-05,
+      "loss": 0.7681,
+      "step": 4974
+    },
+    {
+      "epoch": 0.7612853863810253,
+      "grad_norm": 2.104374866205293,
+      "learning_rate": 1.4216759628009132e-05,
+      "loss": 0.6724,
+      "step": 4975
+    },
+    {
+      "epoch": 0.7614384085692425,
+      "grad_norm": 2.376649516428986,
+      "learning_rate": 1.421451241451215e-05,
+      "loss": 0.7457,
+      "step": 4976
+    },
+    {
+      "epoch": 0.7615914307574598,
+      "grad_norm": 2.5026682947159773,
+      "learning_rate": 1.4212264942185473e-05,
+      "loss": 0.7707,
+      "step": 4977
+    },
+    {
+      "epoch": 0.7617444529456772,
+      "grad_norm": 2.339951422144216,
+      "learning_rate": 1.421001721116713e-05,
+      "loss": 0.7215,
+      "step": 4978
+    },
+    {
+      "epoch": 0.7618974751338944,
+      "grad_norm": 2.5119998784613524,
+      "learning_rate": 1.4207769221595157e-05,
+      "loss": 0.8083,
+      "step": 4979
+    },
+    {
+      "epoch": 0.7620504973221117,
+      "grad_norm": 2.503174353934065,
+      "learning_rate": 1.4205520973607618e-05,
+      "loss": 0.6887,
+      "step": 4980
+    },
+    {
+      "epoch": 0.762203519510329,
+      "grad_norm": 2.4742332404941387,
+      "learning_rate": 1.4203272467342582e-05,
+      "loss": 0.7633,
+      "step": 4981
+    },
+    {
+      "epoch": 0.7623565416985463,
+      "grad_norm": 2.4392406082909655,
+      "learning_rate": 1.4201023702938143e-05,
+      "loss": 0.6757,
+      "step": 4982
+    },
+    {
+      "epoch": 0.7625095638867636,
+      "grad_norm": 2.3098987205477064,
+      "learning_rate": 1.4198774680532403e-05,
+      "loss": 0.7513,
+      "step": 4983
+    },
+    {
+      "epoch": 0.7626625860749808,
+      "grad_norm": 2.258614987166009,
+      "learning_rate": 1.4196525400263482e-05,
+      "loss": 0.7191,
+      "step": 4984
+    },
+    {
+      "epoch": 0.7628156082631982,
+      "grad_norm": 2.3404733220525196,
+      "learning_rate": 1.419427586226952e-05,
+      "loss": 0.799,
+      "step": 4985
+    },
+    {
+      "epoch": 0.7629686304514155,
+      "grad_norm": 2.2779284684505354,
+      "learning_rate": 1.4192026066688673e-05,
+      "loss": 0.8147,
+      "step": 4986
+    },
+    {
+      "epoch": 0.7631216526396327,
+      "grad_norm": 2.31643846075932,
+      "learning_rate": 1.41897760136591e-05,
+      "loss": 0.7568,
+      "step": 4987
+    },
+    {
+      "epoch": 0.76327467482785,
+      "grad_norm": 2.2663948410577337,
+      "learning_rate": 1.4187525703318996e-05,
+      "loss": 0.6588,
+      "step": 4988
+    },
+    {
+      "epoch": 0.7634276970160673,
+      "grad_norm": 2.295491923173907,
+      "learning_rate": 1.4185275135806555e-05,
+      "loss": 0.8117,
+      "step": 4989
+    },
+    {
+      "epoch": 0.7635807192042846,
+      "grad_norm": 2.353699212264283,
+      "learning_rate": 1.4183024311259997e-05,
+      "loss": 0.6288,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7637337413925019,
+      "grad_norm": 2.1939178575915124,
+      "learning_rate": 1.4180773229817548e-05,
+      "loss": 0.6682,
+      "step": 4991
+    },
+    {
+      "epoch": 0.7638867635807192,
+      "grad_norm": 2.2979720867197377,
+      "learning_rate": 1.4178521891617462e-05,
+      "loss": 0.7026,
+      "step": 4992
+    },
+    {
+      "epoch": 0.7640397857689365,
+      "grad_norm": 2.4221273394850487,
+      "learning_rate": 1.4176270296797998e-05,
+      "loss": 0.7267,
+      "step": 4993
+    },
+    {
+      "epoch": 0.7641928079571538,
+      "grad_norm": 2.3662257487773797,
+      "learning_rate": 1.4174018445497439e-05,
+      "loss": 0.7745,
+      "step": 4994
+    },
+    {
+      "epoch": 0.764345830145371,
+      "grad_norm": 2.2759670548636133,
+      "learning_rate": 1.4171766337854083e-05,
+      "loss": 0.6579,
+      "step": 4995
+    },
+    {
+      "epoch": 0.7644988523335884,
+      "grad_norm": 1.9781887539285967,
+      "learning_rate": 1.416951397400623e-05,
+      "loss": 0.6649,
+      "step": 4996
+    },
+    {
+      "epoch": 0.7646518745218056,
+      "grad_norm": 2.2189360667232414,
+      "learning_rate": 1.4167261354092214e-05,
+      "loss": 0.7571,
+      "step": 4997
+    },
+    {
+      "epoch": 0.7648048967100229,
+      "grad_norm": 2.280749037561788,
+      "learning_rate": 1.4165008478250377e-05,
+      "loss": 0.717,
+      "step": 4998
+    },
+    {
+      "epoch": 0.7649579188982403,
+      "grad_norm": 2.362059249926859,
+      "learning_rate": 1.4162755346619075e-05,
+      "loss": 0.6807,
+      "step": 4999
+    },
+    {
+      "epoch": 0.7651109410864575,
+      "grad_norm": 2.2248228523984084,
+      "learning_rate": 1.4160501959336684e-05,
+      "loss": 0.6791,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7652639632746748,
+      "grad_norm": 2.2767134208389717,
+      "learning_rate": 1.415824831654159e-05,
+      "loss": 0.7279,
+      "step": 5001
+    },
+    {
+      "epoch": 0.7654169854628922,
+      "grad_norm": 2.259109443503308,
+      "learning_rate": 1.41559944183722e-05,
+      "loss": 0.6783,
+      "step": 5002
+    },
+    {
+      "epoch": 0.7655700076511094,
+      "grad_norm": 2.119358016956952,
+      "learning_rate": 1.4153740264966935e-05,
+      "loss": 0.7382,
+      "step": 5003
+    },
+    {
+      "epoch": 0.7657230298393267,
+      "grad_norm": 2.2868498026059507,
+      "learning_rate": 1.4151485856464231e-05,
+      "loss": 0.728,
+      "step": 5004
+    },
+    {
+      "epoch": 0.7658760520275439,
+      "grad_norm": 2.3259370090069074,
+      "learning_rate": 1.414923119300254e-05,
+      "loss": 0.7322,
+      "step": 5005
+    },
+    {
+      "epoch": 0.7660290742157613,
+      "grad_norm": 2.0952658010872325,
+      "learning_rate": 1.4146976274720325e-05,
+      "loss": 0.6611,
+      "step": 5006
+    },
+    {
+      "epoch": 0.7661820964039786,
+      "grad_norm": 2.2507490230036615,
+      "learning_rate": 1.414472110175608e-05,
+      "loss": 0.6851,
+      "step": 5007
+    },
+    {
+      "epoch": 0.7663351185921958,
+      "grad_norm": 2.3596475176429577,
+      "learning_rate": 1.4142465674248295e-05,
+      "loss": 0.7103,
+      "step": 5008
+    },
+    {
+      "epoch": 0.7664881407804132,
+      "grad_norm": 2.2428654694936165,
+      "learning_rate": 1.4140209992335488e-05,
+      "loss": 0.7419,
+      "step": 5009
+    },
+    {
+      "epoch": 0.7666411629686305,
+      "grad_norm": 2.412219562600667,
+      "learning_rate": 1.4137954056156189e-05,
+      "loss": 0.822,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7667941851568477,
+      "grad_norm": 2.469683847966789,
+      "learning_rate": 1.4135697865848945e-05,
+      "loss": 0.821,
+      "step": 5011
+    },
+    {
+      "epoch": 0.766947207345065,
+      "grad_norm": 2.380874074809561,
+      "learning_rate": 1.4133441421552312e-05,
+      "loss": 0.7444,
+      "step": 5012
+    },
+    {
+      "epoch": 0.7671002295332823,
+      "grad_norm": 2.54268340852517,
+      "learning_rate": 1.4131184723404876e-05,
+      "loss": 0.8583,
+      "step": 5013
+    },
+    {
+      "epoch": 0.7672532517214996,
+      "grad_norm": 2.1625190284078863,
+      "learning_rate": 1.4128927771545222e-05,
+      "loss": 0.6829,
+      "step": 5014
+    },
+    {
+      "epoch": 0.7674062739097169,
+      "grad_norm": 2.2366367904796176,
+      "learning_rate": 1.4126670566111962e-05,
+      "loss": 0.7066,
+      "step": 5015
+    },
+    {
+      "epoch": 0.7675592960979342,
+      "grad_norm": 2.2869932661750005,
+      "learning_rate": 1.4124413107243718e-05,
+      "loss": 0.7291,
+      "step": 5016
+    },
+    {
+      "epoch": 0.7677123182861515,
+      "grad_norm": 2.461913309036648,
+      "learning_rate": 1.4122155395079132e-05,
+      "loss": 0.7526,
+      "step": 5017
+    },
+    {
+      "epoch": 0.7678653404743688,
+      "grad_norm": 2.4062027749990755,
+      "learning_rate": 1.4119897429756855e-05,
+      "loss": 0.742,
+      "step": 5018
+    },
+    {
+      "epoch": 0.768018362662586,
+      "grad_norm": 2.1281105958816373,
+      "learning_rate": 1.4117639211415561e-05,
+      "loss": 0.6535,
+      "step": 5019
+    },
+    {
+      "epoch": 0.7681713848508034,
+      "grad_norm": 2.442149676668343,
+      "learning_rate": 1.4115380740193936e-05,
+      "loss": 0.6923,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7683244070390206,
+      "grad_norm": 2.3490210282430426,
+      "learning_rate": 1.4113122016230678e-05,
+      "loss": 0.7034,
+      "step": 5021
+    },
+    {
+      "epoch": 0.7684774292272379,
+      "grad_norm": 2.18584937837184,
+      "learning_rate": 1.4110863039664506e-05,
+      "loss": 0.654,
+      "step": 5022
+    },
+    {
+      "epoch": 0.7686304514154553,
+      "grad_norm": 2.132444677822337,
+      "learning_rate": 1.4108603810634157e-05,
+      "loss": 0.6469,
+      "step": 5023
+    },
+    {
+      "epoch": 0.7687834736036725,
+      "grad_norm": 2.2211029054390634,
+      "learning_rate": 1.4106344329278372e-05,
+      "loss": 0.7746,
+      "step": 5024
+    },
+    {
+      "epoch": 0.7689364957918898,
+      "grad_norm": 2.5642196111069344,
+      "learning_rate": 1.4104084595735916e-05,
+      "loss": 0.7419,
+      "step": 5025
+    },
+    {
+      "epoch": 0.7690895179801072,
+      "grad_norm": 2.2008072174788715,
+      "learning_rate": 1.410182461014557e-05,
+      "loss": 0.7026,
+      "step": 5026
+    },
+    {
+      "epoch": 0.7692425401683244,
+      "grad_norm": 2.1514935715003087,
+      "learning_rate": 1.4099564372646132e-05,
+      "loss": 0.7478,
+      "step": 5027
+    },
+    {
+      "epoch": 0.7693955623565417,
+      "grad_norm": 2.1472496992000836,
+      "learning_rate": 1.4097303883376405e-05,
+      "loss": 0.7736,
+      "step": 5028
+    },
+    {
+      "epoch": 0.7695485845447589,
+      "grad_norm": 2.1083158311876997,
+      "learning_rate": 1.409504314247522e-05,
+      "loss": 0.7575,
+      "step": 5029
+    },
+    {
+      "epoch": 0.7697016067329763,
+      "grad_norm": 2.3859948402117506,
+      "learning_rate": 1.4092782150081415e-05,
+      "loss": 0.8054,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7698546289211936,
+      "grad_norm": 2.2696789787921308,
+      "learning_rate": 1.4090520906333844e-05,
+      "loss": 0.6825,
+      "step": 5031
+    },
+    {
+      "epoch": 0.7700076511094108,
+      "grad_norm": 2.5043142859591954,
+      "learning_rate": 1.4088259411371388e-05,
+      "loss": 0.7305,
+      "step": 5032
+    },
+    {
+      "epoch": 0.7701606732976282,
+      "grad_norm": 2.8269162934283716,
+      "learning_rate": 1.4085997665332925e-05,
+      "loss": 0.6334,
+      "step": 5033
+    },
+    {
+      "epoch": 0.7703136954858455,
+      "grad_norm": 2.363998579657459,
+      "learning_rate": 1.4083735668357359e-05,
+      "loss": 0.717,
+      "step": 5034
+    },
+    {
+      "epoch": 0.7704667176740627,
+      "grad_norm": 2.308357524429582,
+      "learning_rate": 1.4081473420583612e-05,
+      "loss": 0.7569,
+      "step": 5035
+    },
+    {
+      "epoch": 0.77061973986228,
+      "grad_norm": 2.1405563499565625,
+      "learning_rate": 1.4079210922150615e-05,
+      "loss": 0.7978,
+      "step": 5036
+    },
+    {
+      "epoch": 0.7707727620504973,
+      "grad_norm": 2.2014674249133113,
+      "learning_rate": 1.4076948173197316e-05,
+      "loss": 0.7363,
+      "step": 5037
+    },
+    {
+      "epoch": 0.7709257842387146,
+      "grad_norm": 2.5632111129243795,
+      "learning_rate": 1.4074685173862684e-05,
+      "loss": 0.7573,
+      "step": 5038
+    },
+    {
+      "epoch": 0.7710788064269319,
+      "grad_norm": 2.440105505427757,
+      "learning_rate": 1.4072421924285693e-05,
+      "loss": 0.7406,
+      "step": 5039
+    },
+    {
+      "epoch": 0.7712318286151492,
+      "grad_norm": 2.328805243013259,
+      "learning_rate": 1.4070158424605338e-05,
+      "loss": 0.7432,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7713848508033665,
+      "grad_norm": 2.265426601466516,
+      "learning_rate": 1.4067894674960637e-05,
+      "loss": 0.6968,
+      "step": 5041
+    },
+    {
+      "epoch": 0.7715378729915838,
+      "grad_norm": 2.268614568748264,
+      "learning_rate": 1.4065630675490605e-05,
+      "loss": 0.6453,
+      "step": 5042
+    },
+    {
+      "epoch": 0.771690895179801,
+      "grad_norm": 2.2010842207373305,
+      "learning_rate": 1.4063366426334293e-05,
+      "loss": 0.7092,
+      "step": 5043
+    },
+    {
+      "epoch": 0.7718439173680184,
+      "grad_norm": 2.0805941230225424,
+      "learning_rate": 1.4061101927630749e-05,
+      "loss": 0.5924,
+      "step": 5044
+    },
+    {
+      "epoch": 0.7719969395562356,
+      "grad_norm": 2.1821898010478353,
+      "learning_rate": 1.405883717951905e-05,
+      "loss": 0.6982,
+      "step": 5045
+    },
+    {
+      "epoch": 0.7721499617444529,
+      "grad_norm": 2.093645143185147,
+      "learning_rate": 1.4056572182138281e-05,
+      "loss": 0.6132,
+      "step": 5046
+    },
+    {
+      "epoch": 0.7723029839326703,
+      "grad_norm": 2.253233064275212,
+      "learning_rate": 1.4054306935627544e-05,
+      "loss": 0.7959,
+      "step": 5047
+    },
+    {
+      "epoch": 0.7724560061208875,
+      "grad_norm": 2.2821754594763073,
+      "learning_rate": 1.405204144012596e-05,
+      "loss": 0.7634,
+      "step": 5048
+    },
+    {
+      "epoch": 0.7726090283091048,
+      "grad_norm": 2.5505280501375576,
+      "learning_rate": 1.404977569577266e-05,
+      "loss": 0.775,
+      "step": 5049
+    },
+    {
+      "epoch": 0.7727620504973222,
+      "grad_norm": 2.1972396897268847,
+      "learning_rate": 1.404750970270679e-05,
+      "loss": 0.6405,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7729150726855394,
+      "grad_norm": 2.0449200641167264,
+      "learning_rate": 1.4045243461067514e-05,
+      "loss": 0.6206,
+      "step": 5051
+    },
+    {
+      "epoch": 0.7730680948737567,
+      "grad_norm": 2.386168396663282,
+      "learning_rate": 1.4042976970994015e-05,
+      "loss": 0.6859,
+      "step": 5052
+    },
+    {
+      "epoch": 0.7732211170619739,
+      "grad_norm": 2.514122014683401,
+      "learning_rate": 1.4040710232625481e-05,
+      "loss": 0.7822,
+      "step": 5053
+    },
+    {
+      "epoch": 0.7733741392501913,
+      "grad_norm": 2.5342070484797476,
+      "learning_rate": 1.4038443246101125e-05,
+      "loss": 0.6788,
+      "step": 5054
+    },
+    {
+      "epoch": 0.7735271614384086,
+      "grad_norm": 2.345617758130968,
+      "learning_rate": 1.4036176011560172e-05,
+      "loss": 0.8571,
+      "step": 5055
+    },
+    {
+      "epoch": 0.7736801836266258,
+      "grad_norm": 2.18626517984038,
+      "learning_rate": 1.403390852914186e-05,
+      "loss": 0.5995,
+      "step": 5056
+    },
+    {
+      "epoch": 0.7738332058148432,
+      "grad_norm": 2.23239353461061,
+      "learning_rate": 1.4031640798985446e-05,
+      "loss": 0.7064,
+      "step": 5057
+    },
+    {
+      "epoch": 0.7739862280030605,
+      "grad_norm": 2.1588061241982417,
+      "learning_rate": 1.4029372821230196e-05,
+      "loss": 0.622,
+      "step": 5058
+    },
+    {
+      "epoch": 0.7741392501912777,
+      "grad_norm": 2.321759106687769,
+      "learning_rate": 1.40271045960154e-05,
+      "loss": 0.746,
+      "step": 5059
+    },
+    {
+      "epoch": 0.774292272379495,
+      "grad_norm": 2.3401183762900266,
+      "learning_rate": 1.4024836123480356e-05,
+      "loss": 0.7694,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7744452945677123,
+      "grad_norm": 2.353276037005396,
+      "learning_rate": 1.402256740376438e-05,
+      "loss": 0.7548,
+      "step": 5061
+    },
+    {
+      "epoch": 0.7745983167559296,
+      "grad_norm": 2.2155907355110984,
+      "learning_rate": 1.4020298437006803e-05,
+      "loss": 0.6635,
+      "step": 5062
+    },
+    {
+      "epoch": 0.7747513389441469,
+      "grad_norm": 2.254233811440101,
+      "learning_rate": 1.4018029223346972e-05,
+      "loss": 0.5666,
+      "step": 5063
+    },
+    {
+      "epoch": 0.7749043611323642,
+      "grad_norm": 2.2481526640750764,
+      "learning_rate": 1.4015759762924246e-05,
+      "loss": 0.6752,
+      "step": 5064
+    },
+    {
+      "epoch": 0.7750573833205815,
+      "grad_norm": 2.148794153992223,
+      "learning_rate": 1.4013490055878008e-05,
+      "loss": 0.7136,
+      "step": 5065
+    },
+    {
+      "epoch": 0.7752104055087988,
+      "grad_norm": 2.3975560671750022,
+      "learning_rate": 1.401122010234764e-05,
+      "loss": 0.7725,
+      "step": 5066
+    },
+    {
+      "epoch": 0.775363427697016,
+      "grad_norm": 2.2347633781944407,
+      "learning_rate": 1.4008949902472554e-05,
+      "loss": 0.637,
+      "step": 5067
+    },
+    {
+      "epoch": 0.7755164498852334,
+      "grad_norm": 2.285143070661817,
+      "learning_rate": 1.4006679456392174e-05,
+      "loss": 0.7154,
+      "step": 5068
+    },
+    {
+      "epoch": 0.7756694720734506,
+      "grad_norm": 2.2031275325618553,
+      "learning_rate": 1.4004408764245934e-05,
+      "loss": 0.7259,
+      "step": 5069
+    },
+    {
+      "epoch": 0.7758224942616679,
+      "grad_norm": 2.315668163634487,
+      "learning_rate": 1.4002137826173286e-05,
+      "loss": 0.7905,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7759755164498853,
+      "grad_norm": 2.541307986854028,
+      "learning_rate": 1.3999866642313698e-05,
+      "loss": 0.6911,
+      "step": 5071
+    },
+    {
+      "epoch": 0.7761285386381025,
+      "grad_norm": 2.27496047507574,
+      "learning_rate": 1.3997595212806648e-05,
+      "loss": 0.6393,
+      "step": 5072
+    },
+    {
+      "epoch": 0.7762815608263198,
+      "grad_norm": 2.247769499454639,
+      "learning_rate": 1.3995323537791643e-05,
+      "loss": 0.76,
+      "step": 5073
+    },
+    {
+      "epoch": 0.7764345830145372,
+      "grad_norm": 2.2888853081384974,
+      "learning_rate": 1.3993051617408186e-05,
+      "loss": 0.7552,
+      "step": 5074
+    },
+    {
+      "epoch": 0.7765876052027544,
+      "grad_norm": 2.2001010956721037,
+      "learning_rate": 1.3990779451795808e-05,
+      "loss": 0.6797,
+      "step": 5075
+    },
+    {
+      "epoch": 0.7767406273909717,
+      "grad_norm": 2.4115578977098164,
+      "learning_rate": 1.3988507041094055e-05,
+      "loss": 0.7154,
+      "step": 5076
+    },
+    {
+      "epoch": 0.7768936495791889,
+      "grad_norm": 2.1463792083802273,
+      "learning_rate": 1.3986234385442481e-05,
+      "loss": 0.6909,
+      "step": 5077
+    },
+    {
+      "epoch": 0.7770466717674063,
+      "grad_norm": 2.0897406819197863,
+      "learning_rate": 1.3983961484980656e-05,
+      "loss": 0.751,
+      "step": 5078
+    },
+    {
+      "epoch": 0.7771996939556236,
+      "grad_norm": 2.1608247723805687,
+      "learning_rate": 1.3981688339848174e-05,
+      "loss": 0.6893,
+      "step": 5079
+    },
+    {
+      "epoch": 0.7773527161438408,
+      "grad_norm": 2.882363155314608,
+      "learning_rate": 1.3979414950184632e-05,
+      "loss": 0.6289,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7775057383320582,
+      "grad_norm": 2.5337887564427883,
+      "learning_rate": 1.3977141316129653e-05,
+      "loss": 0.7907,
+      "step": 5081
+    },
+    {
+      "epoch": 0.7776587605202755,
+      "grad_norm": 2.368738369887408,
+      "learning_rate": 1.3974867437822866e-05,
+      "loss": 0.777,
+      "step": 5082
+    },
+    {
+      "epoch": 0.7778117827084927,
+      "grad_norm": 2.3466429855351505,
+      "learning_rate": 1.3972593315403919e-05,
+      "loss": 0.748,
+      "step": 5083
+    },
+    {
+      "epoch": 0.77796480489671,
+      "grad_norm": 2.269874457550218,
+      "learning_rate": 1.3970318949012475e-05,
+      "loss": 0.7797,
+      "step": 5084
+    },
+    {
+      "epoch": 0.7781178270849273,
+      "grad_norm": 2.284119463233079,
+      "learning_rate": 1.3968044338788216e-05,
+      "loss": 0.7871,
+      "step": 5085
+    },
+    {
+      "epoch": 0.7782708492731446,
+      "grad_norm": 2.2696960329330813,
+      "learning_rate": 1.3965769484870829e-05,
+      "loss": 0.6889,
+      "step": 5086
+    },
+    {
+      "epoch": 0.7784238714613619,
+      "grad_norm": 2.420011775174571,
+      "learning_rate": 1.3963494387400023e-05,
+      "loss": 0.6882,
+      "step": 5087
+    },
+    {
+      "epoch": 0.7785768936495792,
+      "grad_norm": 2.1957556052360916,
+      "learning_rate": 1.3961219046515519e-05,
+      "loss": 0.7425,
+      "step": 5088
+    },
+    {
+      "epoch": 0.7787299158377965,
+      "grad_norm": 2.119582758763422,
+      "learning_rate": 1.3958943462357065e-05,
+      "loss": 0.6109,
+      "step": 5089
+    },
+    {
+      "epoch": 0.7788829380260138,
+      "grad_norm": 2.6124106631663984,
+      "learning_rate": 1.3956667635064398e-05,
+      "loss": 0.7887,
+      "step": 5090
+    },
+    {
+      "epoch": 0.779035960214231,
+      "grad_norm": 2.2893517601019107,
+      "learning_rate": 1.3954391564777295e-05,
+      "loss": 0.6306,
+      "step": 5091
+    },
+    {
+      "epoch": 0.7791889824024484,
+      "grad_norm": 2.1777151683208444,
+      "learning_rate": 1.3952115251635536e-05,
+      "loss": 0.7121,
+      "step": 5092
+    },
+    {
+      "epoch": 0.7793420045906656,
+      "grad_norm": 2.2899237702858746,
+      "learning_rate": 1.3949838695778921e-05,
+      "loss": 0.7505,
+      "step": 5093
+    },
+    {
+      "epoch": 0.7794950267788829,
+      "grad_norm": 2.42586559436103,
+      "learning_rate": 1.3947561897347257e-05,
+      "loss": 0.7946,
+      "step": 5094
+    },
+    {
+      "epoch": 0.7796480489671003,
+      "grad_norm": 1.9766272067621482,
+      "learning_rate": 1.3945284856480376e-05,
+      "loss": 0.6901,
+      "step": 5095
+    },
+    {
+      "epoch": 0.7798010711553175,
+      "grad_norm": 2.156148901403677,
+      "learning_rate": 1.3943007573318117e-05,
+      "loss": 0.7859,
+      "step": 5096
+    },
+    {
+      "epoch": 0.7799540933435348,
+      "grad_norm": 2.5202800450554865,
+      "learning_rate": 1.3940730048000338e-05,
+      "loss": 0.7896,
+      "step": 5097
+    },
+    {
+      "epoch": 0.780107115531752,
+      "grad_norm": 2.4171048421326207,
+      "learning_rate": 1.3938452280666909e-05,
+      "loss": 0.7879,
+      "step": 5098
+    },
+    {
+      "epoch": 0.7802601377199694,
+      "grad_norm": 2.4109847004996174,
+      "learning_rate": 1.3936174271457721e-05,
+      "loss": 0.8669,
+      "step": 5099
+    },
+    {
+      "epoch": 0.7804131599081867,
+      "grad_norm": 2.0993258641086228,
+      "learning_rate": 1.3933896020512671e-05,
+      "loss": 0.727,
+      "step": 5100
+    },
+    {
+      "epoch": 0.7805661820964039,
+      "grad_norm": 2.027143937230042,
+      "learning_rate": 1.3931617527971674e-05,
+      "loss": 0.6448,
+      "step": 5101
+    },
+    {
+      "epoch": 0.7807192042846213,
+      "grad_norm": 1.984107315848014,
+      "learning_rate": 1.3929338793974671e-05,
+      "loss": 0.6419,
+      "step": 5102
+    },
+    {
+      "epoch": 0.7808722264728386,
+      "grad_norm": 1.9833282224258104,
+      "learning_rate": 1.3927059818661596e-05,
+      "loss": 0.5596,
+      "step": 5103
+    },
+    {
+      "epoch": 0.7810252486610558,
+      "grad_norm": 2.1670221414876756,
+      "learning_rate": 1.3924780602172413e-05,
+      "loss": 0.717,
+      "step": 5104
+    },
+    {
+      "epoch": 0.7811782708492732,
+      "grad_norm": 2.2583166303336237,
+      "learning_rate": 1.3922501144647105e-05,
+      "loss": 0.7974,
+      "step": 5105
+    },
+    {
+      "epoch": 0.7813312930374904,
+      "grad_norm": 2.380217191342555,
+      "learning_rate": 1.3920221446225654e-05,
+      "loss": 0.7113,
+      "step": 5106
+    },
+    {
+      "epoch": 0.7814843152257077,
+      "grad_norm": 2.220484377761435,
+      "learning_rate": 1.3917941507048068e-05,
+      "loss": 0.7526,
+      "step": 5107
+    },
+    {
+      "epoch": 0.781637337413925,
+      "grad_norm": 2.298676239650065,
+      "learning_rate": 1.3915661327254367e-05,
+      "loss": 0.7043,
+      "step": 5108
+    },
+    {
+      "epoch": 0.7817903596021423,
+      "grad_norm": 2.1637013431269447,
+      "learning_rate": 1.3913380906984586e-05,
+      "loss": 0.7548,
+      "step": 5109
+    },
+    {
+      "epoch": 0.7819433817903596,
+      "grad_norm": 2.1322701567201006,
+      "learning_rate": 1.3911100246378775e-05,
+      "loss": 0.758,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7820964039785769,
+      "grad_norm": 2.192596038504427,
+      "learning_rate": 1.3908819345576996e-05,
+      "loss": 0.6943,
+      "step": 5111
+    },
+    {
+      "epoch": 0.7822494261667942,
+      "grad_norm": 2.2990794845292317,
+      "learning_rate": 1.3906538204719329e-05,
+      "loss": 0.6867,
+      "step": 5112
+    },
+    {
+      "epoch": 0.7824024483550115,
+      "grad_norm": 2.3025056442381735,
+      "learning_rate": 1.3904256823945868e-05,
+      "loss": 0.7531,
+      "step": 5113
+    },
+    {
+      "epoch": 0.7825554705432287,
+      "grad_norm": 2.318696680589184,
+      "learning_rate": 1.3901975203396724e-05,
+      "loss": 0.6702,
+      "step": 5114
+    },
+    {
+      "epoch": 0.782708492731446,
+      "grad_norm": 2.0791679798970897,
+      "learning_rate": 1.389969334321202e-05,
+      "loss": 0.7031,
+      "step": 5115
+    },
+    {
+      "epoch": 0.7828615149196634,
+      "grad_norm": 2.2745500029402694,
+      "learning_rate": 1.3897411243531886e-05,
+      "loss": 0.6795,
+      "step": 5116
+    },
+    {
+      "epoch": 0.7830145371078806,
+      "grad_norm": 2.4518559288103554,
+      "learning_rate": 1.3895128904496486e-05,
+      "loss": 0.7113,
+      "step": 5117
+    },
+    {
+      "epoch": 0.7831675592960979,
+      "grad_norm": 2.2388389113594553,
+      "learning_rate": 1.3892846326245984e-05,
+      "loss": 0.7826,
+      "step": 5118
+    },
+    {
+      "epoch": 0.7833205814843153,
+      "grad_norm": 2.0864895233637775,
+      "learning_rate": 1.3890563508920554e-05,
+      "loss": 0.7918,
+      "step": 5119
+    },
+    {
+      "epoch": 0.7834736036725325,
+      "grad_norm": 2.250652586337949,
+      "learning_rate": 1.3888280452660401e-05,
+      "loss": 0.8105,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7836266258607498,
+      "grad_norm": 2.5348100337657438,
+      "learning_rate": 1.3885997157605737e-05,
+      "loss": 0.7097,
+      "step": 5121
+    },
+    {
+      "epoch": 0.783779648048967,
+      "grad_norm": 2.2521060385487277,
+      "learning_rate": 1.3883713623896782e-05,
+      "loss": 0.8175,
+      "step": 5122
+    },
+    {
+      "epoch": 0.7839326702371844,
+      "grad_norm": 2.4476407046502446,
+      "learning_rate": 1.3881429851673781e-05,
+      "loss": 0.7518,
+      "step": 5123
+    },
+    {
+      "epoch": 0.7840856924254017,
+      "grad_norm": 2.512045611078853,
+      "learning_rate": 1.3879145841076991e-05,
+      "loss": 0.7752,
+      "step": 5124
+    },
+    {
+      "epoch": 0.7842387146136189,
+      "grad_norm": 2.278880822195872,
+      "learning_rate": 1.3876861592246678e-05,
+      "loss": 0.7291,
+      "step": 5125
+    },
+    {
+      "epoch": 0.7843917368018363,
+      "grad_norm": 1.9592367955208152,
+      "learning_rate": 1.3874577105323127e-05,
+      "loss": 0.5866,
+      "step": 5126
+    },
+    {
+      "epoch": 0.7845447589900536,
+      "grad_norm": 2.1204944930182466,
+      "learning_rate": 1.3872292380446641e-05,
+      "loss": 0.6941,
+      "step": 5127
+    },
+    {
+      "epoch": 0.7846977811782708,
+      "grad_norm": 2.436918071377942,
+      "learning_rate": 1.3870007417757529e-05,
+      "loss": 0.7549,
+      "step": 5128
+    },
+    {
+      "epoch": 0.7848508033664882,
+      "grad_norm": 2.049946768306998,
+      "learning_rate": 1.3867722217396122e-05,
+      "loss": 0.7674,
+      "step": 5129
+    },
+    {
+      "epoch": 0.7850038255547054,
+      "grad_norm": 2.5063329923729554,
+      "learning_rate": 1.3865436779502767e-05,
+      "loss": 0.7467,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7851568477429227,
+      "grad_norm": 1.9886281059847166,
+      "learning_rate": 1.3863151104217816e-05,
+      "loss": 0.6673,
+      "step": 5131
+    },
+    {
+      "epoch": 0.78530986993114,
+      "grad_norm": 2.1184066490210727,
+      "learning_rate": 1.3860865191681639e-05,
+      "loss": 0.6644,
+      "step": 5132
+    },
+    {
+      "epoch": 0.7854628921193573,
+      "grad_norm": 2.2524087014750336,
+      "learning_rate": 1.385857904203463e-05,
+      "loss": 0.8056,
+      "step": 5133
+    },
+    {
+      "epoch": 0.7856159143075746,
+      "grad_norm": 2.1941572669081326,
+      "learning_rate": 1.3856292655417187e-05,
+      "loss": 0.7362,
+      "step": 5134
+    },
+    {
+      "epoch": 0.7857689364957919,
+      "grad_norm": 2.1730871002177836,
+      "learning_rate": 1.3854006031969727e-05,
+      "loss": 0.6781,
+      "step": 5135
+    },
+    {
+      "epoch": 0.7859219586840092,
+      "grad_norm": 2.4598351761040256,
+      "learning_rate": 1.3851719171832678e-05,
+      "loss": 0.8042,
+      "step": 5136
+    },
+    {
+      "epoch": 0.7860749808722265,
+      "grad_norm": 2.38153534428331,
+      "learning_rate": 1.3849432075146485e-05,
+      "loss": 0.7745,
+      "step": 5137
+    },
+    {
+      "epoch": 0.7862280030604437,
+      "grad_norm": 2.1360826368406354,
+      "learning_rate": 1.3847144742051613e-05,
+      "loss": 0.8381,
+      "step": 5138
+    },
+    {
+      "epoch": 0.786381025248661,
+      "grad_norm": 2.85778086002553,
+      "learning_rate": 1.3844857172688531e-05,
+      "loss": 0.8439,
+      "step": 5139
+    },
+    {
+      "epoch": 0.7865340474368784,
+      "grad_norm": 2.2874345985423896,
+      "learning_rate": 1.3842569367197726e-05,
+      "loss": 0.7498,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7866870696250956,
+      "grad_norm": 2.132202408900261,
+      "learning_rate": 1.3840281325719708e-05,
+      "loss": 0.6593,
+      "step": 5141
+    },
+    {
+      "epoch": 0.7868400918133129,
+      "grad_norm": 2.224486844047528,
+      "learning_rate": 1.3837993048394988e-05,
+      "loss": 0.7379,
+      "step": 5142
+    },
+    {
+      "epoch": 0.7869931140015303,
+      "grad_norm": 2.2894622734424734,
+      "learning_rate": 1.3835704535364103e-05,
+      "loss": 0.7231,
+      "step": 5143
+    },
+    {
+      "epoch": 0.7871461361897475,
+      "grad_norm": 2.100328372449939,
+      "learning_rate": 1.3833415786767596e-05,
+      "loss": 0.6338,
+      "step": 5144
+    },
+    {
+      "epoch": 0.7872991583779648,
+      "grad_norm": 2.4355056805050217,
+      "learning_rate": 1.3831126802746026e-05,
+      "loss": 0.6756,
+      "step": 5145
+    },
+    {
+      "epoch": 0.787452180566182,
+      "grad_norm": 2.2133365219845618,
+      "learning_rate": 1.3828837583439975e-05,
+      "loss": 0.7254,
+      "step": 5146
+    },
+    {
+      "epoch": 0.7876052027543994,
+      "grad_norm": 2.286610066393195,
+      "learning_rate": 1.3826548128990031e-05,
+      "loss": 0.7471,
+      "step": 5147
+    },
+    {
+      "epoch": 0.7877582249426167,
+      "grad_norm": 2.545738125909192,
+      "learning_rate": 1.3824258439536793e-05,
+      "loss": 0.7792,
+      "step": 5148
+    },
+    {
+      "epoch": 0.7879112471308339,
+      "grad_norm": 2.1321283437860665,
+      "learning_rate": 1.3821968515220885e-05,
+      "loss": 0.6472,
+      "step": 5149
+    },
+    {
+      "epoch": 0.7880642693190513,
+      "grad_norm": 2.7061865458512995,
+      "learning_rate": 1.381967835618294e-05,
+      "loss": 0.791,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7882172915072686,
+      "grad_norm": 2.5000426098798427,
+      "learning_rate": 1.3817387962563605e-05,
+      "loss": 0.7419,
+      "step": 5151
+    },
+    {
+      "epoch": 0.7883703136954858,
+      "grad_norm": 2.1413176672534,
+      "learning_rate": 1.381509733450354e-05,
+      "loss": 0.6886,
+      "step": 5152
+    },
+    {
+      "epoch": 0.7885233358837032,
+      "grad_norm": 2.0937281811946438,
+      "learning_rate": 1.3812806472143423e-05,
+      "loss": 0.6495,
+      "step": 5153
+    },
+    {
+      "epoch": 0.7886763580719204,
+      "grad_norm": 2.350309442778104,
+      "learning_rate": 1.3810515375623944e-05,
+      "loss": 0.6501,
+      "step": 5154
+    },
+    {
+      "epoch": 0.7888293802601377,
+      "grad_norm": 2.152716288815514,
+      "learning_rate": 1.3808224045085812e-05,
+      "loss": 0.7354,
+      "step": 5155
+    },
+    {
+      "epoch": 0.788982402448355,
+      "grad_norm": 2.15154503852824,
+      "learning_rate": 1.3805932480669739e-05,
+      "loss": 0.7494,
+      "step": 5156
+    },
+    {
+      "epoch": 0.7891354246365723,
+      "grad_norm": 2.0973581972487745,
+      "learning_rate": 1.3803640682516466e-05,
+      "loss": 0.7602,
+      "step": 5157
+    },
+    {
+      "epoch": 0.7892884468247896,
+      "grad_norm": 2.711881599173585,
+      "learning_rate": 1.3801348650766739e-05,
+      "loss": 0.6601,
+      "step": 5158
+    },
+    {
+      "epoch": 0.7894414690130069,
+      "grad_norm": 2.184034530375545,
+      "learning_rate": 1.379905638556132e-05,
+      "loss": 0.6898,
+      "step": 5159
+    },
+    {
+      "epoch": 0.7895944912012242,
+      "grad_norm": 2.095943948257656,
+      "learning_rate": 1.3796763887040987e-05,
+      "loss": 0.7355,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7897475133894415,
+      "grad_norm": 2.2534986734403595,
+      "learning_rate": 1.3794471155346529e-05,
+      "loss": 0.7118,
+      "step": 5161
+    },
+    {
+      "epoch": 0.7899005355776587,
+      "grad_norm": 2.1206432371445545,
+      "learning_rate": 1.3792178190618754e-05,
+      "loss": 0.6405,
+      "step": 5162
+    },
+    {
+      "epoch": 0.790053557765876,
+      "grad_norm": 1.9695127305630176,
+      "learning_rate": 1.3789884992998484e-05,
+      "loss": 0.6778,
+      "step": 5163
+    },
+    {
+      "epoch": 0.7902065799540934,
+      "grad_norm": 2.3350051931180227,
+      "learning_rate": 1.3787591562626545e-05,
+      "loss": 0.8381,
+      "step": 5164
+    },
+    {
+      "epoch": 0.7903596021423106,
+      "grad_norm": 2.52231424297307,
+      "learning_rate": 1.3785297899643797e-05,
+      "loss": 0.8362,
+      "step": 5165
+    },
+    {
+      "epoch": 0.7905126243305279,
+      "grad_norm": 2.1652917778433753,
+      "learning_rate": 1.3783004004191095e-05,
+      "loss": 0.825,
+      "step": 5166
+    },
+    {
+      "epoch": 0.7906656465187453,
+      "grad_norm": 2.156387877386113,
+      "learning_rate": 1.3780709876409315e-05,
+      "loss": 0.5602,
+      "step": 5167
+    },
+    {
+      "epoch": 0.7908186687069625,
+      "grad_norm": 2.42839264979406,
+      "learning_rate": 1.3778415516439352e-05,
+      "loss": 0.7292,
+      "step": 5168
+    },
+    {
+      "epoch": 0.7909716908951798,
+      "grad_norm": 2.4513243449235134,
+      "learning_rate": 1.3776120924422114e-05,
+      "loss": 0.8545,
+      "step": 5169
+    },
+    {
+      "epoch": 0.791124713083397,
+      "grad_norm": 2.2296488916634756,
+      "learning_rate": 1.3773826100498512e-05,
+      "loss": 0.641,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7912777352716144,
+      "grad_norm": 2.2384319518642677,
+      "learning_rate": 1.377153104480949e-05,
+      "loss": 0.7482,
+      "step": 5171
+    },
+    {
+      "epoch": 0.7914307574598317,
+      "grad_norm": 2.2378257415185385,
+      "learning_rate": 1.3769235757495994e-05,
+      "loss": 0.7268,
+      "step": 5172
+    },
+    {
+      "epoch": 0.7915837796480489,
+      "grad_norm": 2.1791575768084375,
+      "learning_rate": 1.3766940238698983e-05,
+      "loss": 0.6738,
+      "step": 5173
+    },
+    {
+      "epoch": 0.7917368018362663,
+      "grad_norm": 2.214967735911992,
+      "learning_rate": 1.3764644488559433e-05,
+      "loss": 0.6761,
+      "step": 5174
+    },
+    {
+      "epoch": 0.7918898240244836,
+      "grad_norm": 2.0880221650456083,
+      "learning_rate": 1.3762348507218342e-05,
+      "loss": 0.701,
+      "step": 5175
+    },
+    {
+      "epoch": 0.7920428462127008,
+      "grad_norm": 1.9249158885011985,
+      "learning_rate": 1.3760052294816708e-05,
+      "loss": 0.6566,
+      "step": 5176
+    },
+    {
+      "epoch": 0.7921958684009182,
+      "grad_norm": 2.304539470343536,
+      "learning_rate": 1.3757755851495553e-05,
+      "loss": 0.7316,
+      "step": 5177
+    },
+    {
+      "epoch": 0.7923488905891354,
+      "grad_norm": 2.040611815803622,
+      "learning_rate": 1.3755459177395911e-05,
+      "loss": 0.7166,
+      "step": 5178
+    },
+    {
+      "epoch": 0.7925019127773527,
+      "grad_norm": 2.3706391834263942,
+      "learning_rate": 1.3753162272658832e-05,
+      "loss": 0.6723,
+      "step": 5179
+    },
+    {
+      "epoch": 0.79265493496557,
+      "grad_norm": 2.1589457526111877,
+      "learning_rate": 1.3750865137425371e-05,
+      "loss": 0.7957,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7928079571537873,
+      "grad_norm": 2.357150090202171,
+      "learning_rate": 1.3748567771836612e-05,
+      "loss": 0.7333,
+      "step": 5181
+    },
+    {
+      "epoch": 0.7929609793420046,
+      "grad_norm": 2.035270416220051,
+      "learning_rate": 1.374627017603364e-05,
+      "loss": 0.6303,
+      "step": 5182
+    },
+    {
+      "epoch": 0.7931140015302219,
+      "grad_norm": 2.310334406515395,
+      "learning_rate": 1.374397235015756e-05,
+      "loss": 0.6799,
+      "step": 5183
+    },
+    {
+      "epoch": 0.7932670237184392,
+      "grad_norm": 2.5329925277281378,
+      "learning_rate": 1.3741674294349494e-05,
+      "loss": 0.7163,
+      "step": 5184
+    },
+    {
+      "epoch": 0.7934200459066565,
+      "grad_norm": 2.3209415935722078,
+      "learning_rate": 1.373937600875057e-05,
+      "loss": 0.6513,
+      "step": 5185
+    },
+    {
+      "epoch": 0.7935730680948737,
+      "grad_norm": 2.2392121898655044,
+      "learning_rate": 1.3737077493501939e-05,
+      "loss": 0.6232,
+      "step": 5186
+    },
+    {
+      "epoch": 0.793726090283091,
+      "grad_norm": 2.053324943153721,
+      "learning_rate": 1.373477874874476e-05,
+      "loss": 0.6207,
+      "step": 5187
+    },
+    {
+      "epoch": 0.7938791124713084,
+      "grad_norm": 2.3153062662700026,
+      "learning_rate": 1.3732479774620206e-05,
+      "loss": 0.7312,
+      "step": 5188
+    },
+    {
+      "epoch": 0.7940321346595256,
+      "grad_norm": 2.202811664872602,
+      "learning_rate": 1.3730180571269465e-05,
+      "loss": 0.736,
+      "step": 5189
+    },
+    {
+      "epoch": 0.7941851568477429,
+      "grad_norm": 2.095998770625852,
+      "learning_rate": 1.3727881138833746e-05,
+      "loss": 0.6731,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7943381790359603,
+      "grad_norm": 1.841345240252431,
+      "learning_rate": 1.3725581477454262e-05,
+      "loss": 0.5946,
+      "step": 5191
+    },
+    {
+      "epoch": 0.7944912012241775,
+      "grad_norm": 2.390734382027951,
+      "learning_rate": 1.3723281587272243e-05,
+      "loss": 0.6396,
+      "step": 5192
+    },
+    {
+      "epoch": 0.7946442234123948,
+      "grad_norm": 2.2841584319362758,
+      "learning_rate": 1.3720981468428938e-05,
+      "loss": 0.7108,
+      "step": 5193
+    },
+    {
+      "epoch": 0.794797245600612,
+      "grad_norm": 2.3818717367728075,
+      "learning_rate": 1.3718681121065605e-05,
+      "loss": 0.7718,
+      "step": 5194
+    },
+    {
+      "epoch": 0.7949502677888294,
+      "grad_norm": 2.5859476893129165,
+      "learning_rate": 1.3716380545323516e-05,
+      "loss": 0.7855,
+      "step": 5195
+    },
+    {
+      "epoch": 0.7951032899770467,
+      "grad_norm": 2.3940293661363654,
+      "learning_rate": 1.371407974134396e-05,
+      "loss": 0.7146,
+      "step": 5196
+    },
+    {
+      "epoch": 0.7952563121652639,
+      "grad_norm": 2.2903896079503316,
+      "learning_rate": 1.3711778709268235e-05,
+      "loss": 0.7184,
+      "step": 5197
+    },
+    {
+      "epoch": 0.7954093343534813,
+      "grad_norm": 2.01581293118149,
+      "learning_rate": 1.3709477449237661e-05,
+      "loss": 0.6431,
+      "step": 5198
+    },
+    {
+      "epoch": 0.7955623565416986,
+      "grad_norm": 1.8836934333964197,
+      "learning_rate": 1.3707175961393564e-05,
+      "loss": 0.616,
+      "step": 5199
+    },
+    {
+      "epoch": 0.7957153787299158,
+      "grad_norm": 2.2744019198062815,
+      "learning_rate": 1.370487424587729e-05,
+      "loss": 0.6745,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7958684009181332,
+      "grad_norm": 2.117675246910853,
+      "learning_rate": 1.3702572302830194e-05,
+      "loss": 0.6382,
+      "step": 5201
+    },
+    {
+      "epoch": 0.7960214231063504,
+      "grad_norm": 2.125438352072235,
+      "learning_rate": 1.370027013239365e-05,
+      "loss": 0.6448,
+      "step": 5202
+    },
+    {
+      "epoch": 0.7961744452945677,
+      "grad_norm": 2.174518702017577,
+      "learning_rate": 1.369796773470904e-05,
+      "loss": 0.5696,
+      "step": 5203
+    },
+    {
+      "epoch": 0.796327467482785,
+      "grad_norm": 2.6649237891451216,
+      "learning_rate": 1.3695665109917764e-05,
+      "loss": 0.7357,
+      "step": 5204
+    },
+    {
+      "epoch": 0.7964804896710023,
+      "grad_norm": 2.130917757031721,
+      "learning_rate": 1.3693362258161239e-05,
+      "loss": 0.7103,
+      "step": 5205
+    },
+    {
+      "epoch": 0.7966335118592196,
+      "grad_norm": 2.308278946221109,
+      "learning_rate": 1.3691059179580888e-05,
+      "loss": 0.7996,
+      "step": 5206
+    },
+    {
+      "epoch": 0.7967865340474368,
+      "grad_norm": 2.2292739591762567,
+      "learning_rate": 1.3688755874318154e-05,
+      "loss": 0.7414,
+      "step": 5207
+    },
+    {
+      "epoch": 0.7969395562356542,
+      "grad_norm": 2.3250780563176474,
+      "learning_rate": 1.3686452342514486e-05,
+      "loss": 0.6129,
+      "step": 5208
+    },
+    {
+      "epoch": 0.7970925784238715,
+      "grad_norm": 2.0736440654210644,
+      "learning_rate": 1.3684148584311365e-05,
+      "loss": 0.6746,
+      "step": 5209
+    },
+    {
+      "epoch": 0.7972456006120887,
+      "grad_norm": 2.3603589942528678,
+      "learning_rate": 1.3681844599850265e-05,
+      "loss": 0.7034,
+      "step": 5210
+    },
+    {
+      "epoch": 0.797398622800306,
+      "grad_norm": 2.349702397627265,
+      "learning_rate": 1.3679540389272683e-05,
+      "loss": 0.6011,
+      "step": 5211
+    },
+    {
+      "epoch": 0.7975516449885234,
+      "grad_norm": 2.366788617289836,
+      "learning_rate": 1.3677235952720132e-05,
+      "loss": 0.6788,
+      "step": 5212
+    },
+    {
+      "epoch": 0.7977046671767406,
+      "grad_norm": 2.0913280489885233,
+      "learning_rate": 1.3674931290334137e-05,
+      "loss": 0.6981,
+      "step": 5213
+    },
+    {
+      "epoch": 0.7978576893649579,
+      "grad_norm": 2.5304173735852045,
+      "learning_rate": 1.3672626402256233e-05,
+      "loss": 0.7291,
+      "step": 5214
+    },
+    {
+      "epoch": 0.7980107115531752,
+      "grad_norm": 2.3922657289236673,
+      "learning_rate": 1.3670321288627975e-05,
+      "loss": 0.7516,
+      "step": 5215
+    },
+    {
+      "epoch": 0.7981637337413925,
+      "grad_norm": 2.198486199079782,
+      "learning_rate": 1.3668015949590929e-05,
+      "loss": 0.6851,
+      "step": 5216
+    },
+    {
+      "epoch": 0.7983167559296098,
+      "grad_norm": 2.3466456024492053,
+      "learning_rate": 1.3665710385286674e-05,
+      "loss": 0.6804,
+      "step": 5217
+    },
+    {
+      "epoch": 0.798469778117827,
+      "grad_norm": 2.166891644272971,
+      "learning_rate": 1.3663404595856804e-05,
+      "loss": 0.7362,
+      "step": 5218
+    },
+    {
+      "epoch": 0.7986228003060444,
+      "grad_norm": 2.3442111181100946,
+      "learning_rate": 1.3661098581442924e-05,
+      "loss": 0.7576,
+      "step": 5219
+    },
+    {
+      "epoch": 0.7987758224942617,
+      "grad_norm": 2.2049727285082357,
+      "learning_rate": 1.3658792342186662e-05,
+      "loss": 0.632,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7989288446824789,
+      "grad_norm": 2.4747037218029666,
+      "learning_rate": 1.3656485878229646e-05,
+      "loss": 0.8662,
+      "step": 5221
+    },
+    {
+      "epoch": 0.7990818668706963,
+      "grad_norm": 2.154683342466119,
+      "learning_rate": 1.365417918971353e-05,
+      "loss": 0.7382,
+      "step": 5222
+    },
+    {
+      "epoch": 0.7992348890589135,
+      "grad_norm": 2.3297008480161994,
+      "learning_rate": 1.3651872276779975e-05,
+      "loss": 0.7879,
+      "step": 5223
+    },
+    {
+      "epoch": 0.7993879112471308,
+      "grad_norm": 2.678458704148208,
+      "learning_rate": 1.3649565139570653e-05,
+      "loss": 0.8128,
+      "step": 5224
+    },
+    {
+      "epoch": 0.7995409334353482,
+      "grad_norm": 2.4406095933985035,
+      "learning_rate": 1.3647257778227263e-05,
+      "loss": 0.6893,
+      "step": 5225
+    },
+    {
+      "epoch": 0.7996939556235654,
+      "grad_norm": 2.533590774082434,
+      "learning_rate": 1.3644950192891502e-05,
+      "loss": 0.8111,
+      "step": 5226
+    },
+    {
+      "epoch": 0.7998469778117827,
+      "grad_norm": 2.1737368302955056,
+      "learning_rate": 1.3642642383705092e-05,
+      "loss": 0.7153,
+      "step": 5227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.033682378494723,
+      "learning_rate": 1.3640334350809763e-05,
+      "loss": 0.7538,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8001530221882173,
+      "grad_norm": 2.2980150041237635,
+      "learning_rate": 1.3638026094347261e-05,
+      "loss": 0.7585,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8003060443764346,
+      "grad_norm": 2.139295198153623,
+      "learning_rate": 1.3635717614459342e-05,
+      "loss": 0.6941,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8004590665646518,
+      "grad_norm": 2.331590755207613,
+      "learning_rate": 1.3633408911287785e-05,
+      "loss": 0.7219,
+      "step": 5231
+    },
+    {
+      "epoch": 0.8006120887528692,
+      "grad_norm": 2.132296410347176,
+      "learning_rate": 1.363109998497437e-05,
+      "loss": 0.742,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8007651109410865,
+      "grad_norm": 2.2753739670012743,
+      "learning_rate": 1.3628790835660901e-05,
+      "loss": 0.6547,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8009181331293037,
+      "grad_norm": 2.213855353645808,
+      "learning_rate": 1.3626481463489189e-05,
+      "loss": 0.6624,
+      "step": 5234
+    },
+    {
+      "epoch": 0.801071155317521,
+      "grad_norm": 2.3897025233653797,
+      "learning_rate": 1.3624171868601066e-05,
+      "loss": 0.7068,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8012241775057384,
+      "grad_norm": 2.161798259768198,
+      "learning_rate": 1.3621862051138368e-05,
+      "loss": 0.6786,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8013771996939556,
+      "grad_norm": 2.063352788797973,
+      "learning_rate": 1.3619552011242956e-05,
+      "loss": 0.6605,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8015302218821729,
+      "grad_norm": 2.191978633874912,
+      "learning_rate": 1.3617241749056693e-05,
+      "loss": 0.6603,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8016832440703902,
+      "grad_norm": 2.236976470780728,
+      "learning_rate": 1.3614931264721464e-05,
+      "loss": 0.7457,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8018362662586075,
+      "grad_norm": 2.348713734882274,
+      "learning_rate": 1.3612620558379164e-05,
+      "loss": 0.788,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8019892884468248,
+      "grad_norm": 2.1540503355147136,
+      "learning_rate": 1.3610309630171703e-05,
+      "loss": 0.6961,
+      "step": 5241
+    },
+    {
+      "epoch": 0.802142310635042,
+      "grad_norm": 2.2481983626353297,
+      "learning_rate": 1.3607998480241005e-05,
+      "loss": 0.7852,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8022953328232594,
+      "grad_norm": 2.223866399302239,
+      "learning_rate": 1.3605687108729005e-05,
+      "loss": 0.7054,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8024483550114767,
+      "grad_norm": 2.2526593082756587,
+      "learning_rate": 1.3603375515777654e-05,
+      "loss": 0.7195,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8026013771996939,
+      "grad_norm": 2.1762612602974545,
+      "learning_rate": 1.3601063701528916e-05,
+      "loss": 0.7121,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8027543993879113,
+      "grad_norm": 2.19828314276474,
+      "learning_rate": 1.359875166612477e-05,
+      "loss": 0.6465,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8029074215761285,
+      "grad_norm": 2.1003393586466186,
+      "learning_rate": 1.3596439409707205e-05,
+      "loss": 0.6349,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8030604437643458,
+      "grad_norm": 2.481715046553168,
+      "learning_rate": 1.3594126932418226e-05,
+      "loss": 0.7672,
+      "step": 5248
+    },
+    {
+      "epoch": 0.8032134659525632,
+      "grad_norm": 2.055351929826298,
+      "learning_rate": 1.359181423439985e-05,
+      "loss": 0.6777,
+      "step": 5249
+    },
+    {
+      "epoch": 0.8033664881407804,
+      "grad_norm": 1.9408465558535808,
+      "learning_rate": 1.3589501315794115e-05,
+      "loss": 0.5616,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8035195103289977,
+      "grad_norm": 2.168856933193797,
+      "learning_rate": 1.358718817674306e-05,
+      "loss": 0.7286,
+      "step": 5251
+    },
+    {
+      "epoch": 0.803672532517215,
+      "grad_norm": 2.2388441038555706,
+      "learning_rate": 1.3584874817388744e-05,
+      "loss": 0.6583,
+      "step": 5252
+    },
+    {
+      "epoch": 0.8038255547054323,
+      "grad_norm": 2.638753746731687,
+      "learning_rate": 1.3582561237873244e-05,
+      "loss": 0.7871,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8039785768936496,
+      "grad_norm": 2.173253012550754,
+      "learning_rate": 1.3580247438338643e-05,
+      "loss": 0.6701,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8041315990818668,
+      "grad_norm": 2.373102282978098,
+      "learning_rate": 1.3577933418927039e-05,
+      "loss": 0.8085,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8042846212700842,
+      "grad_norm": 2.1330407791676613,
+      "learning_rate": 1.3575619179780549e-05,
+      "loss": 0.747,
+      "step": 5256
+    },
+    {
+      "epoch": 0.8044376434583015,
+      "grad_norm": 2.1281332913564484,
+      "learning_rate": 1.3573304721041294e-05,
+      "loss": 0.7252,
+      "step": 5257
+    },
+    {
+      "epoch": 0.8045906656465187,
+      "grad_norm": 2.1458569602813062,
+      "learning_rate": 1.3570990042851419e-05,
+      "loss": 0.7456,
+      "step": 5258
+    },
+    {
+      "epoch": 0.804743687834736,
+      "grad_norm": 2.3701157521990894,
+      "learning_rate": 1.3568675145353076e-05,
+      "loss": 0.8018,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8048967100229534,
+      "grad_norm": 2.478264328190266,
+      "learning_rate": 1.3566360028688432e-05,
+      "loss": 0.7114,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8050497322111706,
+      "grad_norm": 2.2845510559676074,
+      "learning_rate": 1.3564044692999667e-05,
+      "loss": 0.7534,
+      "step": 5261
+    },
+    {
+      "epoch": 0.8052027543993879,
+      "grad_norm": 2.0041111447825783,
+      "learning_rate": 1.3561729138428974e-05,
+      "loss": 0.6483,
+      "step": 5262
+    },
+    {
+      "epoch": 0.8053557765876052,
+      "grad_norm": 2.419128035536536,
+      "learning_rate": 1.3559413365118563e-05,
+      "loss": 0.6902,
+      "step": 5263
+    },
+    {
+      "epoch": 0.8055087987758225,
+      "grad_norm": 2.154837398243689,
+      "learning_rate": 1.3557097373210651e-05,
+      "loss": 0.7418,
+      "step": 5264
+    },
+    {
+      "epoch": 0.8056618209640398,
+      "grad_norm": 2.6319673770565983,
+      "learning_rate": 1.3554781162847472e-05,
+      "loss": 0.7431,
+      "step": 5265
+    },
+    {
+      "epoch": 0.805814843152257,
+      "grad_norm": 2.1323745296359387,
+      "learning_rate": 1.355246473417128e-05,
+      "loss": 0.6952,
+      "step": 5266
+    },
+    {
+      "epoch": 0.8059678653404744,
+      "grad_norm": 2.1242915789939865,
+      "learning_rate": 1.3550148087324329e-05,
+      "loss": 0.7627,
+      "step": 5267
+    },
+    {
+      "epoch": 0.8061208875286917,
+      "grad_norm": 2.72482323781392,
+      "learning_rate": 1.3547831222448893e-05,
+      "loss": 0.789,
+      "step": 5268
+    },
+    {
+      "epoch": 0.8062739097169089,
+      "grad_norm": 2.357414573913534,
+      "learning_rate": 1.3545514139687264e-05,
+      "loss": 0.7448,
+      "step": 5269
+    },
+    {
+      "epoch": 0.8064269319051263,
+      "grad_norm": 2.169457925241073,
+      "learning_rate": 1.3543196839181745e-05,
+      "loss": 0.6584,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8065799540933435,
+      "grad_norm": 2.036547827305924,
+      "learning_rate": 1.3540879321074642e-05,
+      "loss": 0.5961,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8067329762815608,
+      "grad_norm": 2.3964724268378363,
+      "learning_rate": 1.3538561585508292e-05,
+      "loss": 0.8957,
+      "step": 5272
+    },
+    {
+      "epoch": 0.8068859984697782,
+      "grad_norm": 2.3452218195555137,
+      "learning_rate": 1.353624363262503e-05,
+      "loss": 0.6594,
+      "step": 5273
+    },
+    {
+      "epoch": 0.8070390206579954,
+      "grad_norm": 2.1038872850157735,
+      "learning_rate": 1.3533925462567213e-05,
+      "loss": 0.6078,
+      "step": 5274
+    },
+    {
+      "epoch": 0.8071920428462127,
+      "grad_norm": 2.356954054341592,
+      "learning_rate": 1.3531607075477209e-05,
+      "loss": 0.7797,
+      "step": 5275
+    },
+    {
+      "epoch": 0.80734506503443,
+      "grad_norm": 2.0115247230251816,
+      "learning_rate": 1.3529288471497399e-05,
+      "loss": 0.5769,
+      "step": 5276
+    },
+    {
+      "epoch": 0.8074980872226473,
+      "grad_norm": 2.276904775969954,
+      "learning_rate": 1.3526969650770175e-05,
+      "loss": 0.7503,
+      "step": 5277
+    },
+    {
+      "epoch": 0.8076511094108646,
+      "grad_norm": 2.167203694480653,
+      "learning_rate": 1.3524650613437948e-05,
+      "loss": 0.6904,
+      "step": 5278
+    },
+    {
+      "epoch": 0.8078041315990818,
+      "grad_norm": 1.9616634696078104,
+      "learning_rate": 1.3522331359643141e-05,
+      "loss": 0.6396,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8079571537872992,
+      "grad_norm": 2.4533852708261406,
+      "learning_rate": 1.3520011889528185e-05,
+      "loss": 0.7826,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8081101759755165,
+      "grad_norm": 2.5398426188480645,
+      "learning_rate": 1.3517692203235526e-05,
+      "loss": 0.7062,
+      "step": 5281
+    },
+    {
+      "epoch": 0.8082631981637337,
+      "grad_norm": 2.0377247734089594,
+      "learning_rate": 1.351537230090763e-05,
+      "loss": 0.6129,
+      "step": 5282
+    },
+    {
+      "epoch": 0.808416220351951,
+      "grad_norm": 2.5926738747286517,
+      "learning_rate": 1.3513052182686968e-05,
+      "loss": 0.7073,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8085692425401684,
+      "grad_norm": 2.104975770665454,
+      "learning_rate": 1.3510731848716028e-05,
+      "loss": 0.8069,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8087222647283856,
+      "grad_norm": 2.1377050518059746,
+      "learning_rate": 1.350841129913731e-05,
+      "loss": 0.717,
+      "step": 5285
+    },
+    {
+      "epoch": 0.8088752869166029,
+      "grad_norm": 2.2039599414867874,
+      "learning_rate": 1.350609053409333e-05,
+      "loss": 0.6571,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8090283091048202,
+      "grad_norm": 2.1174602800748397,
+      "learning_rate": 1.3503769553726616e-05,
+      "loss": 0.6212,
+      "step": 5287
+    },
+    {
+      "epoch": 0.8091813312930375,
+      "grad_norm": 2.547319193448088,
+      "learning_rate": 1.3501448358179705e-05,
+      "loss": 0.8006,
+      "step": 5288
+    },
+    {
+      "epoch": 0.8093343534812548,
+      "grad_norm": 2.3880553798277693,
+      "learning_rate": 1.3499126947595154e-05,
+      "loss": 0.6515,
+      "step": 5289
+    },
+    {
+      "epoch": 0.809487375669472,
+      "grad_norm": 2.220555933056926,
+      "learning_rate": 1.3496805322115525e-05,
+      "loss": 0.7402,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8096403978576894,
+      "grad_norm": 2.1418695856415404,
+      "learning_rate": 1.3494483481883408e-05,
+      "loss": 0.6952,
+      "step": 5291
+    },
+    {
+      "epoch": 0.8097934200459067,
+      "grad_norm": 2.280869004587261,
+      "learning_rate": 1.3492161427041385e-05,
+      "loss": 0.6293,
+      "step": 5292
+    },
+    {
+      "epoch": 0.8099464422341239,
+      "grad_norm": 2.1942676446078466,
+      "learning_rate": 1.3489839157732067e-05,
+      "loss": 0.6146,
+      "step": 5293
+    },
+    {
+      "epoch": 0.8100994644223413,
+      "grad_norm": 1.964131045316613,
+      "learning_rate": 1.3487516674098076e-05,
+      "loss": 0.6257,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8102524866105585,
+      "grad_norm": 2.192523042159634,
+      "learning_rate": 1.3485193976282045e-05,
+      "loss": 0.7152,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8104055087987758,
+      "grad_norm": 2.2224555260794325,
+      "learning_rate": 1.3482871064426612e-05,
+      "loss": 0.7834,
+      "step": 5296
+    },
+    {
+      "epoch": 0.8105585309869932,
+      "grad_norm": 2.3486851929486092,
+      "learning_rate": 1.3480547938674446e-05,
+      "loss": 0.7778,
+      "step": 5297
+    },
+    {
+      "epoch": 0.8107115531752104,
+      "grad_norm": 2.066083356496167,
+      "learning_rate": 1.3478224599168215e-05,
+      "loss": 0.6438,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8108645753634277,
+      "grad_norm": 2.172729196692573,
+      "learning_rate": 1.3475901046050603e-05,
+      "loss": 0.7957,
+      "step": 5299
+    },
+    {
+      "epoch": 0.811017597551645,
+      "grad_norm": 2.0121964473406675,
+      "learning_rate": 1.3473577279464309e-05,
+      "loss": 0.7376,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8111706197398623,
+      "grad_norm": 2.5865207720030368,
+      "learning_rate": 1.3471253299552049e-05,
+      "loss": 0.7715,
+      "step": 5301
+    },
+    {
+      "epoch": 0.8113236419280796,
+      "grad_norm": 2.2695859310630047,
+      "learning_rate": 1.3468929106456543e-05,
+      "loss": 0.7316,
+      "step": 5302
+    },
+    {
+      "epoch": 0.8114766641162968,
+      "grad_norm": 2.212371993004032,
+      "learning_rate": 1.3466604700320529e-05,
+      "loss": 0.8111,
+      "step": 5303
+    },
+    {
+      "epoch": 0.8116296863045142,
+      "grad_norm": 2.7667086132112297,
+      "learning_rate": 1.346428008128676e-05,
+      "loss": 0.7738,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8117827084927315,
+      "grad_norm": 2.242386698700782,
+      "learning_rate": 1.3461955249497998e-05,
+      "loss": 0.6502,
+      "step": 5305
+    },
+    {
+      "epoch": 0.8119357306809487,
+      "grad_norm": 2.3940374260343606,
+      "learning_rate": 1.3459630205097018e-05,
+      "loss": 0.6801,
+      "step": 5306
+    },
+    {
+      "epoch": 0.812088752869166,
+      "grad_norm": 2.4788937521650114,
+      "learning_rate": 1.345730494822662e-05,
+      "loss": 0.6994,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8122417750573834,
+      "grad_norm": 2.135678351797468,
+      "learning_rate": 1.3454979479029595e-05,
+      "loss": 0.7326,
+      "step": 5308
+    },
+    {
+      "epoch": 0.8123947972456006,
+      "grad_norm": 2.380508428475471,
+      "learning_rate": 1.3452653797648765e-05,
+      "loss": 0.7547,
+      "step": 5309
+    },
+    {
+      "epoch": 0.8125478194338179,
+      "grad_norm": 2.292228512939073,
+      "learning_rate": 1.3450327904226956e-05,
+      "loss": 0.7265,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8127008416220352,
+      "grad_norm": 2.6999034724742144,
+      "learning_rate": 1.3448001798907017e-05,
+      "loss": 0.6915,
+      "step": 5311
+    },
+    {
+      "epoch": 0.8128538638102525,
+      "grad_norm": 2.277333877537579,
+      "learning_rate": 1.3445675481831798e-05,
+      "loss": 0.61,
+      "step": 5312
+    },
+    {
+      "epoch": 0.8130068859984698,
+      "grad_norm": 2.179731789967011,
+      "learning_rate": 1.3443348953144166e-05,
+      "loss": 0.6815,
+      "step": 5313
+    },
+    {
+      "epoch": 0.813159908186687,
+      "grad_norm": 2.1453547823586185,
+      "learning_rate": 1.3441022212987008e-05,
+      "loss": 0.6212,
+      "step": 5314
+    },
+    {
+      "epoch": 0.8133129303749044,
+      "grad_norm": 2.0990096504871953,
+      "learning_rate": 1.343869526150321e-05,
+      "loss": 0.621,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8134659525631217,
+      "grad_norm": 2.181671391425356,
+      "learning_rate": 1.3436368098835689e-05,
+      "loss": 0.612,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8136189747513389,
+      "grad_norm": 2.16318928001326,
+      "learning_rate": 1.3434040725127355e-05,
+      "loss": 0.7547,
+      "step": 5317
+    },
+    {
+      "epoch": 0.8137719969395563,
+      "grad_norm": 2.5676933895963674,
+      "learning_rate": 1.3431713140521146e-05,
+      "loss": 0.9013,
+      "step": 5318
+    },
+    {
+      "epoch": 0.8139250191277735,
+      "grad_norm": 2.342905564848808,
+      "learning_rate": 1.3429385345160012e-05,
+      "loss": 0.7104,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8140780413159908,
+      "grad_norm": 2.142110911327244,
+      "learning_rate": 1.3427057339186906e-05,
+      "loss": 0.7011,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8142310635042082,
+      "grad_norm": 2.2067029947839423,
+      "learning_rate": 1.34247291227448e-05,
+      "loss": 0.6377,
+      "step": 5321
+    },
+    {
+      "epoch": 0.8143840856924254,
+      "grad_norm": 2.2040511888427496,
+      "learning_rate": 1.3422400695976685e-05,
+      "loss": 0.5819,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8145371078806427,
+      "grad_norm": 2.1632943457971745,
+      "learning_rate": 1.3420072059025552e-05,
+      "loss": 0.6302,
+      "step": 5323
+    },
+    {
+      "epoch": 0.8146901300688599,
+      "grad_norm": 2.2258275335537303,
+      "learning_rate": 1.3417743212034415e-05,
+      "loss": 0.8637,
+      "step": 5324
+    },
+    {
+      "epoch": 0.8148431522570773,
+      "grad_norm": 2.3153241322316,
+      "learning_rate": 1.3415414155146299e-05,
+      "loss": 0.8324,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8149961744452946,
+      "grad_norm": 2.3974769360130863,
+      "learning_rate": 1.3413084888504236e-05,
+      "loss": 0.755,
+      "step": 5326
+    },
+    {
+      "epoch": 0.8151491966335118,
+      "grad_norm": 2.2510661996362904,
+      "learning_rate": 1.341075541225128e-05,
+      "loss": 0.6549,
+      "step": 5327
+    },
+    {
+      "epoch": 0.8153022188217292,
+      "grad_norm": 2.279833694007899,
+      "learning_rate": 1.340842572653049e-05,
+      "loss": 0.7515,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8154552410099465,
+      "grad_norm": 2.2168201833394465,
+      "learning_rate": 1.3406095831484943e-05,
+      "loss": 0.748,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8156082631981637,
+      "grad_norm": 2.302319672441321,
+      "learning_rate": 1.3403765727257724e-05,
+      "loss": 0.7128,
+      "step": 5330
+    },
+    {
+      "epoch": 0.815761285386381,
+      "grad_norm": 2.3865943289140295,
+      "learning_rate": 1.3401435413991936e-05,
+      "loss": 0.7773,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8159143075745983,
+      "grad_norm": 2.2644122607612918,
+      "learning_rate": 1.3399104891830698e-05,
+      "loss": 0.6749,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8160673297628156,
+      "grad_norm": 2.220283646827807,
+      "learning_rate": 1.3396774160917128e-05,
+      "loss": 0.6206,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8162203519510329,
+      "grad_norm": 2.2493947518081145,
+      "learning_rate": 1.3394443221394366e-05,
+      "loss": 0.743,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8163733741392502,
+      "grad_norm": 2.4348049499278828,
+      "learning_rate": 1.339211207340557e-05,
+      "loss": 0.8518,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8165263963274675,
+      "grad_norm": 2.2335021156921036,
+      "learning_rate": 1.33897807170939e-05,
+      "loss": 0.8099,
+      "step": 5336
+    },
+    {
+      "epoch": 0.8166794185156848,
+      "grad_norm": 2.2602615444390515,
+      "learning_rate": 1.3387449152602536e-05,
+      "loss": 0.744,
+      "step": 5337
+    },
+    {
+      "epoch": 0.816832440703902,
+      "grad_norm": 2.483487434991995,
+      "learning_rate": 1.3385117380074668e-05,
+      "loss": 0.6676,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8169854628921194,
+      "grad_norm": 2.252465139032877,
+      "learning_rate": 1.3382785399653498e-05,
+      "loss": 0.6504,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8171384850803366,
+      "grad_norm": 2.1653703088602487,
+      "learning_rate": 1.3380453211482243e-05,
+      "loss": 0.6743,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8172915072685539,
+      "grad_norm": 2.36517463966179,
+      "learning_rate": 1.3378120815704132e-05,
+      "loss": 0.7109,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8174445294567713,
+      "grad_norm": 2.15790281403125,
+      "learning_rate": 1.3375788212462408e-05,
+      "loss": 0.6949,
+      "step": 5342
+    },
+    {
+      "epoch": 0.8175975516449885,
+      "grad_norm": 2.0850769500930286,
+      "learning_rate": 1.3373455401900323e-05,
+      "loss": 0.6332,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8177505738332058,
+      "grad_norm": 2.1616905772965547,
+      "learning_rate": 1.3371122384161144e-05,
+      "loss": 0.7172,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8179035960214232,
+      "grad_norm": 2.4995256490816633,
+      "learning_rate": 1.3368789159388151e-05,
+      "loss": 0.8572,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8180566182096404,
+      "grad_norm": 2.178755972070649,
+      "learning_rate": 1.3366455727724638e-05,
+      "loss": 0.6471,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8182096403978577,
+      "grad_norm": 2.061965745724041,
+      "learning_rate": 1.3364122089313907e-05,
+      "loss": 0.6209,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8183626625860749,
+      "grad_norm": 2.0978327032670046,
+      "learning_rate": 1.3361788244299281e-05,
+      "loss": 0.6965,
+      "step": 5348
+    },
+    {
+      "epoch": 0.8185156847742923,
+      "grad_norm": 2.1655873920099435,
+      "learning_rate": 1.3359454192824088e-05,
+      "loss": 0.7474,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8186687069625096,
+      "grad_norm": 2.1707296467790003,
+      "learning_rate": 1.3357119935031668e-05,
+      "loss": 0.6475,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8188217291507268,
+      "grad_norm": 2.4511219553404624,
+      "learning_rate": 1.3354785471065382e-05,
+      "loss": 0.7043,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8189747513389442,
+      "grad_norm": 2.5976694677220986,
+      "learning_rate": 1.3352450801068595e-05,
+      "loss": 0.86,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8191277735271615,
+      "grad_norm": 2.395094174010329,
+      "learning_rate": 1.3350115925184688e-05,
+      "loss": 0.7383,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8192807957153787,
+      "grad_norm": 2.3064301975807098,
+      "learning_rate": 1.334778084355706e-05,
+      "loss": 0.7484,
+      "step": 5354
+    },
+    {
+      "epoch": 0.819433817903596,
+      "grad_norm": 2.2933955948779174,
+      "learning_rate": 1.3345445556329111e-05,
+      "loss": 0.6784,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8195868400918133,
+      "grad_norm": 2.2922480591575276,
+      "learning_rate": 1.334311006364426e-05,
+      "loss": 0.793,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8197398622800306,
+      "grad_norm": 2.157361043385925,
+      "learning_rate": 1.3340774365645948e-05,
+      "loss": 0.649,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8198928844682479,
+      "grad_norm": 2.4028534857255877,
+      "learning_rate": 1.333843846247761e-05,
+      "loss": 0.7454,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8200459066564652,
+      "grad_norm": 2.2779711254618897,
+      "learning_rate": 1.3336102354282706e-05,
+      "loss": 0.6664,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8201989288446825,
+      "grad_norm": 2.1582367759018704,
+      "learning_rate": 1.3333766041204705e-05,
+      "loss": 0.6144,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8203519510328998,
+      "grad_norm": 2.302800374343068,
+      "learning_rate": 1.3331429523387091e-05,
+      "loss": 0.5859,
+      "step": 5361
+    },
+    {
+      "epoch": 0.820504973221117,
+      "grad_norm": 2.1501206310043766,
+      "learning_rate": 1.3329092800973356e-05,
+      "loss": 0.6121,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8206579954093344,
+      "grad_norm": 2.3631056301619995,
+      "learning_rate": 1.332675587410701e-05,
+      "loss": 0.7051,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8208110175975516,
+      "grad_norm": 2.3421798133262715,
+      "learning_rate": 1.3324418742931568e-05,
+      "loss": 0.6834,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8209640397857689,
+      "grad_norm": 2.438015095671974,
+      "learning_rate": 1.332208140759057e-05,
+      "loss": 0.686,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8211170619739863,
+      "grad_norm": 2.3411328591375584,
+      "learning_rate": 1.3319743868227555e-05,
+      "loss": 0.7772,
+      "step": 5366
+    },
+    {
+      "epoch": 0.8212700841622035,
+      "grad_norm": 2.4243773349249,
+      "learning_rate": 1.3317406124986083e-05,
+      "loss": 0.7733,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8214231063504208,
+      "grad_norm": 2.6127991053589907,
+      "learning_rate": 1.3315068178009722e-05,
+      "loss": 0.611,
+      "step": 5368
+    },
+    {
+      "epoch": 0.8215761285386382,
+      "grad_norm": 2.3139864560744186,
+      "learning_rate": 1.3312730027442057e-05,
+      "loss": 0.7795,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8217291507268554,
+      "grad_norm": 2.3885664294242295,
+      "learning_rate": 1.3310391673426681e-05,
+      "loss": 0.7107,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8218821729150727,
+      "grad_norm": 2.4219894732303304,
+      "learning_rate": 1.3308053116107202e-05,
+      "loss": 0.6879,
+      "step": 5371
+    },
+    {
+      "epoch": 0.8220351951032899,
+      "grad_norm": 2.5090665576091546,
+      "learning_rate": 1.3305714355627237e-05,
+      "loss": 0.7902,
+      "step": 5372
+    },
+    {
+      "epoch": 0.8221882172915073,
+      "grad_norm": 2.470907543441836,
+      "learning_rate": 1.3303375392130427e-05,
+      "loss": 0.6743,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8223412394797246,
+      "grad_norm": 1.9891713032992344,
+      "learning_rate": 1.3301036225760408e-05,
+      "loss": 0.5842,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8224942616679418,
+      "grad_norm": 2.1719841885826554,
+      "learning_rate": 1.329869685666084e-05,
+      "loss": 0.7156,
+      "step": 5375
+    },
+    {
+      "epoch": 0.8226472838561592,
+      "grad_norm": 2.6093916724723134,
+      "learning_rate": 1.3296357284975395e-05,
+      "loss": 0.8111,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8228003060443765,
+      "grad_norm": 2.348866234796436,
+      "learning_rate": 1.3294017510847754e-05,
+      "loss": 0.7498,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8229533282325937,
+      "grad_norm": 2.4779202227153503,
+      "learning_rate": 1.3291677534421608e-05,
+      "loss": 0.7308,
+      "step": 5378
+    },
+    {
+      "epoch": 0.823106350420811,
+      "grad_norm": 2.439470281437543,
+      "learning_rate": 1.328933735584067e-05,
+      "loss": 0.8282,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8232593726090283,
+      "grad_norm": 1.9490028637280303,
+      "learning_rate": 1.3286996975248659e-05,
+      "loss": 0.5773,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8234123947972456,
+      "grad_norm": 2.17675951263559,
+      "learning_rate": 1.3284656392789301e-05,
+      "loss": 0.6001,
+      "step": 5381
+    },
+    {
+      "epoch": 0.8235654169854629,
+      "grad_norm": 2.5997970873418765,
+      "learning_rate": 1.3282315608606347e-05,
+      "loss": 0.6896,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8237184391736802,
+      "grad_norm": 2.1650822698710694,
+      "learning_rate": 1.327997462284355e-05,
+      "loss": 0.624,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8238714613618975,
+      "grad_norm": 2.1112801781705475,
+      "learning_rate": 1.3277633435644681e-05,
+      "loss": 0.6039,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8240244835501148,
+      "grad_norm": 2.3871543648388194,
+      "learning_rate": 1.3275292047153516e-05,
+      "loss": 0.7194,
+      "step": 5385
+    },
+    {
+      "epoch": 0.824177505738332,
+      "grad_norm": 2.2010526920746276,
+      "learning_rate": 1.327295045751386e-05,
+      "loss": 0.7322,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8243305279265494,
+      "grad_norm": 2.775859697782073,
+      "learning_rate": 1.3270608666869512e-05,
+      "loss": 0.7952,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8244835501147666,
+      "grad_norm": 2.276716590063767,
+      "learning_rate": 1.3268266675364285e-05,
+      "loss": 0.6301,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8246365723029839,
+      "grad_norm": 2.2539606958094787,
+      "learning_rate": 1.3265924483142021e-05,
+      "loss": 0.726,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8247895944912013,
+      "grad_norm": 2.127389941524374,
+      "learning_rate": 1.3263582090346559e-05,
+      "loss": 0.607,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8249426166794185,
+      "grad_norm": 2.2618517630927593,
+      "learning_rate": 1.3261239497121747e-05,
+      "loss": 0.7103,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8250956388676358,
+      "grad_norm": 2.301198924576191,
+      "learning_rate": 1.3258896703611465e-05,
+      "loss": 0.674,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8252486610558531,
+      "grad_norm": 2.2519800323299823,
+      "learning_rate": 1.3256553709959588e-05,
+      "loss": 0.7118,
+      "step": 5393
+    },
+    {
+      "epoch": 0.8254016832440704,
+      "grad_norm": 2.1158962008831104,
+      "learning_rate": 1.3254210516310004e-05,
+      "loss": 0.6324,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8255547054322877,
+      "grad_norm": 2.1458773057151816,
+      "learning_rate": 1.3251867122806625e-05,
+      "loss": 0.6327,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8257077276205049,
+      "grad_norm": 2.6875604486528704,
+      "learning_rate": 1.3249523529593364e-05,
+      "loss": 0.7069,
+      "step": 5396
+    },
+    {
+      "epoch": 0.8258607498087223,
+      "grad_norm": 2.4063745136874592,
+      "learning_rate": 1.3247179736814149e-05,
+      "loss": 0.6958,
+      "step": 5397
+    },
+    {
+      "epoch": 0.8260137719969396,
+      "grad_norm": 2.2108085408792446,
+      "learning_rate": 1.3244835744612926e-05,
+      "loss": 0.7354,
+      "step": 5398
+    },
+    {
+      "epoch": 0.8261667941851568,
+      "grad_norm": 2.1647786138967207,
+      "learning_rate": 1.3242491553133646e-05,
+      "loss": 0.7654,
+      "step": 5399
+    },
+    {
+      "epoch": 0.8263198163733741,
+      "grad_norm": 2.2781294249844355,
+      "learning_rate": 1.3240147162520272e-05,
+      "loss": 0.69,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8264728385615915,
+      "grad_norm": 2.205081890912215,
+      "learning_rate": 1.3237802572916786e-05,
+      "loss": 0.7011,
+      "step": 5401
+    },
+    {
+      "epoch": 0.8266258607498087,
+      "grad_norm": 2.156765525326034,
+      "learning_rate": 1.3235457784467183e-05,
+      "loss": 0.7145,
+      "step": 5402
+    },
+    {
+      "epoch": 0.826778882938026,
+      "grad_norm": 2.0555102520992077,
+      "learning_rate": 1.3233112797315456e-05,
+      "loss": 0.5552,
+      "step": 5403
+    },
+    {
+      "epoch": 0.8269319051262433,
+      "grad_norm": 2.178522116336929,
+      "learning_rate": 1.3230767611605622e-05,
+      "loss": 0.7269,
+      "step": 5404
+    },
+    {
+      "epoch": 0.8270849273144606,
+      "grad_norm": 2.0747364623531306,
+      "learning_rate": 1.3228422227481712e-05,
+      "loss": 0.6244,
+      "step": 5405
+    },
+    {
+      "epoch": 0.8272379495026779,
+      "grad_norm": 2.2732975783479596,
+      "learning_rate": 1.3226076645087765e-05,
+      "loss": 0.6528,
+      "step": 5406
+    },
+    {
+      "epoch": 0.8273909716908951,
+      "grad_norm": 2.032788279412093,
+      "learning_rate": 1.3223730864567829e-05,
+      "loss": 0.6503,
+      "step": 5407
+    },
+    {
+      "epoch": 0.8275439938791125,
+      "grad_norm": 2.186247863682011,
+      "learning_rate": 1.3221384886065968e-05,
+      "loss": 0.7322,
+      "step": 5408
+    },
+    {
+      "epoch": 0.8276970160673298,
+      "grad_norm": 2.048505319281304,
+      "learning_rate": 1.3219038709726261e-05,
+      "loss": 0.6557,
+      "step": 5409
+    },
+    {
+      "epoch": 0.827850038255547,
+      "grad_norm": 2.1651304187113487,
+      "learning_rate": 1.3216692335692796e-05,
+      "loss": 0.6042,
+      "step": 5410
+    },
+    {
+      "epoch": 0.8280030604437644,
+      "grad_norm": 2.3706714888571576,
+      "learning_rate": 1.3214345764109668e-05,
+      "loss": 0.7171,
+      "step": 5411
+    },
+    {
+      "epoch": 0.8281560826319816,
+      "grad_norm": 2.144590330064656,
+      "learning_rate": 1.321199899512099e-05,
+      "loss": 0.5572,
+      "step": 5412
+    },
+    {
+      "epoch": 0.8283091048201989,
+      "grad_norm": 2.3176192449312727,
+      "learning_rate": 1.3209652028870891e-05,
+      "loss": 0.6542,
+      "step": 5413
+    },
+    {
+      "epoch": 0.8284621270084163,
+      "grad_norm": 2.4363610612490225,
+      "learning_rate": 1.3207304865503508e-05,
+      "loss": 0.6725,
+      "step": 5414
+    },
+    {
+      "epoch": 0.8286151491966335,
+      "grad_norm": 2.239550593859622,
+      "learning_rate": 1.3204957505162983e-05,
+      "loss": 0.6656,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8287681713848508,
+      "grad_norm": 2.2066515473547743,
+      "learning_rate": 1.3202609947993477e-05,
+      "loss": 0.7237,
+      "step": 5416
+    },
+    {
+      "epoch": 0.8289211935730681,
+      "grad_norm": 2.4805510088889613,
+      "learning_rate": 1.3200262194139168e-05,
+      "loss": 0.7478,
+      "step": 5417
+    },
+    {
+      "epoch": 0.8290742157612854,
+      "grad_norm": 2.1960028509697733,
+      "learning_rate": 1.3197914243744237e-05,
+      "loss": 0.6824,
+      "step": 5418
+    },
+    {
+      "epoch": 0.8292272379495027,
+      "grad_norm": 2.3511494818838456,
+      "learning_rate": 1.3195566096952883e-05,
+      "loss": 0.7456,
+      "step": 5419
+    },
+    {
+      "epoch": 0.8293802601377199,
+      "grad_norm": 2.249239523564754,
+      "learning_rate": 1.3193217753909316e-05,
+      "loss": 0.7507,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8295332823259373,
+      "grad_norm": 1.9415800244217254,
+      "learning_rate": 1.3190869214757751e-05,
+      "loss": 0.684,
+      "step": 5421
+    },
+    {
+      "epoch": 0.8296863045141546,
+      "grad_norm": 2.3205062630225353,
+      "learning_rate": 1.3188520479642426e-05,
+      "loss": 0.7321,
+      "step": 5422
+    },
+    {
+      "epoch": 0.8298393267023718,
+      "grad_norm": 2.400336439716699,
+      "learning_rate": 1.3186171548707587e-05,
+      "loss": 0.6574,
+      "step": 5423
+    },
+    {
+      "epoch": 0.8299923488905891,
+      "grad_norm": 2.0677198849039273,
+      "learning_rate": 1.3183822422097486e-05,
+      "loss": 0.634,
+      "step": 5424
+    },
+    {
+      "epoch": 0.8301453710788065,
+      "grad_norm": 2.2584330809992315,
+      "learning_rate": 1.3181473099956397e-05,
+      "loss": 0.6519,
+      "step": 5425
+    },
+    {
+      "epoch": 0.8302983932670237,
+      "grad_norm": 2.124240068152974,
+      "learning_rate": 1.3179123582428594e-05,
+      "loss": 0.6199,
+      "step": 5426
+    },
+    {
+      "epoch": 0.830451415455241,
+      "grad_norm": 2.3273164439976473,
+      "learning_rate": 1.3176773869658383e-05,
+      "loss": 0.6734,
+      "step": 5427
+    },
+    {
+      "epoch": 0.8306044376434583,
+      "grad_norm": 2.149535999528806,
+      "learning_rate": 1.3174423961790057e-05,
+      "loss": 0.6557,
+      "step": 5428
+    },
+    {
+      "epoch": 0.8307574598316756,
+      "grad_norm": 2.481319214602762,
+      "learning_rate": 1.3172073858967935e-05,
+      "loss": 0.7535,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8309104820198929,
+      "grad_norm": 1.9994495591491446,
+      "learning_rate": 1.3169723561336349e-05,
+      "loss": 0.6636,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8310635042081101,
+      "grad_norm": 2.023320842609098,
+      "learning_rate": 1.3167373069039644e-05,
+      "loss": 0.6146,
+      "step": 5431
+    },
+    {
+      "epoch": 0.8312165263963275,
+      "grad_norm": 2.3556441047611125,
+      "learning_rate": 1.3165022382222161e-05,
+      "loss": 0.6654,
+      "step": 5432
+    },
+    {
+      "epoch": 0.8313695485845447,
+      "grad_norm": 2.325532575404804,
+      "learning_rate": 1.3162671501028275e-05,
+      "loss": 0.7858,
+      "step": 5433
+    },
+    {
+      "epoch": 0.831522570772762,
+      "grad_norm": 2.230886816421429,
+      "learning_rate": 1.316032042560236e-05,
+      "loss": 0.7242,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8316755929609794,
+      "grad_norm": 2.1374285377576334,
+      "learning_rate": 1.3157969156088806e-05,
+      "loss": 0.6857,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8318286151491966,
+      "grad_norm": 2.2024531130944087,
+      "learning_rate": 1.315561769263201e-05,
+      "loss": 0.7603,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8319816373374139,
+      "grad_norm": 2.4725523502828186,
+      "learning_rate": 1.3153266035376387e-05,
+      "loss": 0.7892,
+      "step": 5437
+    },
+    {
+      "epoch": 0.8321346595256313,
+      "grad_norm": 1.942259678346756,
+      "learning_rate": 1.3150914184466359e-05,
+      "loss": 0.6643,
+      "step": 5438
+    },
+    {
+      "epoch": 0.8322876817138485,
+      "grad_norm": 1.9805552537307942,
+      "learning_rate": 1.3148562140046368e-05,
+      "loss": 0.6216,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8324407039020658,
+      "grad_norm": 2.0651368907886094,
+      "learning_rate": 1.3146209902260858e-05,
+      "loss": 0.6408,
+      "step": 5440
+    },
+    {
+      "epoch": 0.832593726090283,
+      "grad_norm": 1.8716904461752568,
+      "learning_rate": 1.3143857471254284e-05,
+      "loss": 0.6518,
+      "step": 5441
+    },
+    {
+      "epoch": 0.8327467482785004,
+      "grad_norm": 1.9940867590996474,
+      "learning_rate": 1.3141504847171131e-05,
+      "loss": 0.6708,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8328997704667177,
+      "grad_norm": 2.321143832807135,
+      "learning_rate": 1.3139152030155875e-05,
+      "loss": 0.6505,
+      "step": 5443
+    },
+    {
+      "epoch": 0.8330527926549349,
+      "grad_norm": 2.3643297951532474,
+      "learning_rate": 1.3136799020353009e-05,
+      "loss": 0.7394,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8332058148431523,
+      "grad_norm": 2.4184612753665466,
+      "learning_rate": 1.3134445817907047e-05,
+      "loss": 0.6883,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8333588370313696,
+      "grad_norm": 2.3561896414391144,
+      "learning_rate": 1.3132092422962503e-05,
+      "loss": 0.7043,
+      "step": 5446
+    },
+    {
+      "epoch": 0.8335118592195868,
+      "grad_norm": 2.362497581248584,
+      "learning_rate": 1.3129738835663911e-05,
+      "loss": 0.7542,
+      "step": 5447
+    },
+    {
+      "epoch": 0.8336648814078041,
+      "grad_norm": 2.4107911094529677,
+      "learning_rate": 1.3127385056155818e-05,
+      "loss": 0.7097,
+      "step": 5448
+    },
+    {
+      "epoch": 0.8338179035960214,
+      "grad_norm": 2.151079347885521,
+      "learning_rate": 1.3125031084582769e-05,
+      "loss": 0.677,
+      "step": 5449
+    },
+    {
+      "epoch": 0.8339709257842387,
+      "grad_norm": 2.113665557525543,
+      "learning_rate": 1.3122676921089338e-05,
+      "loss": 0.6911,
+      "step": 5450
+    },
+    {
+      "epoch": 0.834123947972456,
+      "grad_norm": 2.089972690894945,
+      "learning_rate": 1.3120322565820098e-05,
+      "loss": 0.6565,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8342769701606733,
+      "grad_norm": 2.3308970486598404,
+      "learning_rate": 1.3117968018919646e-05,
+      "loss": 0.8009,
+      "step": 5452
+    },
+    {
+      "epoch": 0.8344299923488906,
+      "grad_norm": 2.233752104710811,
+      "learning_rate": 1.3115613280532581e-05,
+      "loss": 0.6564,
+      "step": 5453
+    },
+    {
+      "epoch": 0.8345830145371079,
+      "grad_norm": 2.341965814021548,
+      "learning_rate": 1.3113258350803513e-05,
+      "loss": 0.7691,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8347360367253251,
+      "grad_norm": 2.495347947770681,
+      "learning_rate": 1.3110903229877073e-05,
+      "loss": 0.7702,
+      "step": 5455
+    },
+    {
+      "epoch": 0.8348890589135425,
+      "grad_norm": 2.229124461211568,
+      "learning_rate": 1.3108547917897896e-05,
+      "loss": 0.779,
+      "step": 5456
+    },
+    {
+      "epoch": 0.8350420811017597,
+      "grad_norm": 2.404664452229642,
+      "learning_rate": 1.3106192415010628e-05,
+      "loss": 0.7439,
+      "step": 5457
+    },
+    {
+      "epoch": 0.835195103289977,
+      "grad_norm": 2.144379029046011,
+      "learning_rate": 1.3103836721359934e-05,
+      "loss": 0.7058,
+      "step": 5458
+    },
+    {
+      "epoch": 0.8353481254781944,
+      "grad_norm": 2.2023641817367934,
+      "learning_rate": 1.3101480837090485e-05,
+      "loss": 0.6623,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8355011476664116,
+      "grad_norm": 2.0170490475868936,
+      "learning_rate": 1.3099124762346965e-05,
+      "loss": 0.6674,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8356541698546289,
+      "grad_norm": 1.9195685937200395,
+      "learning_rate": 1.3096768497274069e-05,
+      "loss": 0.6357,
+      "step": 5461
+    },
+    {
+      "epoch": 0.8358071920428463,
+      "grad_norm": 2.024436830598696,
+      "learning_rate": 1.3094412042016504e-05,
+      "loss": 0.5846,
+      "step": 5462
+    },
+    {
+      "epoch": 0.8359602142310635,
+      "grad_norm": 2.198755705976839,
+      "learning_rate": 1.3092055396718992e-05,
+      "loss": 0.7809,
+      "step": 5463
+    },
+    {
+      "epoch": 0.8361132364192808,
+      "grad_norm": 2.2113261586270485,
+      "learning_rate": 1.3089698561526261e-05,
+      "loss": 0.7037,
+      "step": 5464
+    },
+    {
+      "epoch": 0.836266258607498,
+      "grad_norm": 2.354039463335677,
+      "learning_rate": 1.3087341536583054e-05,
+      "loss": 0.8019,
+      "step": 5465
+    },
+    {
+      "epoch": 0.8364192807957154,
+      "grad_norm": 2.2564806325771194,
+      "learning_rate": 1.3084984322034124e-05,
+      "loss": 0.9169,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8365723029839327,
+      "grad_norm": 3.4017957008245157,
+      "learning_rate": 1.3082626918024239e-05,
+      "loss": 0.6874,
+      "step": 5467
+    },
+    {
+      "epoch": 0.8367253251721499,
+      "grad_norm": 2.2833142256607086,
+      "learning_rate": 1.3080269324698179e-05,
+      "loss": 0.6265,
+      "step": 5468
+    },
+    {
+      "epoch": 0.8368783473603673,
+      "grad_norm": 2.057805293848059,
+      "learning_rate": 1.3077911542200727e-05,
+      "loss": 0.6179,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8370313695485846,
+      "grad_norm": 2.3137603844967924,
+      "learning_rate": 1.3075553570676685e-05,
+      "loss": 0.7012,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8371843917368018,
+      "grad_norm": 2.348687565853293,
+      "learning_rate": 1.3073195410270869e-05,
+      "loss": 0.8127,
+      "step": 5471
+    },
+    {
+      "epoch": 0.8373374139250191,
+      "grad_norm": 2.1805314170172223,
+      "learning_rate": 1.3070837061128101e-05,
+      "loss": 0.7328,
+      "step": 5472
+    },
+    {
+      "epoch": 0.8374904361132364,
+      "grad_norm": 2.298133329589823,
+      "learning_rate": 1.3068478523393213e-05,
+      "loss": 0.7008,
+      "step": 5473
+    },
+    {
+      "epoch": 0.8376434583014537,
+      "grad_norm": 2.130736044585862,
+      "learning_rate": 1.3066119797211056e-05,
+      "loss": 0.5503,
+      "step": 5474
+    },
+    {
+      "epoch": 0.837796480489671,
+      "grad_norm": 2.348461660163417,
+      "learning_rate": 1.3063760882726488e-05,
+      "loss": 0.6586,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8379495026778883,
+      "grad_norm": 2.457109691397983,
+      "learning_rate": 1.3061401780084379e-05,
+      "loss": 0.8545,
+      "step": 5476
+    },
+    {
+      "epoch": 0.8381025248661056,
+      "grad_norm": 2.081413285430723,
+      "learning_rate": 1.305904248942961e-05,
+      "loss": 0.7044,
+      "step": 5477
+    },
+    {
+      "epoch": 0.8382555470543229,
+      "grad_norm": 2.2761005336360443,
+      "learning_rate": 1.3056683010907074e-05,
+      "loss": 0.7146,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8384085692425401,
+      "grad_norm": 2.4427202264353096,
+      "learning_rate": 1.3054323344661675e-05,
+      "loss": 0.7789,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8385615914307575,
+      "grad_norm": 2.5737808915465648,
+      "learning_rate": 1.3051963490838336e-05,
+      "loss": 0.747,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8387146136189747,
+      "grad_norm": 2.3321630387102474,
+      "learning_rate": 1.3049603449581976e-05,
+      "loss": 0.7329,
+      "step": 5481
+    },
+    {
+      "epoch": 0.838867635807192,
+      "grad_norm": 2.2436821204503996,
+      "learning_rate": 1.3047243221037537e-05,
+      "loss": 0.8297,
+      "step": 5482
+    },
+    {
+      "epoch": 0.8390206579954094,
+      "grad_norm": 2.1236816351700916,
+      "learning_rate": 1.3044882805349972e-05,
+      "loss": 0.6615,
+      "step": 5483
+    },
+    {
+      "epoch": 0.8391736801836266,
+      "grad_norm": 2.1357048998496286,
+      "learning_rate": 1.3042522202664247e-05,
+      "loss": 0.7149,
+      "step": 5484
+    },
+    {
+      "epoch": 0.8393267023718439,
+      "grad_norm": 2.276524858878027,
+      "learning_rate": 1.3040161413125325e-05,
+      "loss": 0.7097,
+      "step": 5485
+    },
+    {
+      "epoch": 0.8394797245600613,
+      "grad_norm": 2.20618942502793,
+      "learning_rate": 1.3037800436878199e-05,
+      "loss": 0.6871,
+      "step": 5486
+    },
+    {
+      "epoch": 0.8396327467482785,
+      "grad_norm": 2.2991077527046513,
+      "learning_rate": 1.3035439274067865e-05,
+      "loss": 0.6966,
+      "step": 5487
+    },
+    {
+      "epoch": 0.8397857689364958,
+      "grad_norm": 2.3943691068471344,
+      "learning_rate": 1.303307792483933e-05,
+      "loss": 0.774,
+      "step": 5488
+    },
+    {
+      "epoch": 0.839938791124713,
+      "grad_norm": 2.4172819833948886,
+      "learning_rate": 1.3030716389337614e-05,
+      "loss": 0.7621,
+      "step": 5489
+    },
+    {
+      "epoch": 0.8400918133129304,
+      "grad_norm": 2.0557590442430276,
+      "learning_rate": 1.302835466770775e-05,
+      "loss": 0.6875,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8402448355011477,
+      "grad_norm": 2.1172952300702517,
+      "learning_rate": 1.3025992760094778e-05,
+      "loss": 0.7656,
+      "step": 5491
+    },
+    {
+      "epoch": 0.8403978576893649,
+      "grad_norm": 2.1069628676461667,
+      "learning_rate": 1.302363066664375e-05,
+      "loss": 0.6424,
+      "step": 5492
+    },
+    {
+      "epoch": 0.8405508798775823,
+      "grad_norm": 2.0851667419856033,
+      "learning_rate": 1.302126838749974e-05,
+      "loss": 0.6485,
+      "step": 5493
+    },
+    {
+      "epoch": 0.8407039020657996,
+      "grad_norm": 2.345547762432767,
+      "learning_rate": 1.3018905922807814e-05,
+      "loss": 0.7328,
+      "step": 5494
+    },
+    {
+      "epoch": 0.8408569242540168,
+      "grad_norm": 2.418693674830027,
+      "learning_rate": 1.3016543272713066e-05,
+      "loss": 0.7418,
+      "step": 5495
+    },
+    {
+      "epoch": 0.8410099464422341,
+      "grad_norm": 2.371023973660028,
+      "learning_rate": 1.3014180437360595e-05,
+      "loss": 0.5673,
+      "step": 5496
+    },
+    {
+      "epoch": 0.8411629686304514,
+      "grad_norm": 2.265839060637262,
+      "learning_rate": 1.3011817416895516e-05,
+      "loss": 0.6141,
+      "step": 5497
+    },
+    {
+      "epoch": 0.8413159908186687,
+      "grad_norm": 2.561937095874154,
+      "learning_rate": 1.3009454211462941e-05,
+      "loss": 0.7998,
+      "step": 5498
+    },
+    {
+      "epoch": 0.841469013006886,
+      "grad_norm": 2.264911983777347,
+      "learning_rate": 1.3007090821208013e-05,
+      "loss": 0.6413,
+      "step": 5499
+    },
+    {
+      "epoch": 0.8416220351951033,
+      "grad_norm": 2.38515727129013,
+      "learning_rate": 1.3004727246275875e-05,
+      "loss": 0.6904,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8417750573833206,
+      "grad_norm": 2.2272191122576475,
+      "learning_rate": 1.300236348681168e-05,
+      "loss": 0.67,
+      "step": 5501
+    },
+    {
+      "epoch": 0.8419280795715379,
+      "grad_norm": 2.328983070075472,
+      "learning_rate": 1.2999999542960597e-05,
+      "loss": 0.7334,
+      "step": 5502
+    },
+    {
+      "epoch": 0.8420811017597551,
+      "grad_norm": 2.2285007466959206,
+      "learning_rate": 1.2997635414867808e-05,
+      "loss": 0.7039,
+      "step": 5503
+    },
+    {
+      "epoch": 0.8422341239479725,
+      "grad_norm": 2.475196568822425,
+      "learning_rate": 1.2995271102678503e-05,
+      "loss": 0.6824,
+      "step": 5504
+    },
+    {
+      "epoch": 0.8423871461361897,
+      "grad_norm": 2.145354123730406,
+      "learning_rate": 1.2992906606537878e-05,
+      "loss": 0.6814,
+      "step": 5505
+    },
+    {
+      "epoch": 0.842540168324407,
+      "grad_norm": 2.648650178635418,
+      "learning_rate": 1.2990541926591153e-05,
+      "loss": 0.7099,
+      "step": 5506
+    },
+    {
+      "epoch": 0.8426931905126244,
+      "grad_norm": 2.2924356285560554,
+      "learning_rate": 1.298817706298355e-05,
+      "loss": 0.7504,
+      "step": 5507
+    },
+    {
+      "epoch": 0.8428462127008416,
+      "grad_norm": 2.260371927743499,
+      "learning_rate": 1.2985812015860297e-05,
+      "loss": 0.8017,
+      "step": 5508
+    },
+    {
+      "epoch": 0.8429992348890589,
+      "grad_norm": 2.1438114603897898,
+      "learning_rate": 1.2983446785366656e-05,
+      "loss": 0.6886,
+      "step": 5509
+    },
+    {
+      "epoch": 0.8431522570772763,
+      "grad_norm": 2.328522698350745,
+      "learning_rate": 1.2981081371647872e-05,
+      "loss": 0.6653,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8433052792654935,
+      "grad_norm": 2.2418473468868627,
+      "learning_rate": 1.297871577484922e-05,
+      "loss": 0.7969,
+      "step": 5511
+    },
+    {
+      "epoch": 0.8434583014537108,
+      "grad_norm": 2.079868183628365,
+      "learning_rate": 1.2976349995115979e-05,
+      "loss": 0.6576,
+      "step": 5512
+    },
+    {
+      "epoch": 0.843611323641928,
+      "grad_norm": 2.2551328175381657,
+      "learning_rate": 1.2973984032593446e-05,
+      "loss": 0.7583,
+      "step": 5513
+    },
+    {
+      "epoch": 0.8437643458301454,
+      "grad_norm": 1.9291727183519736,
+      "learning_rate": 1.2971617887426911e-05,
+      "loss": 0.6354,
+      "step": 5514
+    },
+    {
+      "epoch": 0.8439173680183627,
+      "grad_norm": 2.0798484589827226,
+      "learning_rate": 1.2969251559761702e-05,
+      "loss": 0.653,
+      "step": 5515
+    },
+    {
+      "epoch": 0.8440703902065799,
+      "grad_norm": 2.418587228020786,
+      "learning_rate": 1.2966885049743136e-05,
+      "loss": 0.6181,
+      "step": 5516
+    },
+    {
+      "epoch": 0.8442234123947973,
+      "grad_norm": 2.19342314918509,
+      "learning_rate": 1.2964518357516553e-05,
+      "loss": 0.6955,
+      "step": 5517
+    },
+    {
+      "epoch": 0.8443764345830146,
+      "grad_norm": 2.338148065171502,
+      "learning_rate": 1.2962151483227303e-05,
+      "loss": 0.7878,
+      "step": 5518
+    },
+    {
+      "epoch": 0.8445294567712318,
+      "grad_norm": 1.9721749570907623,
+      "learning_rate": 1.295978442702074e-05,
+      "loss": 0.6078,
+      "step": 5519
+    },
+    {
+      "epoch": 0.8446824789594491,
+      "grad_norm": 2.0817212942936845,
+      "learning_rate": 1.2957417189042237e-05,
+      "loss": 0.6564,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8448355011476664,
+      "grad_norm": 2.289786601333639,
+      "learning_rate": 1.2955049769437173e-05,
+      "loss": 0.6076,
+      "step": 5521
+    },
+    {
+      "epoch": 0.8449885233358837,
+      "grad_norm": 2.400216562893098,
+      "learning_rate": 1.2952682168350949e-05,
+      "loss": 0.8167,
+      "step": 5522
+    },
+    {
+      "epoch": 0.845141545524101,
+      "grad_norm": 2.423170432612747,
+      "learning_rate": 1.2950314385928957e-05,
+      "loss": 0.7366,
+      "step": 5523
+    },
+    {
+      "epoch": 0.8452945677123183,
+      "grad_norm": 2.079069856562734,
+      "learning_rate": 1.2947946422316617e-05,
+      "loss": 0.6885,
+      "step": 5524
+    },
+    {
+      "epoch": 0.8454475899005356,
+      "grad_norm": 2.314286472410869,
+      "learning_rate": 1.2945578277659357e-05,
+      "loss": 0.7336,
+      "step": 5525
+    },
+    {
+      "epoch": 0.8456006120887529,
+      "grad_norm": 2.261984315977617,
+      "learning_rate": 1.2943209952102608e-05,
+      "loss": 0.6953,
+      "step": 5526
+    },
+    {
+      "epoch": 0.8457536342769701,
+      "grad_norm": 2.0233508566664793,
+      "learning_rate": 1.2940841445791826e-05,
+      "loss": 0.6739,
+      "step": 5527
+    },
+    {
+      "epoch": 0.8459066564651875,
+      "grad_norm": 2.288650722875104,
+      "learning_rate": 1.2938472758872464e-05,
+      "loss": 0.7251,
+      "step": 5528
+    },
+    {
+      "epoch": 0.8460596786534047,
+      "grad_norm": 2.5941529869810522,
+      "learning_rate": 1.2936103891489995e-05,
+      "loss": 0.6892,
+      "step": 5529
+    },
+    {
+      "epoch": 0.846212700841622,
+      "grad_norm": 2.488575302432984,
+      "learning_rate": 1.29337348437899e-05,
+      "loss": 0.6759,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8463657230298394,
+      "grad_norm": 2.1952089497466396,
+      "learning_rate": 1.293136561591767e-05,
+      "loss": 0.748,
+      "step": 5531
+    },
+    {
+      "epoch": 0.8465187452180566,
+      "grad_norm": 2.2910733629818756,
+      "learning_rate": 1.2928996208018813e-05,
+      "loss": 0.6614,
+      "step": 5532
+    },
+    {
+      "epoch": 0.8466717674062739,
+      "grad_norm": 2.2937953646424787,
+      "learning_rate": 1.2926626620238836e-05,
+      "loss": 0.7134,
+      "step": 5533
+    },
+    {
+      "epoch": 0.8468247895944913,
+      "grad_norm": 2.260542552534389,
+      "learning_rate": 1.2924256852723276e-05,
+      "loss": 0.6862,
+      "step": 5534
+    },
+    {
+      "epoch": 0.8469778117827085,
+      "grad_norm": 2.2815687375045615,
+      "learning_rate": 1.2921886905617658e-05,
+      "loss": 0.6566,
+      "step": 5535
+    },
+    {
+      "epoch": 0.8471308339709258,
+      "grad_norm": 2.355462711241084,
+      "learning_rate": 1.2919516779067533e-05,
+      "loss": 0.7204,
+      "step": 5536
+    },
+    {
+      "epoch": 0.847283856159143,
+      "grad_norm": 2.3383308189087186,
+      "learning_rate": 1.2917146473218464e-05,
+      "loss": 0.76,
+      "step": 5537
+    },
+    {
+      "epoch": 0.8474368783473604,
+      "grad_norm": 2.294371471880201,
+      "learning_rate": 1.2914775988216021e-05,
+      "loss": 0.6685,
+      "step": 5538
+    },
+    {
+      "epoch": 0.8475899005355777,
+      "grad_norm": 2.237221973534265,
+      "learning_rate": 1.2912405324205776e-05,
+      "loss": 0.7479,
+      "step": 5539
+    },
+    {
+      "epoch": 0.8477429227237949,
+      "grad_norm": 2.4085148292399468,
+      "learning_rate": 1.2910034481333331e-05,
+      "loss": 0.6868,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8478959449120123,
+      "grad_norm": 2.1331468654542154,
+      "learning_rate": 1.2907663459744283e-05,
+      "loss": 0.6263,
+      "step": 5541
+    },
+    {
+      "epoch": 0.8480489671002295,
+      "grad_norm": 2.416430462046604,
+      "learning_rate": 1.2905292259584248e-05,
+      "loss": 0.5775,
+      "step": 5542
+    },
+    {
+      "epoch": 0.8482019892884468,
+      "grad_norm": 2.154148517983704,
+      "learning_rate": 1.2902920880998848e-05,
+      "loss": 0.704,
+      "step": 5543
+    },
+    {
+      "epoch": 0.8483550114766641,
+      "grad_norm": 2.221789554064452,
+      "learning_rate": 1.2900549324133722e-05,
+      "loss": 0.6048,
+      "step": 5544
+    },
+    {
+      "epoch": 0.8485080336648814,
+      "grad_norm": 2.48005927838095,
+      "learning_rate": 1.2898177589134515e-05,
+      "loss": 0.715,
+      "step": 5545
+    },
+    {
+      "epoch": 0.8486610558530987,
+      "grad_norm": 2.144770996988103,
+      "learning_rate": 1.2895805676146882e-05,
+      "loss": 0.6509,
+      "step": 5546
+    },
+    {
+      "epoch": 0.848814078041316,
+      "grad_norm": 1.9436185628607763,
+      "learning_rate": 1.2893433585316497e-05,
+      "loss": 0.5823,
+      "step": 5547
+    },
+    {
+      "epoch": 0.8489671002295333,
+      "grad_norm": 2.3484164597566304,
+      "learning_rate": 1.2891061316789036e-05,
+      "loss": 0.6889,
+      "step": 5548
+    },
+    {
+      "epoch": 0.8491201224177506,
+      "grad_norm": 2.4994443211846296,
+      "learning_rate": 1.2888688870710186e-05,
+      "loss": 0.7833,
+      "step": 5549
+    },
+    {
+      "epoch": 0.8492731446059678,
+      "grad_norm": 2.156946521054402,
+      "learning_rate": 1.2886316247225657e-05,
+      "loss": 0.76,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8494261667941851,
+      "grad_norm": 2.1799637821710482,
+      "learning_rate": 1.2883943446481153e-05,
+      "loss": 0.6765,
+      "step": 5551
+    },
+    {
+      "epoch": 0.8495791889824025,
+      "grad_norm": 2.0568551184076886,
+      "learning_rate": 1.28815704686224e-05,
+      "loss": 0.6281,
+      "step": 5552
+    },
+    {
+      "epoch": 0.8497322111706197,
+      "grad_norm": 2.2126970022565726,
+      "learning_rate": 1.2879197313795131e-05,
+      "loss": 0.6563,
+      "step": 5553
+    },
+    {
+      "epoch": 0.849885233358837,
+      "grad_norm": 2.522319909037303,
+      "learning_rate": 1.2876823982145095e-05,
+      "loss": 0.8035,
+      "step": 5554
+    },
+    {
+      "epoch": 0.8500382555470544,
+      "grad_norm": 2.498837348760292,
+      "learning_rate": 1.287445047381804e-05,
+      "loss": 0.7085,
+      "step": 5555
+    },
+    {
+      "epoch": 0.8501912777352716,
+      "grad_norm": 2.363108474252094,
+      "learning_rate": 1.2872076788959738e-05,
+      "loss": 0.706,
+      "step": 5556
+    },
+    {
+      "epoch": 0.8503442999234889,
+      "grad_norm": 2.1018992973643207,
+      "learning_rate": 1.2869702927715966e-05,
+      "loss": 0.6308,
+      "step": 5557
+    },
+    {
+      "epoch": 0.8504973221117061,
+      "grad_norm": 2.5754528738479117,
+      "learning_rate": 1.2867328890232506e-05,
+      "loss": 0.7338,
+      "step": 5558
+    },
+    {
+      "epoch": 0.8506503442999235,
+      "grad_norm": 2.2465891179344237,
+      "learning_rate": 1.286495467665517e-05,
+      "loss": 0.7204,
+      "step": 5559
+    },
+    {
+      "epoch": 0.8508033664881408,
+      "grad_norm": 2.282158755636477,
+      "learning_rate": 1.2862580287129754e-05,
+      "loss": 0.709,
+      "step": 5560
+    },
+    {
+      "epoch": 0.850956388676358,
+      "grad_norm": 2.253648961989343,
+      "learning_rate": 1.2860205721802083e-05,
+      "loss": 0.6625,
+      "step": 5561
+    },
+    {
+      "epoch": 0.8511094108645754,
+      "grad_norm": 2.224500447042734,
+      "learning_rate": 1.2857830980817994e-05,
+      "loss": 0.6685,
+      "step": 5562
+    },
+    {
+      "epoch": 0.8512624330527927,
+      "grad_norm": 2.25169449960483,
+      "learning_rate": 1.2855456064323323e-05,
+      "loss": 0.6976,
+      "step": 5563
+    },
+    {
+      "epoch": 0.8514154552410099,
+      "grad_norm": 2.254082007689388,
+      "learning_rate": 1.2853080972463923e-05,
+      "loss": 0.7009,
+      "step": 5564
+    },
+    {
+      "epoch": 0.8515684774292273,
+      "grad_norm": 2.111325957451995,
+      "learning_rate": 1.285070570538566e-05,
+      "loss": 0.5968,
+      "step": 5565
+    },
+    {
+      "epoch": 0.8517214996174445,
+      "grad_norm": 2.212773489566472,
+      "learning_rate": 1.2848330263234408e-05,
+      "loss": 0.6656,
+      "step": 5566
+    },
+    {
+      "epoch": 0.8518745218056618,
+      "grad_norm": 2.354877615111615,
+      "learning_rate": 1.2845954646156051e-05,
+      "loss": 0.7605,
+      "step": 5567
+    },
+    {
+      "epoch": 0.8520275439938791,
+      "grad_norm": 2.166128846465111,
+      "learning_rate": 1.2843578854296485e-05,
+      "loss": 0.7061,
+      "step": 5568
+    },
+    {
+      "epoch": 0.8521805661820964,
+      "grad_norm": 2.3172994145565164,
+      "learning_rate": 1.2841202887801618e-05,
+      "loss": 0.7018,
+      "step": 5569
+    },
+    {
+      "epoch": 0.8523335883703137,
+      "grad_norm": 2.4927482090533193,
+      "learning_rate": 1.283882674681737e-05,
+      "loss": 0.7816,
+      "step": 5570
+    },
+    {
+      "epoch": 0.852486610558531,
+      "grad_norm": 2.327043385481717,
+      "learning_rate": 1.2836450431489662e-05,
+      "loss": 0.7731,
+      "step": 5571
+    },
+    {
+      "epoch": 0.8526396327467483,
+      "grad_norm": 2.309999254781278,
+      "learning_rate": 1.2834073941964438e-05,
+      "loss": 0.756,
+      "step": 5572
+    },
+    {
+      "epoch": 0.8527926549349656,
+      "grad_norm": 2.4280514724840363,
+      "learning_rate": 1.2831697278387647e-05,
+      "loss": 0.7067,
+      "step": 5573
+    },
+    {
+      "epoch": 0.8529456771231828,
+      "grad_norm": 2.0447441695223914,
+      "learning_rate": 1.2829320440905244e-05,
+      "loss": 0.6752,
+      "step": 5574
+    },
+    {
+      "epoch": 0.8530986993114001,
+      "grad_norm": 2.0921672371070223,
+      "learning_rate": 1.2826943429663212e-05,
+      "loss": 0.5819,
+      "step": 5575
+    },
+    {
+      "epoch": 0.8532517214996175,
+      "grad_norm": 2.0878343760248512,
+      "learning_rate": 1.2824566244807522e-05,
+      "loss": 0.7016,
+      "step": 5576
+    },
+    {
+      "epoch": 0.8534047436878347,
+      "grad_norm": 2.306674208187633,
+      "learning_rate": 1.2822188886484165e-05,
+      "loss": 0.754,
+      "step": 5577
+    },
+    {
+      "epoch": 0.853557765876052,
+      "grad_norm": 2.4509282856897,
+      "learning_rate": 1.2819811354839151e-05,
+      "loss": 0.7932,
+      "step": 5578
+    },
+    {
+      "epoch": 0.8537107880642694,
+      "grad_norm": 2.20326195730903,
+      "learning_rate": 1.2817433650018493e-05,
+      "loss": 0.6907,
+      "step": 5579
+    },
+    {
+      "epoch": 0.8538638102524866,
+      "grad_norm": 2.267707095823843,
+      "learning_rate": 1.2815055772168208e-05,
+      "loss": 0.7413,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8540168324407039,
+      "grad_norm": 2.246442637791065,
+      "learning_rate": 1.2812677721434338e-05,
+      "loss": 0.6501,
+      "step": 5581
+    },
+    {
+      "epoch": 0.8541698546289211,
+      "grad_norm": 2.344967709598487,
+      "learning_rate": 1.2810299497962925e-05,
+      "loss": 0.7843,
+      "step": 5582
+    },
+    {
+      "epoch": 0.8543228768171385,
+      "grad_norm": 2.4466452469658604,
+      "learning_rate": 1.280792110190003e-05,
+      "loss": 0.6965,
+      "step": 5583
+    },
+    {
+      "epoch": 0.8544758990053558,
+      "grad_norm": 2.405752794561989,
+      "learning_rate": 1.280554253339171e-05,
+      "loss": 0.7889,
+      "step": 5584
+    },
+    {
+      "epoch": 0.854628921193573,
+      "grad_norm": 2.349194728354978,
+      "learning_rate": 1.280316379258405e-05,
+      "loss": 0.6264,
+      "step": 5585
+    },
+    {
+      "epoch": 0.8547819433817904,
+      "grad_norm": 2.3163214380738117,
+      "learning_rate": 1.2800784879623135e-05,
+      "loss": 0.7097,
+      "step": 5586
+    },
+    {
+      "epoch": 0.8549349655700077,
+      "grad_norm": 2.205955621353382,
+      "learning_rate": 1.2798405794655064e-05,
+      "loss": 0.7304,
+      "step": 5587
+    },
+    {
+      "epoch": 0.8550879877582249,
+      "grad_norm": 2.2160837810993526,
+      "learning_rate": 1.2796026537825947e-05,
+      "loss": 0.5901,
+      "step": 5588
+    },
+    {
+      "epoch": 0.8552410099464423,
+      "grad_norm": 2.076799915175384,
+      "learning_rate": 1.2793647109281901e-05,
+      "loss": 0.6659,
+      "step": 5589
+    },
+    {
+      "epoch": 0.8553940321346595,
+      "grad_norm": 2.169267298600903,
+      "learning_rate": 1.2791267509169054e-05,
+      "loss": 0.707,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8555470543228768,
+      "grad_norm": 2.0806303549381364,
+      "learning_rate": 1.2788887737633555e-05,
+      "loss": 0.648,
+      "step": 5591
+    },
+    {
+      "epoch": 0.8557000765110941,
+      "grad_norm": 2.1414594681949635,
+      "learning_rate": 1.278650779482155e-05,
+      "loss": 0.6572,
+      "step": 5592
+    },
+    {
+      "epoch": 0.8558530986993114,
+      "grad_norm": 2.3353430053689435,
+      "learning_rate": 1.2784127680879195e-05,
+      "loss": 0.6605,
+      "step": 5593
+    },
+    {
+      "epoch": 0.8560061208875287,
+      "grad_norm": 2.4855997648787653,
+      "learning_rate": 1.278174739595267e-05,
+      "loss": 0.6597,
+      "step": 5594
+    },
+    {
+      "epoch": 0.856159143075746,
+      "grad_norm": 2.139860057474892,
+      "learning_rate": 1.2779366940188159e-05,
+      "loss": 0.6893,
+      "step": 5595
+    },
+    {
+      "epoch": 0.8563121652639633,
+      "grad_norm": 2.229662131253198,
+      "learning_rate": 1.2776986313731847e-05,
+      "loss": 0.6621,
+      "step": 5596
+    },
+    {
+      "epoch": 0.8564651874521806,
+      "grad_norm": 2.482093964955999,
+      "learning_rate": 1.2774605516729944e-05,
+      "loss": 0.8149,
+      "step": 5597
+    },
+    {
+      "epoch": 0.8566182096403978,
+      "grad_norm": 2.3260474144446115,
+      "learning_rate": 1.277222454932866e-05,
+      "loss": 0.6988,
+      "step": 5598
+    },
+    {
+      "epoch": 0.8567712318286151,
+      "grad_norm": 2.238022446025169,
+      "learning_rate": 1.2769843411674221e-05,
+      "loss": 0.7234,
+      "step": 5599
+    },
+    {
+      "epoch": 0.8569242540168325,
+      "grad_norm": 1.9938349722793565,
+      "learning_rate": 1.2767462103912864e-05,
+      "loss": 0.597,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8570772762050497,
+      "grad_norm": 2.4546219970379726,
+      "learning_rate": 1.2765080626190834e-05,
+      "loss": 0.7452,
+      "step": 5601
+    },
+    {
+      "epoch": 0.857230298393267,
+      "grad_norm": 2.3423061899227244,
+      "learning_rate": 1.2762698978654381e-05,
+      "loss": 0.7442,
+      "step": 5602
+    },
+    {
+      "epoch": 0.8573833205814844,
+      "grad_norm": 2.5320781222404176,
+      "learning_rate": 1.276031716144978e-05,
+      "loss": 0.772,
+      "step": 5603
+    },
+    {
+      "epoch": 0.8575363427697016,
+      "grad_norm": 2.3118002124605623,
+      "learning_rate": 1.2757935174723306e-05,
+      "loss": 0.6711,
+      "step": 5604
+    },
+    {
+      "epoch": 0.8576893649579189,
+      "grad_norm": 2.400968927044133,
+      "learning_rate": 1.2755553018621238e-05,
+      "loss": 0.6895,
+      "step": 5605
+    },
+    {
+      "epoch": 0.8578423871461361,
+      "grad_norm": 1.9936304706669457,
+      "learning_rate": 1.2753170693289882e-05,
+      "loss": 0.5834,
+      "step": 5606
+    },
+    {
+      "epoch": 0.8579954093343535,
+      "grad_norm": 2.5147260168293384,
+      "learning_rate": 1.2750788198875545e-05,
+      "loss": 0.6648,
+      "step": 5607
+    },
+    {
+      "epoch": 0.8581484315225708,
+      "grad_norm": 2.202869381008538,
+      "learning_rate": 1.2748405535524541e-05,
+      "loss": 0.6608,
+      "step": 5608
+    },
+    {
+      "epoch": 0.858301453710788,
+      "grad_norm": 2.2267686720968394,
+      "learning_rate": 1.2746022703383203e-05,
+      "loss": 0.7458,
+      "step": 5609
+    },
+    {
+      "epoch": 0.8584544758990054,
+      "grad_norm": 2.1783234217642153,
+      "learning_rate": 1.2743639702597868e-05,
+      "loss": 0.7113,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8586074980872227,
+      "grad_norm": 2.2367728821011785,
+      "learning_rate": 1.2741256533314885e-05,
+      "loss": 0.6847,
+      "step": 5611
+    },
+    {
+      "epoch": 0.8587605202754399,
+      "grad_norm": 2.4930263847632457,
+      "learning_rate": 1.2738873195680615e-05,
+      "loss": 0.7769,
+      "step": 5612
+    },
+    {
+      "epoch": 0.8589135424636573,
+      "grad_norm": 2.3443625908959858,
+      "learning_rate": 1.2736489689841427e-05,
+      "loss": 0.7359,
+      "step": 5613
+    },
+    {
+      "epoch": 0.8590665646518745,
+      "grad_norm": 2.305054821326658,
+      "learning_rate": 1.2734106015943702e-05,
+      "loss": 0.6783,
+      "step": 5614
+    },
+    {
+      "epoch": 0.8592195868400918,
+      "grad_norm": 1.9663379003968546,
+      "learning_rate": 1.2731722174133827e-05,
+      "loss": 0.6796,
+      "step": 5615
+    },
+    {
+      "epoch": 0.8593726090283091,
+      "grad_norm": 2.092756435546732,
+      "learning_rate": 1.2729338164558214e-05,
+      "loss": 0.713,
+      "step": 5616
+    },
+    {
+      "epoch": 0.8595256312165264,
+      "grad_norm": 2.24491708283434,
+      "learning_rate": 1.2726953987363264e-05,
+      "loss": 0.6345,
+      "step": 5617
+    },
+    {
+      "epoch": 0.8596786534047437,
+      "grad_norm": 2.1594981539794276,
+      "learning_rate": 1.2724569642695399e-05,
+      "loss": 0.6694,
+      "step": 5618
+    },
+    {
+      "epoch": 0.859831675592961,
+      "grad_norm": 2.2943585124865384,
+      "learning_rate": 1.2722185130701057e-05,
+      "loss": 0.7367,
+      "step": 5619
+    },
+    {
+      "epoch": 0.8599846977811783,
+      "grad_norm": 2.3976683676703767,
+      "learning_rate": 1.2719800451526676e-05,
+      "loss": 0.7081,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8601377199693956,
+      "grad_norm": 2.2946408098661064,
+      "learning_rate": 1.2717415605318708e-05,
+      "loss": 0.7871,
+      "step": 5621
+    },
+    {
+      "epoch": 0.8602907421576128,
+      "grad_norm": 2.3980471919531876,
+      "learning_rate": 1.2715030592223619e-05,
+      "loss": 0.758,
+      "step": 5622
+    },
+    {
+      "epoch": 0.8604437643458301,
+      "grad_norm": 2.2339319776161037,
+      "learning_rate": 1.2712645412387878e-05,
+      "loss": 0.6481,
+      "step": 5623
+    },
+    {
+      "epoch": 0.8605967865340475,
+      "grad_norm": 2.370078313360965,
+      "learning_rate": 1.2710260065957973e-05,
+      "loss": 0.8031,
+      "step": 5624
+    },
+    {
+      "epoch": 0.8607498087222647,
+      "grad_norm": 2.361702795265424,
+      "learning_rate": 1.2707874553080393e-05,
+      "loss": 0.6998,
+      "step": 5625
+    },
+    {
+      "epoch": 0.860902830910482,
+      "grad_norm": 2.185328558542397,
+      "learning_rate": 1.2705488873901643e-05,
+      "loss": 0.6882,
+      "step": 5626
+    },
+    {
+      "epoch": 0.8610558530986994,
+      "grad_norm": 2.432654884382053,
+      "learning_rate": 1.270310302856824e-05,
+      "loss": 0.662,
+      "step": 5627
+    },
+    {
+      "epoch": 0.8612088752869166,
+      "grad_norm": 2.4952824090910624,
+      "learning_rate": 1.27007170172267e-05,
+      "loss": 0.8529,
+      "step": 5628
+    },
+    {
+      "epoch": 0.8613618974751339,
+      "grad_norm": 2.236267328529113,
+      "learning_rate": 1.2698330840023569e-05,
+      "loss": 0.7145,
+      "step": 5629
+    },
+    {
+      "epoch": 0.8615149196633511,
+      "grad_norm": 2.2009751715707746,
+      "learning_rate": 1.2695944497105383e-05,
+      "loss": 0.7445,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8616679418515685,
+      "grad_norm": 2.671644999870953,
+      "learning_rate": 1.2693557988618696e-05,
+      "loss": 0.6987,
+      "step": 5631
+    },
+    {
+      "epoch": 0.8618209640397858,
+      "grad_norm": 2.1200922727768194,
+      "learning_rate": 1.2691171314710078e-05,
+      "loss": 0.6588,
+      "step": 5632
+    },
+    {
+      "epoch": 0.861973986228003,
+      "grad_norm": 2.20322726646221,
+      "learning_rate": 1.2688784475526103e-05,
+      "loss": 0.5955,
+      "step": 5633
+    },
+    {
+      "epoch": 0.8621270084162204,
+      "grad_norm": 2.0969035364933095,
+      "learning_rate": 1.2686397471213352e-05,
+      "loss": 0.6224,
+      "step": 5634
+    },
+    {
+      "epoch": 0.8622800306044377,
+      "grad_norm": 2.128200156708807,
+      "learning_rate": 1.2684010301918424e-05,
+      "loss": 0.6266,
+      "step": 5635
+    },
+    {
+      "epoch": 0.8624330527926549,
+      "grad_norm": 2.1941487004202114,
+      "learning_rate": 1.2681622967787925e-05,
+      "loss": 0.6647,
+      "step": 5636
+    },
+    {
+      "epoch": 0.8625860749808723,
+      "grad_norm": 1.9520577900142773,
+      "learning_rate": 1.2679235468968468e-05,
+      "loss": 0.6282,
+      "step": 5637
+    },
+    {
+      "epoch": 0.8627390971690895,
+      "grad_norm": 2.36577545218808,
+      "learning_rate": 1.267684780560668e-05,
+      "loss": 0.6948,
+      "step": 5638
+    },
+    {
+      "epoch": 0.8628921193573068,
+      "grad_norm": 2.030714318416365,
+      "learning_rate": 1.2674459977849197e-05,
+      "loss": 0.6446,
+      "step": 5639
+    },
+    {
+      "epoch": 0.8630451415455241,
+      "grad_norm": 2.3029069648527463,
+      "learning_rate": 1.267207198584266e-05,
+      "loss": 0.7449,
+      "step": 5640
+    },
+    {
+      "epoch": 0.8631981637337414,
+      "grad_norm": 2.240610425205807,
+      "learning_rate": 1.2669683829733734e-05,
+      "loss": 0.6766,
+      "step": 5641
+    },
+    {
+      "epoch": 0.8633511859219587,
+      "grad_norm": 2.3727769689851743,
+      "learning_rate": 1.2667295509669077e-05,
+      "loss": 0.681,
+      "step": 5642
+    },
+    {
+      "epoch": 0.863504208110176,
+      "grad_norm": 2.2735964610799932,
+      "learning_rate": 1.2664907025795369e-05,
+      "loss": 0.689,
+      "step": 5643
+    },
+    {
+      "epoch": 0.8636572302983933,
+      "grad_norm": 2.3451657930943632,
+      "learning_rate": 1.2662518378259297e-05,
+      "loss": 0.7561,
+      "step": 5644
+    },
+    {
+      "epoch": 0.8638102524866106,
+      "grad_norm": 2.3808378476510583,
+      "learning_rate": 1.2660129567207555e-05,
+      "loss": 0.7952,
+      "step": 5645
+    },
+    {
+      "epoch": 0.8639632746748278,
+      "grad_norm": 2.264345473222062,
+      "learning_rate": 1.2657740592786847e-05,
+      "loss": 0.7002,
+      "step": 5646
+    },
+    {
+      "epoch": 0.8641162968630451,
+      "grad_norm": 2.079793665356641,
+      "learning_rate": 1.2655351455143896e-05,
+      "loss": 0.6575,
+      "step": 5647
+    },
+    {
+      "epoch": 0.8642693190512625,
+      "grad_norm": 2.4275830013363175,
+      "learning_rate": 1.2652962154425423e-05,
+      "loss": 0.7579,
+      "step": 5648
+    },
+    {
+      "epoch": 0.8644223412394797,
+      "grad_norm": 2.2748583552247994,
+      "learning_rate": 1.2650572690778164e-05,
+      "loss": 0.6597,
+      "step": 5649
+    },
+    {
+      "epoch": 0.864575363427697,
+      "grad_norm": 2.1601232807271673,
+      "learning_rate": 1.2648183064348868e-05,
+      "loss": 0.5999,
+      "step": 5650
+    },
+    {
+      "epoch": 0.8647283856159143,
+      "grad_norm": 2.2760902133080467,
+      "learning_rate": 1.2645793275284293e-05,
+      "loss": 0.6744,
+      "step": 5651
+    },
+    {
+      "epoch": 0.8648814078041316,
+      "grad_norm": 2.5328811384211405,
+      "learning_rate": 1.2643403323731198e-05,
+      "loss": 0.7486,
+      "step": 5652
+    },
+    {
+      "epoch": 0.8650344299923489,
+      "grad_norm": 1.9191237464359976,
+      "learning_rate": 1.2641013209836369e-05,
+      "loss": 0.5882,
+      "step": 5653
+    },
+    {
+      "epoch": 0.8651874521805661,
+      "grad_norm": 2.1442818040053564,
+      "learning_rate": 1.2638622933746583e-05,
+      "loss": 0.6574,
+      "step": 5654
+    },
+    {
+      "epoch": 0.8653404743687835,
+      "grad_norm": 2.313223313915067,
+      "learning_rate": 1.263623249560864e-05,
+      "loss": 0.705,
+      "step": 5655
+    },
+    {
+      "epoch": 0.8654934965570008,
+      "grad_norm": 2.1582983437738363,
+      "learning_rate": 1.2633841895569346e-05,
+      "loss": 0.6208,
+      "step": 5656
+    },
+    {
+      "epoch": 0.865646518745218,
+      "grad_norm": 2.140699620669154,
+      "learning_rate": 1.263145113377552e-05,
+      "loss": 0.6216,
+      "step": 5657
+    },
+    {
+      "epoch": 0.8657995409334354,
+      "grad_norm": 2.51001916251359,
+      "learning_rate": 1.2629060210373986e-05,
+      "loss": 0.8365,
+      "step": 5658
+    },
+    {
+      "epoch": 0.8659525631216526,
+      "grad_norm": 2.1467206121687714,
+      "learning_rate": 1.2626669125511578e-05,
+      "loss": 0.6975,
+      "step": 5659
+    },
+    {
+      "epoch": 0.8661055853098699,
+      "grad_norm": 2.1179526840995955,
+      "learning_rate": 1.2624277879335142e-05,
+      "loss": 0.6771,
+      "step": 5660
+    },
+    {
+      "epoch": 0.8662586074980873,
+      "grad_norm": 2.242352909508502,
+      "learning_rate": 1.2621886471991538e-05,
+      "loss": 0.7168,
+      "step": 5661
+    },
+    {
+      "epoch": 0.8664116296863045,
+      "grad_norm": 2.4601217594406193,
+      "learning_rate": 1.261949490362763e-05,
+      "loss": 0.6644,
+      "step": 5662
+    },
+    {
+      "epoch": 0.8665646518745218,
+      "grad_norm": 2.1506121299772327,
+      "learning_rate": 1.2617103174390289e-05,
+      "loss": 0.6518,
+      "step": 5663
+    },
+    {
+      "epoch": 0.8667176740627391,
+      "grad_norm": 2.535854595188026,
+      "learning_rate": 1.2614711284426406e-05,
+      "loss": 0.7968,
+      "step": 5664
+    },
+    {
+      "epoch": 0.8668706962509564,
+      "grad_norm": 2.3806302754076936,
+      "learning_rate": 1.2612319233882873e-05,
+      "loss": 0.6961,
+      "step": 5665
+    },
+    {
+      "epoch": 0.8670237184391737,
+      "grad_norm": 2.0825689145311705,
+      "learning_rate": 1.2609927022906598e-05,
+      "loss": 0.4917,
+      "step": 5666
+    },
+    {
+      "epoch": 0.8671767406273909,
+      "grad_norm": 2.296561591485694,
+      "learning_rate": 1.2607534651644493e-05,
+      "loss": 0.6687,
+      "step": 5667
+    },
+    {
+      "epoch": 0.8673297628156083,
+      "grad_norm": 2.0016062923376703,
+      "learning_rate": 1.2605142120243485e-05,
+      "loss": 0.5966,
+      "step": 5668
+    },
+    {
+      "epoch": 0.8674827850038256,
+      "grad_norm": 2.2328943779206676,
+      "learning_rate": 1.2602749428850506e-05,
+      "loss": 0.6952,
+      "step": 5669
+    },
+    {
+      "epoch": 0.8676358071920428,
+      "grad_norm": 2.1730370472915164,
+      "learning_rate": 1.2600356577612507e-05,
+      "loss": 0.6091,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8677888293802601,
+      "grad_norm": 2.30791377621154,
+      "learning_rate": 1.2597963566676435e-05,
+      "loss": 0.7486,
+      "step": 5671
+    },
+    {
+      "epoch": 0.8679418515684775,
+      "grad_norm": 2.276237653997365,
+      "learning_rate": 1.2595570396189254e-05,
+      "loss": 0.7926,
+      "step": 5672
+    },
+    {
+      "epoch": 0.8680948737566947,
+      "grad_norm": 2.3754804501301514,
+      "learning_rate": 1.2593177066297945e-05,
+      "loss": 0.6826,
+      "step": 5673
+    },
+    {
+      "epoch": 0.868247895944912,
+      "grad_norm": 2.2435259533338314,
+      "learning_rate": 1.2590783577149488e-05,
+      "loss": 0.6384,
+      "step": 5674
+    },
+    {
+      "epoch": 0.8684009181331293,
+      "grad_norm": 2.087817049910248,
+      "learning_rate": 1.2588389928890873e-05,
+      "loss": 0.6228,
+      "step": 5675
+    },
+    {
+      "epoch": 0.8685539403213466,
+      "grad_norm": 2.3075610370217516,
+      "learning_rate": 1.258599612166911e-05,
+      "loss": 0.6741,
+      "step": 5676
+    },
+    {
+      "epoch": 0.8687069625095639,
+      "grad_norm": 2.105883253531245,
+      "learning_rate": 1.258360215563121e-05,
+      "loss": 0.6242,
+      "step": 5677
+    },
+    {
+      "epoch": 0.8688599846977811,
+      "grad_norm": 2.2765438530308564,
+      "learning_rate": 1.258120803092419e-05,
+      "loss": 0.7679,
+      "step": 5678
+    },
+    {
+      "epoch": 0.8690130068859985,
+      "grad_norm": 2.4505312365492844,
+      "learning_rate": 1.2578813747695091e-05,
+      "loss": 0.6735,
+      "step": 5679
+    },
+    {
+      "epoch": 0.8691660290742158,
+      "grad_norm": 2.3676073507608604,
+      "learning_rate": 1.2576419306090948e-05,
+      "loss": 0.6729,
+      "step": 5680
+    },
+    {
+      "epoch": 0.869319051262433,
+      "grad_norm": 2.44965356160525,
+      "learning_rate": 1.2574024706258817e-05,
+      "loss": 0.7476,
+      "step": 5681
+    },
+    {
+      "epoch": 0.8694720734506504,
+      "grad_norm": 2.1538693697394486,
+      "learning_rate": 1.2571629948345763e-05,
+      "loss": 0.6866,
+      "step": 5682
+    },
+    {
+      "epoch": 0.8696250956388676,
+      "grad_norm": 2.0656645571127887,
+      "learning_rate": 1.256923503249885e-05,
+      "loss": 0.7174,
+      "step": 5683
+    },
+    {
+      "epoch": 0.8697781178270849,
+      "grad_norm": 2.21545987856496,
+      "learning_rate": 1.2566839958865162e-05,
+      "loss": 0.6976,
+      "step": 5684
+    },
+    {
+      "epoch": 0.8699311400153023,
+      "grad_norm": 2.2435429627080268,
+      "learning_rate": 1.2564444727591791e-05,
+      "loss": 0.7416,
+      "step": 5685
+    },
+    {
+      "epoch": 0.8700841622035195,
+      "grad_norm": 2.288122679843375,
+      "learning_rate": 1.2562049338825842e-05,
+      "loss": 0.6695,
+      "step": 5686
+    },
+    {
+      "epoch": 0.8702371843917368,
+      "grad_norm": 2.4078595060538452,
+      "learning_rate": 1.2559653792714414e-05,
+      "loss": 0.7819,
+      "step": 5687
+    },
+    {
+      "epoch": 0.8703902065799541,
+      "grad_norm": 2.4138389650474097,
+      "learning_rate": 1.2557258089404635e-05,
+      "loss": 0.6232,
+      "step": 5688
+    },
+    {
+      "epoch": 0.8705432287681714,
+      "grad_norm": 2.1081792762846945,
+      "learning_rate": 1.2554862229043633e-05,
+      "loss": 0.6831,
+      "step": 5689
+    },
+    {
+      "epoch": 0.8706962509563887,
+      "grad_norm": 2.0366851001301596,
+      "learning_rate": 1.255246621177855e-05,
+      "loss": 0.7232,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8708492731446059,
+      "grad_norm": 2.0327500614427536,
+      "learning_rate": 1.2550070037756527e-05,
+      "loss": 0.6628,
+      "step": 5691
+    },
+    {
+      "epoch": 0.8710022953328233,
+      "grad_norm": 2.361742724788852,
+      "learning_rate": 1.254767370712473e-05,
+      "loss": 0.8272,
+      "step": 5692
+    },
+    {
+      "epoch": 0.8711553175210406,
+      "grad_norm": 2.18870554680478,
+      "learning_rate": 1.2545277220030324e-05,
+      "loss": 0.6787,
+      "step": 5693
+    },
+    {
+      "epoch": 0.8713083397092578,
+      "grad_norm": 2.3066595929772613,
+      "learning_rate": 1.2542880576620484e-05,
+      "loss": 0.7596,
+      "step": 5694
+    },
+    {
+      "epoch": 0.8714613618974751,
+      "grad_norm": 2.3582637357450023,
+      "learning_rate": 1.2540483777042403e-05,
+      "loss": 0.6326,
+      "step": 5695
+    },
+    {
+      "epoch": 0.8716143840856925,
+      "grad_norm": 2.1035213548968312,
+      "learning_rate": 1.2538086821443273e-05,
+      "loss": 0.6634,
+      "step": 5696
+    },
+    {
+      "epoch": 0.8717674062739097,
+      "grad_norm": 2.2947630934368837,
+      "learning_rate": 1.25356897099703e-05,
+      "loss": 0.7888,
+      "step": 5697
+    },
+    {
+      "epoch": 0.871920428462127,
+      "grad_norm": 2.3707709917608204,
+      "learning_rate": 1.2533292442770705e-05,
+      "loss": 0.78,
+      "step": 5698
+    },
+    {
+      "epoch": 0.8720734506503443,
+      "grad_norm": 2.1437011373550003,
+      "learning_rate": 1.2530895019991714e-05,
+      "loss": 0.7513,
+      "step": 5699
+    },
+    {
+      "epoch": 0.8722264728385616,
+      "grad_norm": 2.132187427093857,
+      "learning_rate": 1.2528497441780554e-05,
+      "loss": 0.6085,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8723794950267789,
+      "grad_norm": 2.25455470161129,
+      "learning_rate": 1.2526099708284476e-05,
+      "loss": 0.7394,
+      "step": 5701
+    },
+    {
+      "epoch": 0.8725325172149961,
+      "grad_norm": 2.5745994986985106,
+      "learning_rate": 1.2523701819650733e-05,
+      "loss": 0.7067,
+      "step": 5702
+    },
+    {
+      "epoch": 0.8726855394032135,
+      "grad_norm": 2.139572994127622,
+      "learning_rate": 1.252130377602659e-05,
+      "loss": 0.6948,
+      "step": 5703
+    },
+    {
+      "epoch": 0.8728385615914308,
+      "grad_norm": 2.336370835834088,
+      "learning_rate": 1.2518905577559317e-05,
+      "loss": 0.6016,
+      "step": 5704
+    },
+    {
+      "epoch": 0.872991583779648,
+      "grad_norm": 2.212567829784752,
+      "learning_rate": 1.2516507224396199e-05,
+      "loss": 0.7168,
+      "step": 5705
+    },
+    {
+      "epoch": 0.8731446059678654,
+      "grad_norm": 2.1173283124728144,
+      "learning_rate": 1.2514108716684527e-05,
+      "loss": 0.7108,
+      "step": 5706
+    },
+    {
+      "epoch": 0.8732976281560826,
+      "grad_norm": 2.286564364188799,
+      "learning_rate": 1.2511710054571603e-05,
+      "loss": 0.6567,
+      "step": 5707
+    },
+    {
+      "epoch": 0.8734506503442999,
+      "grad_norm": 2.4319669016285155,
+      "learning_rate": 1.2509311238204742e-05,
+      "loss": 0.7369,
+      "step": 5708
+    },
+    {
+      "epoch": 0.8736036725325173,
+      "grad_norm": 2.3105859541550804,
+      "learning_rate": 1.250691226773126e-05,
+      "loss": 0.6155,
+      "step": 5709
+    },
+    {
+      "epoch": 0.8737566947207345,
+      "grad_norm": 2.299770180808811,
+      "learning_rate": 1.2504513143298485e-05,
+      "loss": 0.7094,
+      "step": 5710
+    },
+    {
+      "epoch": 0.8739097169089518,
+      "grad_norm": 2.0337893937536595,
+      "learning_rate": 1.2502113865053764e-05,
+      "loss": 0.6748,
+      "step": 5711
+    },
+    {
+      "epoch": 0.8740627390971691,
+      "grad_norm": 2.385295588774363,
+      "learning_rate": 1.2499714433144441e-05,
+      "loss": 0.7138,
+      "step": 5712
+    },
+    {
+      "epoch": 0.8742157612853864,
+      "grad_norm": 2.284489191336518,
+      "learning_rate": 1.2497314847717875e-05,
+      "loss": 0.743,
+      "step": 5713
+    },
+    {
+      "epoch": 0.8743687834736037,
+      "grad_norm": 2.3990280322523367,
+      "learning_rate": 1.2494915108921434e-05,
+      "loss": 0.6998,
+      "step": 5714
+    },
+    {
+      "epoch": 0.8745218056618209,
+      "grad_norm": 2.2891843264651905,
+      "learning_rate": 1.2492515216902499e-05,
+      "loss": 0.6992,
+      "step": 5715
+    },
+    {
+      "epoch": 0.8746748278500383,
+      "grad_norm": 2.3026551519865683,
+      "learning_rate": 1.249011517180845e-05,
+      "loss": 0.7515,
+      "step": 5716
+    },
+    {
+      "epoch": 0.8748278500382556,
+      "grad_norm": 2.337520385134906,
+      "learning_rate": 1.248771497378669e-05,
+      "loss": 0.7088,
+      "step": 5717
+    },
+    {
+      "epoch": 0.8749808722264728,
+      "grad_norm": 2.1926541321569006,
+      "learning_rate": 1.248531462298462e-05,
+      "loss": 0.7873,
+      "step": 5718
+    },
+    {
+      "epoch": 0.8751338944146901,
+      "grad_norm": 2.4455148838618244,
+      "learning_rate": 1.2482914119549658e-05,
+      "loss": 0.6451,
+      "step": 5719
+    },
+    {
+      "epoch": 0.8752869166029075,
+      "grad_norm": 2.5339680205683743,
+      "learning_rate": 1.2480513463629224e-05,
+      "loss": 0.7793,
+      "step": 5720
+    },
+    {
+      "epoch": 0.8754399387911247,
+      "grad_norm": 2.3813626401774775,
+      "learning_rate": 1.2478112655370758e-05,
+      "loss": 0.5941,
+      "step": 5721
+    },
+    {
+      "epoch": 0.875592960979342,
+      "grad_norm": 2.4186647306912046,
+      "learning_rate": 1.2475711694921695e-05,
+      "loss": 0.723,
+      "step": 5722
+    },
+    {
+      "epoch": 0.8757459831675592,
+      "grad_norm": 2.033642018028739,
+      "learning_rate": 1.2473310582429496e-05,
+      "loss": 0.6569,
+      "step": 5723
+    },
+    {
+      "epoch": 0.8758990053557766,
+      "grad_norm": 2.050104551897003,
+      "learning_rate": 1.2470909318041618e-05,
+      "loss": 0.7118,
+      "step": 5724
+    },
+    {
+      "epoch": 0.8760520275439939,
+      "grad_norm": 2.098302900404374,
+      "learning_rate": 1.2468507901905527e-05,
+      "loss": 0.5665,
+      "step": 5725
+    },
+    {
+      "epoch": 0.8762050497322111,
+      "grad_norm": 2.238378059736463,
+      "learning_rate": 1.2466106334168713e-05,
+      "loss": 0.6099,
+      "step": 5726
+    },
+    {
+      "epoch": 0.8763580719204285,
+      "grad_norm": 2.264652562759667,
+      "learning_rate": 1.2463704614978664e-05,
+      "loss": 0.7528,
+      "step": 5727
+    },
+    {
+      "epoch": 0.8765110941086458,
+      "grad_norm": 2.315603460621223,
+      "learning_rate": 1.2461302744482873e-05,
+      "loss": 0.7416,
+      "step": 5728
+    },
+    {
+      "epoch": 0.876664116296863,
+      "grad_norm": 1.9928379922302875,
+      "learning_rate": 1.2458900722828852e-05,
+      "loss": 0.6208,
+      "step": 5729
+    },
+    {
+      "epoch": 0.8768171384850804,
+      "grad_norm": 2.196223804754041,
+      "learning_rate": 1.245649855016412e-05,
+      "loss": 0.789,
+      "step": 5730
+    },
+    {
+      "epoch": 0.8769701606732976,
+      "grad_norm": 1.9968093015394444,
+      "learning_rate": 1.24540962266362e-05,
+      "loss": 0.6108,
+      "step": 5731
+    },
+    {
+      "epoch": 0.8771231828615149,
+      "grad_norm": 2.1538731630302754,
+      "learning_rate": 1.2451693752392634e-05,
+      "loss": 0.6491,
+      "step": 5732
+    },
+    {
+      "epoch": 0.8772762050497322,
+      "grad_norm": 2.336782860051827,
+      "learning_rate": 1.244929112758096e-05,
+      "loss": 0.6887,
+      "step": 5733
+    },
+    {
+      "epoch": 0.8774292272379495,
+      "grad_norm": 2.0968379693636265,
+      "learning_rate": 1.2446888352348738e-05,
+      "loss": 0.6052,
+      "step": 5734
+    },
+    {
+      "epoch": 0.8775822494261668,
+      "grad_norm": 2.8413364803514463,
+      "learning_rate": 1.2444485426843527e-05,
+      "loss": 0.7297,
+      "step": 5735
+    },
+    {
+      "epoch": 0.8777352716143841,
+      "grad_norm": 2.414859416385035,
+      "learning_rate": 1.2442082351212908e-05,
+      "loss": 0.6594,
+      "step": 5736
+    },
+    {
+      "epoch": 0.8778882938026014,
+      "grad_norm": 2.1397332273037857,
+      "learning_rate": 1.2439679125604455e-05,
+      "loss": 0.6262,
+      "step": 5737
+    },
+    {
+      "epoch": 0.8780413159908187,
+      "grad_norm": 2.302960630598352,
+      "learning_rate": 1.2437275750165763e-05,
+      "loss": 0.6868,
+      "step": 5738
+    },
+    {
+      "epoch": 0.8781943381790359,
+      "grad_norm": 2.159627088120593,
+      "learning_rate": 1.2434872225044432e-05,
+      "loss": 0.6593,
+      "step": 5739
+    },
+    {
+      "epoch": 0.8783473603672532,
+      "grad_norm": 2.303777890381595,
+      "learning_rate": 1.2432468550388078e-05,
+      "loss": 0.7509,
+      "step": 5740
+    },
+    {
+      "epoch": 0.8785003825554706,
+      "grad_norm": 2.082765533333303,
+      "learning_rate": 1.2430064726344306e-05,
+      "loss": 0.6463,
+      "step": 5741
+    },
+    {
+      "epoch": 0.8786534047436878,
+      "grad_norm": 2.300303736175458,
+      "learning_rate": 1.2427660753060758e-05,
+      "loss": 0.7133,
+      "step": 5742
+    },
+    {
+      "epoch": 0.8788064269319051,
+      "grad_norm": 2.2364378062050254,
+      "learning_rate": 1.2425256630685063e-05,
+      "loss": 0.6881,
+      "step": 5743
+    },
+    {
+      "epoch": 0.8789594491201225,
+      "grad_norm": 2.0318322741252146,
+      "learning_rate": 1.2422852359364873e-05,
+      "loss": 0.6728,
+      "step": 5744
+    },
+    {
+      "epoch": 0.8791124713083397,
+      "grad_norm": 2.2990306189375223,
+      "learning_rate": 1.2420447939247841e-05,
+      "loss": 0.623,
+      "step": 5745
+    },
+    {
+      "epoch": 0.879265493496557,
+      "grad_norm": 2.4502465737162504,
+      "learning_rate": 1.2418043370481631e-05,
+      "loss": 0.6824,
+      "step": 5746
+    },
+    {
+      "epoch": 0.8794185156847742,
+      "grad_norm": 2.2065393096891803,
+      "learning_rate": 1.2415638653213919e-05,
+      "loss": 0.631,
+      "step": 5747
+    },
+    {
+      "epoch": 0.8795715378729916,
+      "grad_norm": 2.3645950652061565,
+      "learning_rate": 1.2413233787592388e-05,
+      "loss": 0.7174,
+      "step": 5748
+    },
+    {
+      "epoch": 0.8797245600612089,
+      "grad_norm": 2.184415402062973,
+      "learning_rate": 1.241082877376473e-05,
+      "loss": 0.6933,
+      "step": 5749
+    },
+    {
+      "epoch": 0.8798775822494261,
+      "grad_norm": 2.2344101746435574,
+      "learning_rate": 1.2408423611878643e-05,
+      "loss": 0.7379,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8800306044376435,
+      "grad_norm": 2.2362564806831107,
+      "learning_rate": 1.2406018302081841e-05,
+      "loss": 0.6887,
+      "step": 5751
+    },
+    {
+      "epoch": 0.8801836266258608,
+      "grad_norm": 2.1149926363107814,
+      "learning_rate": 1.2403612844522046e-05,
+      "loss": 0.6424,
+      "step": 5752
+    },
+    {
+      "epoch": 0.880336648814078,
+      "grad_norm": 2.310775641769557,
+      "learning_rate": 1.2401207239346982e-05,
+      "loss": 0.731,
+      "step": 5753
+    },
+    {
+      "epoch": 0.8804896710022954,
+      "grad_norm": 2.254344287576404,
+      "learning_rate": 1.2398801486704384e-05,
+      "loss": 0.7344,
+      "step": 5754
+    },
+    {
+      "epoch": 0.8806426931905126,
+      "grad_norm": 2.159264355383588,
+      "learning_rate": 1.2396395586742005e-05,
+      "loss": 0.7299,
+      "step": 5755
+    },
+    {
+      "epoch": 0.8807957153787299,
+      "grad_norm": 2.4997945440793528,
+      "learning_rate": 1.2393989539607601e-05,
+      "loss": 0.6929,
+      "step": 5756
+    },
+    {
+      "epoch": 0.8809487375669472,
+      "grad_norm": 2.019321161103983,
+      "learning_rate": 1.239158334544893e-05,
+      "loss": 0.5885,
+      "step": 5757
+    },
+    {
+      "epoch": 0.8811017597551645,
+      "grad_norm": 2.4218191192077176,
+      "learning_rate": 1.238917700441377e-05,
+      "loss": 0.7094,
+      "step": 5758
+    },
+    {
+      "epoch": 0.8812547819433818,
+      "grad_norm": 2.0924771769119808,
+      "learning_rate": 1.2386770516649904e-05,
+      "loss": 0.6709,
+      "step": 5759
+    },
+    {
+      "epoch": 0.881407804131599,
+      "grad_norm": 2.48339718936483,
+      "learning_rate": 1.2384363882305125e-05,
+      "loss": 0.7068,
+      "step": 5760
+    },
+    {
+      "epoch": 0.8815608263198164,
+      "grad_norm": 2.565722271968007,
+      "learning_rate": 1.2381957101527233e-05,
+      "loss": 0.6858,
+      "step": 5761
+    },
+    {
+      "epoch": 0.8817138485080337,
+      "grad_norm": 2.2301004887844513,
+      "learning_rate": 1.2379550174464034e-05,
+      "loss": 0.5637,
+      "step": 5762
+    },
+    {
+      "epoch": 0.8818668706962509,
+      "grad_norm": 2.1331296366939356,
+      "learning_rate": 1.2377143101263351e-05,
+      "loss": 0.6533,
+      "step": 5763
+    },
+    {
+      "epoch": 0.8820198928844682,
+      "grad_norm": 2.0132594420000984,
+      "learning_rate": 1.2374735882073014e-05,
+      "loss": 0.6407,
+      "step": 5764
+    },
+    {
+      "epoch": 0.8821729150726856,
+      "grad_norm": 2.11184387213977,
+      "learning_rate": 1.2372328517040854e-05,
+      "loss": 0.7467,
+      "step": 5765
+    },
+    {
+      "epoch": 0.8823259372609028,
+      "grad_norm": 2.1063721445520156,
+      "learning_rate": 1.2369921006314716e-05,
+      "loss": 0.6223,
+      "step": 5766
+    },
+    {
+      "epoch": 0.8824789594491201,
+      "grad_norm": 2.135649260419935,
+      "learning_rate": 1.2367513350042461e-05,
+      "loss": 0.7001,
+      "step": 5767
+    },
+    {
+      "epoch": 0.8826319816373374,
+      "grad_norm": 2.4573013608444167,
+      "learning_rate": 1.2365105548371949e-05,
+      "loss": 0.7085,
+      "step": 5768
+    },
+    {
+      "epoch": 0.8827850038255547,
+      "grad_norm": 2.4521500514536734,
+      "learning_rate": 1.2362697601451055e-05,
+      "loss": 0.6949,
+      "step": 5769
+    },
+    {
+      "epoch": 0.882938026013772,
+      "grad_norm": 2.3315675840027112,
+      "learning_rate": 1.2360289509427657e-05,
+      "loss": 0.6517,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8830910482019892,
+      "grad_norm": 2.4804572529648365,
+      "learning_rate": 1.2357881272449645e-05,
+      "loss": 0.7527,
+      "step": 5771
+    },
+    {
+      "epoch": 0.8832440703902066,
+      "grad_norm": 2.3055905996779162,
+      "learning_rate": 1.2355472890664921e-05,
+      "loss": 0.6958,
+      "step": 5772
+    },
+    {
+      "epoch": 0.8833970925784239,
+      "grad_norm": 2.2077589919529896,
+      "learning_rate": 1.2353064364221394e-05,
+      "loss": 0.7385,
+      "step": 5773
+    },
+    {
+      "epoch": 0.8835501147666411,
+      "grad_norm": 2.5673498996652393,
+      "learning_rate": 1.2350655693266977e-05,
+      "loss": 0.7746,
+      "step": 5774
+    },
+    {
+      "epoch": 0.8837031369548585,
+      "grad_norm": 2.327867947254067,
+      "learning_rate": 1.23482468779496e-05,
+      "loss": 0.7871,
+      "step": 5775
+    },
+    {
+      "epoch": 0.8838561591430757,
+      "grad_norm": 1.994862966333433,
+      "learning_rate": 1.2345837918417192e-05,
+      "loss": 0.5689,
+      "step": 5776
+    },
+    {
+      "epoch": 0.884009181331293,
+      "grad_norm": 2.1628077023188577,
+      "learning_rate": 1.2343428814817704e-05,
+      "loss": 0.6696,
+      "step": 5777
+    },
+    {
+      "epoch": 0.8841622035195104,
+      "grad_norm": 2.4281221978724177,
+      "learning_rate": 1.2341019567299084e-05,
+      "loss": 0.7309,
+      "step": 5778
+    },
+    {
+      "epoch": 0.8843152257077276,
+      "grad_norm": 2.415771519746245,
+      "learning_rate": 1.2338610176009294e-05,
+      "loss": 0.7888,
+      "step": 5779
+    },
+    {
+      "epoch": 0.8844682478959449,
+      "grad_norm": 2.0146648291440132,
+      "learning_rate": 1.2336200641096302e-05,
+      "loss": 0.7183,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8846212700841622,
+      "grad_norm": 2.101105174830713,
+      "learning_rate": 1.2333790962708096e-05,
+      "loss": 0.6244,
+      "step": 5781
+    },
+    {
+      "epoch": 0.8847742922723795,
+      "grad_norm": 2.747095291584874,
+      "learning_rate": 1.233138114099265e-05,
+      "loss": 0.7718,
+      "step": 5782
+    },
+    {
+      "epoch": 0.8849273144605968,
+      "grad_norm": 2.1914791612843842,
+      "learning_rate": 1.2328971176097973e-05,
+      "loss": 0.7515,
+      "step": 5783
+    },
+    {
+      "epoch": 0.885080336648814,
+      "grad_norm": 2.2870028208167317,
+      "learning_rate": 1.2326561068172063e-05,
+      "loss": 0.6468,
+      "step": 5784
+    },
+    {
+      "epoch": 0.8852333588370314,
+      "grad_norm": 2.27622510477052,
+      "learning_rate": 1.2324150817362934e-05,
+      "loss": 0.6705,
+      "step": 5785
+    },
+    {
+      "epoch": 0.8853863810252487,
+      "grad_norm": 2.3932459194155125,
+      "learning_rate": 1.2321740423818614e-05,
+      "loss": 0.72,
+      "step": 5786
+    },
+    {
+      "epoch": 0.8855394032134659,
+      "grad_norm": 2.0056690093491496,
+      "learning_rate": 1.2319329887687132e-05,
+      "loss": 0.6506,
+      "step": 5787
+    },
+    {
+      "epoch": 0.8856924254016832,
+      "grad_norm": 2.3515431421733064,
+      "learning_rate": 1.2316919209116527e-05,
+      "loss": 0.718,
+      "step": 5788
+    },
+    {
+      "epoch": 0.8858454475899006,
+      "grad_norm": 2.564976261897152,
+      "learning_rate": 1.2314508388254848e-05,
+      "loss": 0.7486,
+      "step": 5789
+    },
+    {
+      "epoch": 0.8859984697781178,
+      "grad_norm": 2.286392560336342,
+      "learning_rate": 1.2312097425250157e-05,
+      "loss": 0.5844,
+      "step": 5790
+    },
+    {
+      "epoch": 0.8861514919663351,
+      "grad_norm": 2.40196016256603,
+      "learning_rate": 1.230968632025052e-05,
+      "loss": 0.6324,
+      "step": 5791
+    },
+    {
+      "epoch": 0.8863045141545524,
+      "grad_norm": 2.1219767397301377,
+      "learning_rate": 1.2307275073404005e-05,
+      "loss": 0.6549,
+      "step": 5792
+    },
+    {
+      "epoch": 0.8864575363427697,
+      "grad_norm": 2.1545723623205353,
+      "learning_rate": 1.2304863684858708e-05,
+      "loss": 0.6666,
+      "step": 5793
+    },
+    {
+      "epoch": 0.886610558530987,
+      "grad_norm": 2.1982785238169624,
+      "learning_rate": 1.2302452154762711e-05,
+      "loss": 0.7322,
+      "step": 5794
+    },
+    {
+      "epoch": 0.8867635807192042,
+      "grad_norm": 1.9445798805354053,
+      "learning_rate": 1.230004048326412e-05,
+      "loss": 0.5734,
+      "step": 5795
+    },
+    {
+      "epoch": 0.8869166029074216,
+      "grad_norm": 2.097172733466339,
+      "learning_rate": 1.2297628670511046e-05,
+      "loss": 0.5911,
+      "step": 5796
+    },
+    {
+      "epoch": 0.8870696250956389,
+      "grad_norm": 2.2876608565036736,
+      "learning_rate": 1.229521671665161e-05,
+      "loss": 0.823,
+      "step": 5797
+    },
+    {
+      "epoch": 0.8872226472838561,
+      "grad_norm": 2.130982677926774,
+      "learning_rate": 1.2292804621833932e-05,
+      "loss": 0.6003,
+      "step": 5798
+    },
+    {
+      "epoch": 0.8873756694720735,
+      "grad_norm": 2.3899589697261012,
+      "learning_rate": 1.2290392386206153e-05,
+      "loss": 0.7014,
+      "step": 5799
+    },
+    {
+      "epoch": 0.8875286916602907,
+      "grad_norm": 2.463939146997431,
+      "learning_rate": 1.228798000991642e-05,
+      "loss": 0.6992,
+      "step": 5800
+    },
+    {
+      "epoch": 0.887681713848508,
+      "grad_norm": 2.405366156872562,
+      "learning_rate": 1.228556749311288e-05,
+      "loss": 0.7654,
+      "step": 5801
+    },
+    {
+      "epoch": 0.8878347360367254,
+      "grad_norm": 1.8746932186360155,
+      "learning_rate": 1.2283154835943704e-05,
+      "loss": 0.5216,
+      "step": 5802
+    },
+    {
+      "epoch": 0.8879877582249426,
+      "grad_norm": 2.1661546901574704,
+      "learning_rate": 1.2280742038557056e-05,
+      "loss": 0.7372,
+      "step": 5803
+    },
+    {
+      "epoch": 0.8881407804131599,
+      "grad_norm": 2.493655678620059,
+      "learning_rate": 1.2278329101101116e-05,
+      "loss": 0.7774,
+      "step": 5804
+    },
+    {
+      "epoch": 0.8882938026013772,
+      "grad_norm": 2.0972516435366106,
+      "learning_rate": 1.2275916023724072e-05,
+      "loss": 0.6212,
+      "step": 5805
+    },
+    {
+      "epoch": 0.8884468247895945,
+      "grad_norm": 2.050606086919499,
+      "learning_rate": 1.2273502806574126e-05,
+      "loss": 0.7213,
+      "step": 5806
+    },
+    {
+      "epoch": 0.8885998469778118,
+      "grad_norm": 2.2270183694432175,
+      "learning_rate": 1.2271089449799476e-05,
+      "loss": 0.7031,
+      "step": 5807
+    },
+    {
+      "epoch": 0.888752869166029,
+      "grad_norm": 2.2753592897527097,
+      "learning_rate": 1.2268675953548336e-05,
+      "loss": 0.6464,
+      "step": 5808
+    },
+    {
+      "epoch": 0.8889058913542464,
+      "grad_norm": 2.2711193654457777,
+      "learning_rate": 1.2266262317968934e-05,
+      "loss": 0.7809,
+      "step": 5809
+    },
+    {
+      "epoch": 0.8890589135424637,
+      "grad_norm": 2.1135815709154793,
+      "learning_rate": 1.2263848543209496e-05,
+      "loss": 0.6955,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8892119357306809,
+      "grad_norm": 2.4245861015340573,
+      "learning_rate": 1.2261434629418263e-05,
+      "loss": 0.7699,
+      "step": 5811
+    },
+    {
+      "epoch": 0.8893649579188982,
+      "grad_norm": 2.102041027083013,
+      "learning_rate": 1.2259020576743484e-05,
+      "loss": 0.6489,
+      "step": 5812
+    },
+    {
+      "epoch": 0.8895179801071156,
+      "grad_norm": 2.0571284723371948,
+      "learning_rate": 1.2256606385333411e-05,
+      "loss": 0.7392,
+      "step": 5813
+    },
+    {
+      "epoch": 0.8896710022953328,
+      "grad_norm": 2.232074088650882,
+      "learning_rate": 1.2254192055336315e-05,
+      "loss": 0.7079,
+      "step": 5814
+    },
+    {
+      "epoch": 0.8898240244835501,
+      "grad_norm": 2.3912417483625164,
+      "learning_rate": 1.2251777586900466e-05,
+      "loss": 0.7092,
+      "step": 5815
+    },
+    {
+      "epoch": 0.8899770466717674,
+      "grad_norm": 2.055983923638346,
+      "learning_rate": 1.2249362980174144e-05,
+      "loss": 0.6946,
+      "step": 5816
+    },
+    {
+      "epoch": 0.8901300688599847,
+      "grad_norm": 2.4091786231449146,
+      "learning_rate": 1.2246948235305642e-05,
+      "loss": 0.7588,
+      "step": 5817
+    },
+    {
+      "epoch": 0.890283091048202,
+      "grad_norm": 2.4550247748254446,
+      "learning_rate": 1.2244533352443262e-05,
+      "loss": 0.8143,
+      "step": 5818
+    },
+    {
+      "epoch": 0.8904361132364192,
+      "grad_norm": 2.44563168398848,
+      "learning_rate": 1.2242118331735306e-05,
+      "loss": 0.7611,
+      "step": 5819
+    },
+    {
+      "epoch": 0.8905891354246366,
+      "grad_norm": 2.0083209726168167,
+      "learning_rate": 1.223970317333009e-05,
+      "loss": 0.6456,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8907421576128539,
+      "grad_norm": 2.261824846327099,
+      "learning_rate": 1.2237287877375943e-05,
+      "loss": 0.659,
+      "step": 5821
+    },
+    {
+      "epoch": 0.8908951798010711,
+      "grad_norm": 2.415318778732395,
+      "learning_rate": 1.2234872444021197e-05,
+      "loss": 0.6324,
+      "step": 5822
+    },
+    {
+      "epoch": 0.8910482019892885,
+      "grad_norm": 2.491753442524876,
+      "learning_rate": 1.223245687341419e-05,
+      "loss": 0.8115,
+      "step": 5823
+    },
+    {
+      "epoch": 0.8912012241775057,
+      "grad_norm": 2.4285888845880415,
+      "learning_rate": 1.223004116570327e-05,
+      "loss": 0.7954,
+      "step": 5824
+    },
+    {
+      "epoch": 0.891354246365723,
+      "grad_norm": 2.238293963593015,
+      "learning_rate": 1.22276253210368e-05,
+      "loss": 0.7294,
+      "step": 5825
+    },
+    {
+      "epoch": 0.8915072685539404,
+      "grad_norm": 2.326987572185359,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.5903,
+      "step": 5826
+    },
+    {
+      "epoch": 0.8916602907421576,
+      "grad_norm": 2.1780223917114836,
+      "learning_rate": 1.222279322143068e-05,
+      "loss": 0.6964,
+      "step": 5827
+    },
+    {
+      "epoch": 0.8918133129303749,
+      "grad_norm": 2.048347926322804,
+      "learning_rate": 1.2220376966787785e-05,
+      "loss": 0.6111,
+      "step": 5828
+    },
+    {
+      "epoch": 0.8919663351185922,
+      "grad_norm": 2.003970443987533,
+      "learning_rate": 1.2217960575782856e-05,
+      "loss": 0.6416,
+      "step": 5829
+    },
+    {
+      "epoch": 0.8921193573068095,
+      "grad_norm": 2.55460997045148,
+      "learning_rate": 1.2215544048564294e-05,
+      "loss": 0.7348,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8922723794950268,
+      "grad_norm": 2.3815038136371496,
+      "learning_rate": 1.22131273852805e-05,
+      "loss": 0.7828,
+      "step": 5831
+    },
+    {
+      "epoch": 0.892425401683244,
+      "grad_norm": 2.185261352231627,
+      "learning_rate": 1.22107105860799e-05,
+      "loss": 0.6699,
+      "step": 5832
+    },
+    {
+      "epoch": 0.8925784238714614,
+      "grad_norm": 1.9530762615679256,
+      "learning_rate": 1.220829365111091e-05,
+      "loss": 0.5621,
+      "step": 5833
+    },
+    {
+      "epoch": 0.8927314460596787,
+      "grad_norm": 2.5679065405768693,
+      "learning_rate": 1.2205876580521971e-05,
+      "loss": 0.7618,
+      "step": 5834
+    },
+    {
+      "epoch": 0.8928844682478959,
+      "grad_norm": 1.9323644230313388,
+      "learning_rate": 1.2203459374461522e-05,
+      "loss": 0.7202,
+      "step": 5835
+    },
+    {
+      "epoch": 0.8930374904361132,
+      "grad_norm": 2.1426859186740352,
+      "learning_rate": 1.220104203307801e-05,
+      "loss": 0.672,
+      "step": 5836
+    },
+    {
+      "epoch": 0.8931905126243306,
+      "grad_norm": 2.4556090557348873,
+      "learning_rate": 1.2198624556519899e-05,
+      "loss": 0.6719,
+      "step": 5837
+    },
+    {
+      "epoch": 0.8933435348125478,
+      "grad_norm": 2.411343782082261,
+      "learning_rate": 1.219620694493565e-05,
+      "loss": 0.8695,
+      "step": 5838
+    },
+    {
+      "epoch": 0.8934965570007651,
+      "grad_norm": 1.9229137910731127,
+      "learning_rate": 1.2193789198473743e-05,
+      "loss": 0.5691,
+      "step": 5839
+    },
+    {
+      "epoch": 0.8936495791889824,
+      "grad_norm": 2.7917967312533083,
+      "learning_rate": 1.2191371317282659e-05,
+      "loss": 0.9142,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8938026013771997,
+      "grad_norm": 2.7345209669423927,
+      "learning_rate": 1.2188953301510891e-05,
+      "loss": 0.7448,
+      "step": 5841
+    },
+    {
+      "epoch": 0.893955623565417,
+      "grad_norm": 2.2727678554130857,
+      "learning_rate": 1.2186535151306934e-05,
+      "loss": 0.6421,
+      "step": 5842
+    },
+    {
+      "epoch": 0.8941086457536342,
+      "grad_norm": 2.258694735574518,
+      "learning_rate": 1.2184116866819303e-05,
+      "loss": 0.7776,
+      "step": 5843
+    },
+    {
+      "epoch": 0.8942616679418516,
+      "grad_norm": 2.098915575747473,
+      "learning_rate": 1.2181698448196508e-05,
+      "loss": 0.6931,
+      "step": 5844
+    },
+    {
+      "epoch": 0.8944146901300689,
+      "grad_norm": 2.3098715398217706,
+      "learning_rate": 1.2179279895587078e-05,
+      "loss": 0.7966,
+      "step": 5845
+    },
+    {
+      "epoch": 0.8945677123182861,
+      "grad_norm": 2.1839510482142255,
+      "learning_rate": 1.2176861209139543e-05,
+      "loss": 0.601,
+      "step": 5846
+    },
+    {
+      "epoch": 0.8947207345065035,
+      "grad_norm": 2.186745048637725,
+      "learning_rate": 1.217444238900245e-05,
+      "loss": 0.6878,
+      "step": 5847
+    },
+    {
+      "epoch": 0.8948737566947207,
+      "grad_norm": 2.3464306103822428,
+      "learning_rate": 1.2172023435324336e-05,
+      "loss": 0.6573,
+      "step": 5848
+    },
+    {
+      "epoch": 0.895026778882938,
+      "grad_norm": 2.1034600448438043,
+      "learning_rate": 1.2169604348253772e-05,
+      "loss": 0.7129,
+      "step": 5849
+    },
+    {
+      "epoch": 0.8951798010711554,
+      "grad_norm": 2.4272691681079164,
+      "learning_rate": 1.2167185127939314e-05,
+      "loss": 0.8234,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8953328232593726,
+      "grad_norm": 2.3142269494809873,
+      "learning_rate": 1.2164765774529541e-05,
+      "loss": 0.8082,
+      "step": 5851
+    },
+    {
+      "epoch": 0.8954858454475899,
+      "grad_norm": 2.2118916809347557,
+      "learning_rate": 1.2162346288173033e-05,
+      "loss": 0.7194,
+      "step": 5852
+    },
+    {
+      "epoch": 0.8956388676358072,
+      "grad_norm": 2.654478274331671,
+      "learning_rate": 1.215992666901838e-05,
+      "loss": 0.7258,
+      "step": 5853
+    },
+    {
+      "epoch": 0.8957918898240245,
+      "grad_norm": 2.1917989705330174,
+      "learning_rate": 1.2157506917214179e-05,
+      "loss": 0.7053,
+      "step": 5854
+    },
+    {
+      "epoch": 0.8959449120122418,
+      "grad_norm": 2.295237595115881,
+      "learning_rate": 1.215508703290904e-05,
+      "loss": 0.8176,
+      "step": 5855
+    },
+    {
+      "epoch": 0.896097934200459,
+      "grad_norm": 2.0125203316125724,
+      "learning_rate": 1.2152667016251575e-05,
+      "loss": 0.6165,
+      "step": 5856
+    },
+    {
+      "epoch": 0.8962509563886764,
+      "grad_norm": 2.137048919743919,
+      "learning_rate": 1.2150246867390407e-05,
+      "loss": 0.6266,
+      "step": 5857
+    },
+    {
+      "epoch": 0.8964039785768937,
+      "grad_norm": 2.3382790666248825,
+      "learning_rate": 1.2147826586474168e-05,
+      "loss": 0.7635,
+      "step": 5858
+    },
+    {
+      "epoch": 0.8965570007651109,
+      "grad_norm": 2.1318790415160183,
+      "learning_rate": 1.2145406173651497e-05,
+      "loss": 0.6476,
+      "step": 5859
+    },
+    {
+      "epoch": 0.8967100229533282,
+      "grad_norm": 2.169522268623894,
+      "learning_rate": 1.2142985629071037e-05,
+      "loss": 0.6822,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8968630451415456,
+      "grad_norm": 2.0656815045352417,
+      "learning_rate": 1.2140564952881446e-05,
+      "loss": 0.5719,
+      "step": 5861
+    },
+    {
+      "epoch": 0.8970160673297628,
+      "grad_norm": 2.0940298380060023,
+      "learning_rate": 1.2138144145231387e-05,
+      "loss": 0.6189,
+      "step": 5862
+    },
+    {
+      "epoch": 0.8971690895179801,
+      "grad_norm": 2.1912680322446243,
+      "learning_rate": 1.2135723206269535e-05,
+      "loss": 0.7313,
+      "step": 5863
+    },
+    {
+      "epoch": 0.8973221117061974,
+      "grad_norm": 2.042693256307619,
+      "learning_rate": 1.2133302136144564e-05,
+      "loss": 0.5883,
+      "step": 5864
+    },
+    {
+      "epoch": 0.8974751338944147,
+      "grad_norm": 2.162572034162192,
+      "learning_rate": 1.2130880935005165e-05,
+      "loss": 0.781,
+      "step": 5865
+    },
+    {
+      "epoch": 0.897628156082632,
+      "grad_norm": 2.0853415311565113,
+      "learning_rate": 1.212845960300003e-05,
+      "loss": 0.6414,
+      "step": 5866
+    },
+    {
+      "epoch": 0.8977811782708492,
+      "grad_norm": 2.306821724711259,
+      "learning_rate": 1.2126038140277863e-05,
+      "loss": 0.6421,
+      "step": 5867
+    },
+    {
+      "epoch": 0.8979342004590666,
+      "grad_norm": 2.3155477340665698,
+      "learning_rate": 1.212361654698738e-05,
+      "loss": 0.6216,
+      "step": 5868
+    },
+    {
+      "epoch": 0.8980872226472839,
+      "grad_norm": 2.2402002639166936,
+      "learning_rate": 1.2121194823277294e-05,
+      "loss": 0.6306,
+      "step": 5869
+    },
+    {
+      "epoch": 0.8982402448355011,
+      "grad_norm": 2.2497974374833793,
+      "learning_rate": 1.2118772969296337e-05,
+      "loss": 0.6934,
+      "step": 5870
+    },
+    {
+      "epoch": 0.8983932670237185,
+      "grad_norm": 2.2281448098479264,
+      "learning_rate": 1.2116350985193243e-05,
+      "loss": 0.6245,
+      "step": 5871
+    },
+    {
+      "epoch": 0.8985462892119357,
+      "grad_norm": 2.199538303015666,
+      "learning_rate": 1.2113928871116758e-05,
+      "loss": 0.7062,
+      "step": 5872
+    },
+    {
+      "epoch": 0.898699311400153,
+      "grad_norm": 2.0709420388921207,
+      "learning_rate": 1.211150662721563e-05,
+      "loss": 0.6909,
+      "step": 5873
+    },
+    {
+      "epoch": 0.8988523335883704,
+      "grad_norm": 2.5142622500922154,
+      "learning_rate": 1.2109084253638617e-05,
+      "loss": 0.8169,
+      "step": 5874
+    },
+    {
+      "epoch": 0.8990053557765876,
+      "grad_norm": 2.140104777112769,
+      "learning_rate": 1.210666175053449e-05,
+      "loss": 0.6359,
+      "step": 5875
+    },
+    {
+      "epoch": 0.8991583779648049,
+      "grad_norm": 2.164375390414777,
+      "learning_rate": 1.2104239118052027e-05,
+      "loss": 0.6977,
+      "step": 5876
+    },
+    {
+      "epoch": 0.8993114001530221,
+      "grad_norm": 2.1929526697694,
+      "learning_rate": 1.2101816356340004e-05,
+      "loss": 0.6797,
+      "step": 5877
+    },
+    {
+      "epoch": 0.8994644223412395,
+      "grad_norm": 2.3775107486140894,
+      "learning_rate": 1.2099393465547218e-05,
+      "loss": 0.8075,
+      "step": 5878
+    },
+    {
+      "epoch": 0.8996174445294568,
+      "grad_norm": 2.19374146725586,
+      "learning_rate": 1.2096970445822467e-05,
+      "loss": 0.6913,
+      "step": 5879
+    },
+    {
+      "epoch": 0.899770466717674,
+      "grad_norm": 2.356353536060419,
+      "learning_rate": 1.209454729731456e-05,
+      "loss": 0.6564,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8999234889058914,
+      "grad_norm": 2.2970947492536826,
+      "learning_rate": 1.2092124020172304e-05,
+      "loss": 0.7165,
+      "step": 5881
+    },
+    {
+      "epoch": 0.9000765110941087,
+      "grad_norm": 2.2264397828108335,
+      "learning_rate": 1.208970061454453e-05,
+      "loss": 0.7145,
+      "step": 5882
+    },
+    {
+      "epoch": 0.9002295332823259,
+      "grad_norm": 2.2895364047526754,
+      "learning_rate": 1.2087277080580064e-05,
+      "loss": 0.7901,
+      "step": 5883
+    },
+    {
+      "epoch": 0.9003825554705432,
+      "grad_norm": 2.6582101773237494,
+      "learning_rate": 1.2084853418427754e-05,
+      "loss": 0.729,
+      "step": 5884
+    },
+    {
+      "epoch": 0.9005355776587605,
+      "grad_norm": 2.4029001974784734,
+      "learning_rate": 1.2082429628236433e-05,
+      "loss": 0.663,
+      "step": 5885
+    },
+    {
+      "epoch": 0.9006885998469778,
+      "grad_norm": 2.2788746677911753,
+      "learning_rate": 1.2080005710154962e-05,
+      "loss": 0.6018,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9008416220351951,
+      "grad_norm": 2.2037783198820864,
+      "learning_rate": 1.2077581664332205e-05,
+      "loss": 0.6541,
+      "step": 5887
+    },
+    {
+      "epoch": 0.9009946442234124,
+      "grad_norm": 2.5472288046633587,
+      "learning_rate": 1.2075157490917033e-05,
+      "loss": 0.7121,
+      "step": 5888
+    },
+    {
+      "epoch": 0.9011476664116297,
+      "grad_norm": 2.263169854265418,
+      "learning_rate": 1.2072733190058319e-05,
+      "loss": 0.723,
+      "step": 5889
+    },
+    {
+      "epoch": 0.901300688599847,
+      "grad_norm": 2.565421276878311,
+      "learning_rate": 1.2070308761904949e-05,
+      "loss": 0.7627,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9014537107880642,
+      "grad_norm": 2.33516891773541,
+      "learning_rate": 1.2067884206605821e-05,
+      "loss": 0.6602,
+      "step": 5891
+    },
+    {
+      "epoch": 0.9016067329762816,
+      "grad_norm": 2.369632550813917,
+      "learning_rate": 1.2065459524309838e-05,
+      "loss": 0.7074,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9017597551644988,
+      "grad_norm": 2.245894801979679,
+      "learning_rate": 1.20630347151659e-05,
+      "loss": 0.7945,
+      "step": 5893
+    },
+    {
+      "epoch": 0.9019127773527161,
+      "grad_norm": 2.133776733386824,
+      "learning_rate": 1.2060609779322932e-05,
+      "loss": 0.6325,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9020657995409335,
+      "grad_norm": 2.42818076004384,
+      "learning_rate": 1.2058184716929858e-05,
+      "loss": 0.8039,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9022188217291507,
+      "grad_norm": 2.1864609949528937,
+      "learning_rate": 1.2055759528135604e-05,
+      "loss": 0.6626,
+      "step": 5896
+    },
+    {
+      "epoch": 0.902371843917368,
+      "grad_norm": 2.3056942365670166,
+      "learning_rate": 1.2053334213089119e-05,
+      "loss": 0.5935,
+      "step": 5897
+    },
+    {
+      "epoch": 0.9025248661055854,
+      "grad_norm": 2.4474426076241613,
+      "learning_rate": 1.2050908771939347e-05,
+      "loss": 0.6779,
+      "step": 5898
+    },
+    {
+      "epoch": 0.9026778882938026,
+      "grad_norm": 2.245242913146272,
+      "learning_rate": 1.204848320483524e-05,
+      "loss": 0.672,
+      "step": 5899
+    },
+    {
+      "epoch": 0.9028309104820199,
+      "grad_norm": 2.4002007892851376,
+      "learning_rate": 1.2046057511925773e-05,
+      "loss": 0.6885,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9029839326702371,
+      "grad_norm": 2.0513556880468804,
+      "learning_rate": 1.2043631693359906e-05,
+      "loss": 0.6375,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9031369548584545,
+      "grad_norm": 2.185620019488159,
+      "learning_rate": 1.2041205749286617e-05,
+      "loss": 0.6056,
+      "step": 5902
+    },
+    {
+      "epoch": 0.9032899770466718,
+      "grad_norm": 2.5696711510264727,
+      "learning_rate": 1.2038779679854905e-05,
+      "loss": 0.8615,
+      "step": 5903
+    },
+    {
+      "epoch": 0.903442999234889,
+      "grad_norm": 2.427270547334622,
+      "learning_rate": 1.2036353485213756e-05,
+      "loss": 0.7441,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9035960214231064,
+      "grad_norm": 2.6216911775274254,
+      "learning_rate": 1.2033927165512171e-05,
+      "loss": 0.6633,
+      "step": 5905
+    },
+    {
+      "epoch": 0.9037490436113237,
+      "grad_norm": 2.3969872880103655,
+      "learning_rate": 1.2031500720899164e-05,
+      "loss": 0.7292,
+      "step": 5906
+    },
+    {
+      "epoch": 0.9039020657995409,
+      "grad_norm": 2.3206810022930435,
+      "learning_rate": 1.2029074151523747e-05,
+      "loss": 0.6184,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9040550879877582,
+      "grad_norm": 2.1248876858842998,
+      "learning_rate": 1.2026647457534953e-05,
+      "loss": 0.6063,
+      "step": 5908
+    },
+    {
+      "epoch": 0.9042081101759755,
+      "grad_norm": 2.105010540329758,
+      "learning_rate": 1.2024220639081807e-05,
+      "loss": 0.6701,
+      "step": 5909
+    },
+    {
+      "epoch": 0.9043611323641928,
+      "grad_norm": 2.390961362332351,
+      "learning_rate": 1.2021793696313355e-05,
+      "loss": 0.8372,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9045141545524101,
+      "grad_norm": 2.3906835802026634,
+      "learning_rate": 1.2019366629378642e-05,
+      "loss": 0.6914,
+      "step": 5911
+    },
+    {
+      "epoch": 0.9046671767406274,
+      "grad_norm": 2.331666879806483,
+      "learning_rate": 1.2016939438426722e-05,
+      "loss": 0.7861,
+      "step": 5912
+    },
+    {
+      "epoch": 0.9048201989288447,
+      "grad_norm": 2.2358574061590173,
+      "learning_rate": 1.2014512123606665e-05,
+      "loss": 0.6219,
+      "step": 5913
+    },
+    {
+      "epoch": 0.904973221117062,
+      "grad_norm": 2.106600490640814,
+      "learning_rate": 1.2012084685067535e-05,
+      "loss": 0.7043,
+      "step": 5914
+    },
+    {
+      "epoch": 0.9051262433052792,
+      "grad_norm": 2.2539604314687325,
+      "learning_rate": 1.2009657122958413e-05,
+      "loss": 0.6648,
+      "step": 5915
+    },
+    {
+      "epoch": 0.9052792654934966,
+      "grad_norm": 2.0671088263249464,
+      "learning_rate": 1.2007229437428387e-05,
+      "loss": 0.5475,
+      "step": 5916
+    },
+    {
+      "epoch": 0.9054322876817138,
+      "grad_norm": 2.1790247947962507,
+      "learning_rate": 1.200480162862655e-05,
+      "loss": 0.735,
+      "step": 5917
+    },
+    {
+      "epoch": 0.9055853098699311,
+      "grad_norm": 2.160285319705886,
+      "learning_rate": 1.2002373696701999e-05,
+      "loss": 0.6605,
+      "step": 5918
+    },
+    {
+      "epoch": 0.9057383320581485,
+      "grad_norm": 2.669653172417057,
+      "learning_rate": 1.1999945641803845e-05,
+      "loss": 0.7651,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9058913542463657,
+      "grad_norm": 2.332008350846348,
+      "learning_rate": 1.199751746408121e-05,
+      "loss": 0.7304,
+      "step": 5920
+    },
+    {
+      "epoch": 0.906044376434583,
+      "grad_norm": 2.3630854316639627,
+      "learning_rate": 1.199508916368321e-05,
+      "loss": 0.7975,
+      "step": 5921
+    },
+    {
+      "epoch": 0.9061973986228004,
+      "grad_norm": 2.1925117093837962,
+      "learning_rate": 1.1992660740758981e-05,
+      "loss": 0.6704,
+      "step": 5922
+    },
+    {
+      "epoch": 0.9063504208110176,
+      "grad_norm": 2.2723588771999528,
+      "learning_rate": 1.1990232195457659e-05,
+      "loss": 0.7037,
+      "step": 5923
+    },
+    {
+      "epoch": 0.9065034429992349,
+      "grad_norm": 2.3270779882957937,
+      "learning_rate": 1.1987803527928393e-05,
+      "loss": 0.6958,
+      "step": 5924
+    },
+    {
+      "epoch": 0.9066564651874521,
+      "grad_norm": 2.2060500688080227,
+      "learning_rate": 1.1985374738320337e-05,
+      "loss": 0.5766,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9068094873756695,
+      "grad_norm": 2.2317586164719,
+      "learning_rate": 1.1982945826782651e-05,
+      "loss": 0.6561,
+      "step": 5926
+    },
+    {
+      "epoch": 0.9069625095638868,
+      "grad_norm": 2.322289561129461,
+      "learning_rate": 1.19805167934645e-05,
+      "loss": 0.6501,
+      "step": 5927
+    },
+    {
+      "epoch": 0.907115531752104,
+      "grad_norm": 2.515443301034443,
+      "learning_rate": 1.197808763851507e-05,
+      "loss": 0.7345,
+      "step": 5928
+    },
+    {
+      "epoch": 0.9072685539403214,
+      "grad_norm": 2.2527717426764338,
+      "learning_rate": 1.1975658362083542e-05,
+      "loss": 0.7293,
+      "step": 5929
+    },
+    {
+      "epoch": 0.9074215761285387,
+      "grad_norm": 2.444160848812312,
+      "learning_rate": 1.19732289643191e-05,
+      "loss": 0.7744,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9075745983167559,
+      "grad_norm": 2.0008299026167546,
+      "learning_rate": 1.1970799445370948e-05,
+      "loss": 0.5623,
+      "step": 5931
+    },
+    {
+      "epoch": 0.9077276205049732,
+      "grad_norm": 2.2752478064076023,
+      "learning_rate": 1.1968369805388295e-05,
+      "loss": 0.6978,
+      "step": 5932
+    },
+    {
+      "epoch": 0.9078806426931905,
+      "grad_norm": 2.583491640352064,
+      "learning_rate": 1.196594004452035e-05,
+      "loss": 0.7084,
+      "step": 5933
+    },
+    {
+      "epoch": 0.9080336648814078,
+      "grad_norm": 2.6861883035754706,
+      "learning_rate": 1.1963510162916339e-05,
+      "loss": 0.754,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9081866870696251,
+      "grad_norm": 2.094923336464147,
+      "learning_rate": 1.1961080160725484e-05,
+      "loss": 0.6658,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9083397092578424,
+      "grad_norm": 2.3754655574983934,
+      "learning_rate": 1.1958650038097029e-05,
+      "loss": 0.7497,
+      "step": 5936
+    },
+    {
+      "epoch": 0.9084927314460597,
+      "grad_norm": 2.247236567509479,
+      "learning_rate": 1.1956219795180209e-05,
+      "loss": 0.6799,
+      "step": 5937
+    },
+    {
+      "epoch": 0.908645753634277,
+      "grad_norm": 2.176455805806765,
+      "learning_rate": 1.1953789432124279e-05,
+      "loss": 0.7143,
+      "step": 5938
+    },
+    {
+      "epoch": 0.9087987758224942,
+      "grad_norm": 2.1382985224906443,
+      "learning_rate": 1.1951358949078497e-05,
+      "loss": 0.6692,
+      "step": 5939
+    },
+    {
+      "epoch": 0.9089517980107116,
+      "grad_norm": 2.2784821848914487,
+      "learning_rate": 1.1948928346192128e-05,
+      "loss": 0.7507,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9091048201989288,
+      "grad_norm": 2.094814546778022,
+      "learning_rate": 1.1946497623614448e-05,
+      "loss": 0.5985,
+      "step": 5941
+    },
+    {
+      "epoch": 0.9092578423871461,
+      "grad_norm": 2.0659105778735283,
+      "learning_rate": 1.194406678149473e-05,
+      "loss": 0.5888,
+      "step": 5942
+    },
+    {
+      "epoch": 0.9094108645753635,
+      "grad_norm": 2.529484983416761,
+      "learning_rate": 1.1941635819982267e-05,
+      "loss": 0.7306,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9095638867635807,
+      "grad_norm": 2.316500778173214,
+      "learning_rate": 1.1939204739226353e-05,
+      "loss": 0.6232,
+      "step": 5944
+    },
+    {
+      "epoch": 0.909716908951798,
+      "grad_norm": 2.2416933756176545,
+      "learning_rate": 1.193677353937629e-05,
+      "loss": 0.6686,
+      "step": 5945
+    },
+    {
+      "epoch": 0.9098699311400154,
+      "grad_norm": 2.296889261532724,
+      "learning_rate": 1.1934342220581384e-05,
+      "loss": 0.5793,
+      "step": 5946
+    },
+    {
+      "epoch": 0.9100229533282326,
+      "grad_norm": 2.4069543424400677,
+      "learning_rate": 1.193191078299096e-05,
+      "loss": 0.8436,
+      "step": 5947
+    },
+    {
+      "epoch": 0.9101759755164499,
+      "grad_norm": 2.162792615318407,
+      "learning_rate": 1.1929479226754338e-05,
+      "loss": 0.7032,
+      "step": 5948
+    },
+    {
+      "epoch": 0.9103289977046671,
+      "grad_norm": 2.2406697992043254,
+      "learning_rate": 1.1927047552020845e-05,
+      "loss": 0.6408,
+      "step": 5949
+    },
+    {
+      "epoch": 0.9104820198928845,
+      "grad_norm": 2.4229962303297103,
+      "learning_rate": 1.1924615758939824e-05,
+      "loss": 0.6942,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9106350420811018,
+      "grad_norm": 2.467996134227574,
+      "learning_rate": 1.192218384766062e-05,
+      "loss": 0.782,
+      "step": 5951
+    },
+    {
+      "epoch": 0.910788064269319,
+      "grad_norm": 2.4342339180377603,
+      "learning_rate": 1.1919751818332586e-05,
+      "loss": 0.688,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9109410864575364,
+      "grad_norm": 2.1932437749285167,
+      "learning_rate": 1.191731967110508e-05,
+      "loss": 0.6378,
+      "step": 5953
+    },
+    {
+      "epoch": 0.9110941086457537,
+      "grad_norm": 2.055801571139704,
+      "learning_rate": 1.1914887406127478e-05,
+      "loss": 0.6698,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9112471308339709,
+      "grad_norm": 2.2684464246899,
+      "learning_rate": 1.1912455023549147e-05,
+      "loss": 0.6851,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9114001530221882,
+      "grad_norm": 2.5037197092461922,
+      "learning_rate": 1.1910022523519468e-05,
+      "loss": 0.7673,
+      "step": 5956
+    },
+    {
+      "epoch": 0.9115531752104055,
+      "grad_norm": 1.9131788391694138,
+      "learning_rate": 1.1907589906187837e-05,
+      "loss": 0.5243,
+      "step": 5957
+    },
+    {
+      "epoch": 0.9117061973986228,
+      "grad_norm": 2.234134469025724,
+      "learning_rate": 1.1905157171703651e-05,
+      "loss": 0.6361,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9118592195868401,
+      "grad_norm": 2.210064411480152,
+      "learning_rate": 1.1902724320216304e-05,
+      "loss": 0.6753,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9120122417750574,
+      "grad_norm": 2.3378206504402086,
+      "learning_rate": 1.1900291351875215e-05,
+      "loss": 0.6271,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9121652639632747,
+      "grad_norm": 2.2103620527103556,
+      "learning_rate": 1.18978582668298e-05,
+      "loss": 0.6063,
+      "step": 5961
+    },
+    {
+      "epoch": 0.912318286151492,
+      "grad_norm": 2.265094089348472,
+      "learning_rate": 1.1895425065229487e-05,
+      "loss": 0.6786,
+      "step": 5962
+    },
+    {
+      "epoch": 0.9124713083397092,
+      "grad_norm": 2.1542940090981553,
+      "learning_rate": 1.1892991747223704e-05,
+      "loss": 0.6719,
+      "step": 5963
+    },
+    {
+      "epoch": 0.9126243305279266,
+      "grad_norm": 2.1440181946848904,
+      "learning_rate": 1.189055831296189e-05,
+      "loss": 0.5932,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9127773527161438,
+      "grad_norm": 2.0963045032970906,
+      "learning_rate": 1.1888124762593496e-05,
+      "loss": 0.6104,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9129303749043611,
+      "grad_norm": 2.5262671669907633,
+      "learning_rate": 1.1885691096267975e-05,
+      "loss": 0.717,
+      "step": 5966
+    },
+    {
+      "epoch": 0.9130833970925785,
+      "grad_norm": 2.2639544983140274,
+      "learning_rate": 1.1883257314134787e-05,
+      "loss": 0.6589,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9132364192807957,
+      "grad_norm": 2.020210103595384,
+      "learning_rate": 1.1880823416343397e-05,
+      "loss": 0.6291,
+      "step": 5968
+    },
+    {
+      "epoch": 0.913389441469013,
+      "grad_norm": 2.1045346358957007,
+      "learning_rate": 1.1878389403043284e-05,
+      "loss": 0.6635,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9135424636572304,
+      "grad_norm": 2.0845494979682653,
+      "learning_rate": 1.1875955274383934e-05,
+      "loss": 0.6641,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9136954858454476,
+      "grad_norm": 2.449931136792063,
+      "learning_rate": 1.1873521030514826e-05,
+      "loss": 0.7608,
+      "step": 5971
+    },
+    {
+      "epoch": 0.9138485080336649,
+      "grad_norm": 2.4427533981065226,
+      "learning_rate": 1.1871086671585465e-05,
+      "loss": 0.804,
+      "step": 5972
+    },
+    {
+      "epoch": 0.9140015302218821,
+      "grad_norm": 2.196649531828342,
+      "learning_rate": 1.1868652197745351e-05,
+      "loss": 0.5977,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9141545524100995,
+      "grad_norm": 2.0401281352602343,
+      "learning_rate": 1.1866217609143998e-05,
+      "loss": 0.7304,
+      "step": 5974
+    },
+    {
+      "epoch": 0.9143075745983168,
+      "grad_norm": 2.2225569238419274,
+      "learning_rate": 1.1863782905930918e-05,
+      "loss": 0.6879,
+      "step": 5975
+    },
+    {
+      "epoch": 0.914460596786534,
+      "grad_norm": 2.415433043337009,
+      "learning_rate": 1.186134808825564e-05,
+      "loss": 0.7098,
+      "step": 5976
+    },
+    {
+      "epoch": 0.9146136189747514,
+      "grad_norm": 2.1208798515694767,
+      "learning_rate": 1.1858913156267694e-05,
+      "loss": 0.638,
+      "step": 5977
+    },
+    {
+      "epoch": 0.9147666411629687,
+      "grad_norm": 2.186872985982911,
+      "learning_rate": 1.1856478110116613e-05,
+      "loss": 0.6302,
+      "step": 5978
+    },
+    {
+      "epoch": 0.9149196633511859,
+      "grad_norm": 2.153919248477717,
+      "learning_rate": 1.1854042949951958e-05,
+      "loss": 0.7054,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9150726855394032,
+      "grad_norm": 2.397388197874348,
+      "learning_rate": 1.1851607675923269e-05,
+      "loss": 0.7276,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9152257077276205,
+      "grad_norm": 2.0648750992816165,
+      "learning_rate": 1.1849172288180106e-05,
+      "loss": 0.6745,
+      "step": 5981
+    },
+    {
+      "epoch": 0.9153787299158378,
+      "grad_norm": 2.067192519982237,
+      "learning_rate": 1.1846736786872042e-05,
+      "loss": 0.6398,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9155317521040551,
+      "grad_norm": 2.045318240934085,
+      "learning_rate": 1.1844301172148649e-05,
+      "loss": 0.6044,
+      "step": 5983
+    },
+    {
+      "epoch": 0.9156847742922724,
+      "grad_norm": 2.1862827497217547,
+      "learning_rate": 1.1841865444159502e-05,
+      "loss": 0.69,
+      "step": 5984
+    },
+    {
+      "epoch": 0.9158377964804897,
+      "grad_norm": 2.411569778709733,
+      "learning_rate": 1.1839429603054195e-05,
+      "loss": 0.6993,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9159908186687069,
+      "grad_norm": 2.163870964236157,
+      "learning_rate": 1.183699364898232e-05,
+      "loss": 0.7071,
+      "step": 5986
+    },
+    {
+      "epoch": 0.9161438408569242,
+      "grad_norm": 2.2216540837646543,
+      "learning_rate": 1.1834557582093478e-05,
+      "loss": 0.7181,
+      "step": 5987
+    },
+    {
+      "epoch": 0.9162968630451416,
+      "grad_norm": 2.4990352433843683,
+      "learning_rate": 1.183212140253728e-05,
+      "loss": 0.76,
+      "step": 5988
+    },
+    {
+      "epoch": 0.9164498852333588,
+      "grad_norm": 2.4982223206565495,
+      "learning_rate": 1.1829685110463339e-05,
+      "loss": 0.7631,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9166029074215761,
+      "grad_norm": 2.5748308659023333,
+      "learning_rate": 1.1827248706021274e-05,
+      "loss": 0.6416,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9167559296097935,
+      "grad_norm": 2.3410439692766816,
+      "learning_rate": 1.182481218936072e-05,
+      "loss": 0.6934,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9169089517980107,
+      "grad_norm": 2.55217646983151,
+      "learning_rate": 1.1822375560631311e-05,
+      "loss": 0.6742,
+      "step": 5992
+    },
+    {
+      "epoch": 0.917061973986228,
+      "grad_norm": 2.460463784312502,
+      "learning_rate": 1.181993881998269e-05,
+      "loss": 0.7407,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9172149961744452,
+      "grad_norm": 2.094408911904513,
+      "learning_rate": 1.1817501967564503e-05,
+      "loss": 0.5756,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9173680183626626,
+      "grad_norm": 2.3667932565329397,
+      "learning_rate": 1.1815065003526417e-05,
+      "loss": 0.6282,
+      "step": 5995
+    },
+    {
+      "epoch": 0.9175210405508799,
+      "grad_norm": 2.259307626828805,
+      "learning_rate": 1.1812627928018086e-05,
+      "loss": 0.6107,
+      "step": 5996
+    },
+    {
+      "epoch": 0.9176740627390971,
+      "grad_norm": 2.6719058170793133,
+      "learning_rate": 1.1810190741189177e-05,
+      "loss": 0.7363,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9178270849273145,
+      "grad_norm": 2.253953083863052,
+      "learning_rate": 1.180775344318938e-05,
+      "loss": 0.6705,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9179801071155318,
+      "grad_norm": 2.2485300966189326,
+      "learning_rate": 1.1805316034168373e-05,
+      "loss": 0.6667,
+      "step": 5999
+    },
+    {
+      "epoch": 0.918133129303749,
+      "grad_norm": 2.1631800211031407,
+      "learning_rate": 1.1802878514275839e-05,
+      "loss": 0.6333,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9182861514919664,
+      "grad_norm": 2.3348928082724445,
+      "learning_rate": 1.1800440883661485e-05,
+      "loss": 0.7369,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9184391736801836,
+      "grad_norm": 2.592997532053126,
+      "learning_rate": 1.1798003142475017e-05,
+      "loss": 0.6496,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9185921958684009,
+      "grad_norm": 2.2719886310328548,
+      "learning_rate": 1.1795565290866137e-05,
+      "loss": 0.659,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9187452180566182,
+      "grad_norm": 2.1908026163259957,
+      "learning_rate": 1.1793127328984572e-05,
+      "loss": 0.6543,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9188982402448355,
+      "grad_norm": 2.2403925541148357,
+      "learning_rate": 1.1790689256980042e-05,
+      "loss": 0.7169,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9190512624330528,
+      "grad_norm": 2.1158419300219986,
+      "learning_rate": 1.1788251075002277e-05,
+      "loss": 0.6643,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9192042846212701,
+      "grad_norm": 2.256627339715977,
+      "learning_rate": 1.178581278320102e-05,
+      "loss": 0.6332,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9193573068094874,
+      "grad_norm": 2.2967103658681163,
+      "learning_rate": 1.1783374381726016e-05,
+      "loss": 0.6572,
+      "step": 6008
+    },
+    {
+      "epoch": 0.9195103289977047,
+      "grad_norm": 2.3176786154668316,
+      "learning_rate": 1.178093587072701e-05,
+      "loss": 0.6278,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9196633511859219,
+      "grad_norm": 2.276335975164193,
+      "learning_rate": 1.1778497250353767e-05,
+      "loss": 0.7025,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9198163733741392,
+      "grad_norm": 2.198538269825206,
+      "learning_rate": 1.1776058520756053e-05,
+      "loss": 0.708,
+      "step": 6011
+    },
+    {
+      "epoch": 0.9199693955623566,
+      "grad_norm": 2.0093992303204287,
+      "learning_rate": 1.1773619682083632e-05,
+      "loss": 0.5572,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9201224177505738,
+      "grad_norm": 2.147072844426607,
+      "learning_rate": 1.1771180734486292e-05,
+      "loss": 0.6686,
+      "step": 6013
+    },
+    {
+      "epoch": 0.9202754399387911,
+      "grad_norm": 2.120096783175212,
+      "learning_rate": 1.1768741678113814e-05,
+      "loss": 0.692,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9204284621270085,
+      "grad_norm": 2.18494526069704,
+      "learning_rate": 1.1766302513115988e-05,
+      "loss": 0.6957,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9205814843152257,
+      "grad_norm": 2.183623529461029,
+      "learning_rate": 1.1763863239642617e-05,
+      "loss": 0.6199,
+      "step": 6016
+    },
+    {
+      "epoch": 0.920734506503443,
+      "grad_norm": 2.2086171453443346,
+      "learning_rate": 1.1761423857843504e-05,
+      "loss": 0.6994,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9208875286916602,
+      "grad_norm": 2.175576432459934,
+      "learning_rate": 1.175898436786846e-05,
+      "loss": 0.6868,
+      "step": 6018
+    },
+    {
+      "epoch": 0.9210405508798776,
+      "grad_norm": 2.18905913642669,
+      "learning_rate": 1.1756544769867304e-05,
+      "loss": 0.6596,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9211935730680949,
+      "grad_norm": 2.1552159728330467,
+      "learning_rate": 1.1754105063989865e-05,
+      "loss": 0.628,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9213465952563121,
+      "grad_norm": 2.352853585156651,
+      "learning_rate": 1.175166525038597e-05,
+      "loss": 0.7225,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9214996174445295,
+      "grad_norm": 2.2186163270481645,
+      "learning_rate": 1.1749225329205458e-05,
+      "loss": 0.5838,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9216526396327468,
+      "grad_norm": 2.316156419640685,
+      "learning_rate": 1.1746785300598178e-05,
+      "loss": 0.6525,
+      "step": 6023
+    },
+    {
+      "epoch": 0.921805661820964,
+      "grad_norm": 2.5308854754596917,
+      "learning_rate": 1.174434516471398e-05,
+      "loss": 0.6507,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9219586840091814,
+      "grad_norm": 2.038668645225144,
+      "learning_rate": 1.1741904921702718e-05,
+      "loss": 0.7139,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9221117061973986,
+      "grad_norm": 2.118866474655756,
+      "learning_rate": 1.1739464571714263e-05,
+      "loss": 0.6408,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9222647283856159,
+      "grad_norm": 2.4219096622082366,
+      "learning_rate": 1.1737024114898483e-05,
+      "loss": 0.9817,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9224177505738332,
+      "grad_norm": 2.3920546815691632,
+      "learning_rate": 1.1734583551405257e-05,
+      "loss": 0.731,
+      "step": 6028
+    },
+    {
+      "epoch": 0.9225707727620505,
+      "grad_norm": 2.2898608411847734,
+      "learning_rate": 1.1732142881384469e-05,
+      "loss": 0.7989,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9227237949502678,
+      "grad_norm": 2.275071139556073,
+      "learning_rate": 1.1729702104986011e-05,
+      "loss": 0.6128,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9228768171384851,
+      "grad_norm": 2.2660964406470048,
+      "learning_rate": 1.1727261222359781e-05,
+      "loss": 0.7475,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9230298393267024,
+      "grad_norm": 1.9638194324248897,
+      "learning_rate": 1.172482023365568e-05,
+      "loss": 0.6469,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9231828615149197,
+      "grad_norm": 2.061803171551883,
+      "learning_rate": 1.1722379139023623e-05,
+      "loss": 0.7059,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9233358837031369,
+      "grad_norm": 2.2447062635254897,
+      "learning_rate": 1.1719937938613523e-05,
+      "loss": 0.6911,
+      "step": 6034
+    },
+    {
+      "epoch": 0.9234889058913542,
+      "grad_norm": 2.1651544199285295,
+      "learning_rate": 1.1717496632575304e-05,
+      "loss": 0.5678,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9236419280795716,
+      "grad_norm": 2.4604106984706773,
+      "learning_rate": 1.1715055221058903e-05,
+      "loss": 0.7333,
+      "step": 6036
+    },
+    {
+      "epoch": 0.9237949502677888,
+      "grad_norm": 2.0659823412706104,
+      "learning_rate": 1.171261370421425e-05,
+      "loss": 0.5865,
+      "step": 6037
+    },
+    {
+      "epoch": 0.9239479724560061,
+      "grad_norm": 2.2467918440005628,
+      "learning_rate": 1.1710172082191288e-05,
+      "loss": 0.6591,
+      "step": 6038
+    },
+    {
+      "epoch": 0.9241009946442235,
+      "grad_norm": 2.1449493890857787,
+      "learning_rate": 1.170773035513997e-05,
+      "loss": 0.7061,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9242540168324407,
+      "grad_norm": 2.2560181924361755,
+      "learning_rate": 1.1705288523210253e-05,
+      "loss": 0.6814,
+      "step": 6040
+    },
+    {
+      "epoch": 0.924407039020658,
+      "grad_norm": 2.141905549714062,
+      "learning_rate": 1.1702846586552088e-05,
+      "loss": 0.6385,
+      "step": 6041
+    },
+    {
+      "epoch": 0.9245600612088752,
+      "grad_norm": 2.2548252680890135,
+      "learning_rate": 1.1700404545315458e-05,
+      "loss": 0.6793,
+      "step": 6042
+    },
+    {
+      "epoch": 0.9247130833970926,
+      "grad_norm": 2.482889662318577,
+      "learning_rate": 1.1697962399650333e-05,
+      "loss": 0.7582,
+      "step": 6043
+    },
+    {
+      "epoch": 0.9248661055853099,
+      "grad_norm": 2.5506460166887264,
+      "learning_rate": 1.1695520149706693e-05,
+      "loss": 0.7141,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9250191277735271,
+      "grad_norm": 2.145699244583287,
+      "learning_rate": 1.1693077795634531e-05,
+      "loss": 0.6328,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9251721499617445,
+      "grad_norm": 2.4752175224389577,
+      "learning_rate": 1.1690635337583836e-05,
+      "loss": 0.6299,
+      "step": 6046
+    },
+    {
+      "epoch": 0.9253251721499618,
+      "grad_norm": 2.2683649200809675,
+      "learning_rate": 1.168819277570461e-05,
+      "loss": 0.6869,
+      "step": 6047
+    },
+    {
+      "epoch": 0.925478194338179,
+      "grad_norm": 2.2004469435637457,
+      "learning_rate": 1.168575011014686e-05,
+      "loss": 0.6129,
+      "step": 6048
+    },
+    {
+      "epoch": 0.9256312165263963,
+      "grad_norm": 2.4371132806076994,
+      "learning_rate": 1.1683307341060608e-05,
+      "loss": 0.7319,
+      "step": 6049
+    },
+    {
+      "epoch": 0.9257842387146136,
+      "grad_norm": 2.2224151080045402,
+      "learning_rate": 1.168086446859586e-05,
+      "loss": 0.6627,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9259372609028309,
+      "grad_norm": 2.0430737085760287,
+      "learning_rate": 1.1678421492902652e-05,
+      "loss": 0.6371,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9260902830910482,
+      "grad_norm": 2.253282435206617,
+      "learning_rate": 1.1675978414131013e-05,
+      "loss": 0.6545,
+      "step": 6052
+    },
+    {
+      "epoch": 0.9262433052792655,
+      "grad_norm": 2.2800025754564257,
+      "learning_rate": 1.1673535232430986e-05,
+      "loss": 0.6834,
+      "step": 6053
+    },
+    {
+      "epoch": 0.9263963274674828,
+      "grad_norm": 2.107954815478626,
+      "learning_rate": 1.1671091947952611e-05,
+      "loss": 0.5648,
+      "step": 6054
+    },
+    {
+      "epoch": 0.9265493496557001,
+      "grad_norm": 2.4385308339899248,
+      "learning_rate": 1.1668648560845944e-05,
+      "loss": 0.6888,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9267023718439173,
+      "grad_norm": 2.2467214330301633,
+      "learning_rate": 1.1666205071261038e-05,
+      "loss": 0.7371,
+      "step": 6056
+    },
+    {
+      "epoch": 0.9268553940321347,
+      "grad_norm": 2.507942422969182,
+      "learning_rate": 1.1663761479347964e-05,
+      "loss": 0.7353,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9270084162203519,
+      "grad_norm": 2.196286306638163,
+      "learning_rate": 1.1661317785256787e-05,
+      "loss": 0.717,
+      "step": 6058
+    },
+    {
+      "epoch": 0.9271614384085692,
+      "grad_norm": 2.2098159183285606,
+      "learning_rate": 1.1658873989137585e-05,
+      "loss": 0.7388,
+      "step": 6059
+    },
+    {
+      "epoch": 0.9273144605967866,
+      "grad_norm": 2.539805991550103,
+      "learning_rate": 1.165643009114044e-05,
+      "loss": 0.7721,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9274674827850038,
+      "grad_norm": 2.6044090069624524,
+      "learning_rate": 1.165398609141545e-05,
+      "loss": 0.6935,
+      "step": 6061
+    },
+    {
+      "epoch": 0.9276205049732211,
+      "grad_norm": 2.384150928114248,
+      "learning_rate": 1.1651541990112698e-05,
+      "loss": 0.7076,
+      "step": 6062
+    },
+    {
+      "epoch": 0.9277735271614385,
+      "grad_norm": 2.4976738478057343,
+      "learning_rate": 1.164909778738229e-05,
+      "loss": 0.7126,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9279265493496557,
+      "grad_norm": 2.2947249137280923,
+      "learning_rate": 1.164665348337434e-05,
+      "loss": 0.6781,
+      "step": 6064
+    },
+    {
+      "epoch": 0.928079571537873,
+      "grad_norm": 2.3117604725656293,
+      "learning_rate": 1.164420907823896e-05,
+      "loss": 0.6601,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9282325937260902,
+      "grad_norm": 2.1388291258950605,
+      "learning_rate": 1.1641764572126262e-05,
+      "loss": 0.66,
+      "step": 6066
+    },
+    {
+      "epoch": 0.9283856159143076,
+      "grad_norm": 2.2545193918675506,
+      "learning_rate": 1.1639319965186382e-05,
+      "loss": 0.6792,
+      "step": 6067
+    },
+    {
+      "epoch": 0.9285386381025249,
+      "grad_norm": 2.2752674938211506,
+      "learning_rate": 1.1636875257569448e-05,
+      "loss": 0.6095,
+      "step": 6068
+    },
+    {
+      "epoch": 0.9286916602907421,
+      "grad_norm": 2.2575997598142723,
+      "learning_rate": 1.1634430449425604e-05,
+      "loss": 0.6722,
+      "step": 6069
+    },
+    {
+      "epoch": 0.9288446824789595,
+      "grad_norm": 2.2098103933749904,
+      "learning_rate": 1.1631985540904991e-05,
+      "loss": 0.6803,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9289977046671768,
+      "grad_norm": 2.294225796155611,
+      "learning_rate": 1.1629540532157758e-05,
+      "loss": 0.7256,
+      "step": 6071
+    },
+    {
+      "epoch": 0.929150726855394,
+      "grad_norm": 2.458554414220298,
+      "learning_rate": 1.1627095423334072e-05,
+      "loss": 0.802,
+      "step": 6072
+    },
+    {
+      "epoch": 0.9293037490436113,
+      "grad_norm": 2.3884126063112157,
+      "learning_rate": 1.1624650214584088e-05,
+      "loss": 0.7631,
+      "step": 6073
+    },
+    {
+      "epoch": 0.9294567712318286,
+      "grad_norm": 2.1963002613069618,
+      "learning_rate": 1.1622204906057979e-05,
+      "loss": 0.6559,
+      "step": 6074
+    },
+    {
+      "epoch": 0.9296097934200459,
+      "grad_norm": 2.3884546744501933,
+      "learning_rate": 1.161975949790592e-05,
+      "loss": 0.6752,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9297628156082632,
+      "grad_norm": 2.2694496861180737,
+      "learning_rate": 1.1617313990278093e-05,
+      "loss": 0.6918,
+      "step": 6076
+    },
+    {
+      "epoch": 0.9299158377964805,
+      "grad_norm": 2.3172171693412937,
+      "learning_rate": 1.161486838332469e-05,
+      "loss": 0.6784,
+      "step": 6077
+    },
+    {
+      "epoch": 0.9300688599846978,
+      "grad_norm": 2.1329383279977665,
+      "learning_rate": 1.16124226771959e-05,
+      "loss": 0.5508,
+      "step": 6078
+    },
+    {
+      "epoch": 0.9302218821729151,
+      "grad_norm": 2.293796679557291,
+      "learning_rate": 1.1609976872041924e-05,
+      "loss": 0.658,
+      "step": 6079
+    },
+    {
+      "epoch": 0.9303749043611323,
+      "grad_norm": 2.210358770465247,
+      "learning_rate": 1.1607530968012971e-05,
+      "loss": 0.7028,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9305279265493497,
+      "grad_norm": 2.213691145452231,
+      "learning_rate": 1.1605084965259256e-05,
+      "loss": 0.615,
+      "step": 6081
+    },
+    {
+      "epoch": 0.9306809487375669,
+      "grad_norm": 2.4649032540344344,
+      "learning_rate": 1.160263886393099e-05,
+      "loss": 0.7268,
+      "step": 6082
+    },
+    {
+      "epoch": 0.9308339709257842,
+      "grad_norm": 2.017207465174122,
+      "learning_rate": 1.1600192664178405e-05,
+      "loss": 0.6671,
+      "step": 6083
+    },
+    {
+      "epoch": 0.9309869931140016,
+      "grad_norm": 1.9259330149007499,
+      "learning_rate": 1.1597746366151725e-05,
+      "loss": 0.5948,
+      "step": 6084
+    },
+    {
+      "epoch": 0.9311400153022188,
+      "grad_norm": 2.167872374879758,
+      "learning_rate": 1.159529997000119e-05,
+      "loss": 0.7672,
+      "step": 6085
+    },
+    {
+      "epoch": 0.9312930374904361,
+      "grad_norm": 2.551354194085087,
+      "learning_rate": 1.1592853475877049e-05,
+      "loss": 0.7649,
+      "step": 6086
+    },
+    {
+      "epoch": 0.9314460596786535,
+      "grad_norm": 2.3450320571655348,
+      "learning_rate": 1.159040688392954e-05,
+      "loss": 0.6592,
+      "step": 6087
+    },
+    {
+      "epoch": 0.9315990818668707,
+      "grad_norm": 2.413844869725996,
+      "learning_rate": 1.158796019430892e-05,
+      "loss": 0.7055,
+      "step": 6088
+    },
+    {
+      "epoch": 0.931752104055088,
+      "grad_norm": 2.0778539002805014,
+      "learning_rate": 1.1585513407165456e-05,
+      "loss": 0.578,
+      "step": 6089
+    },
+    {
+      "epoch": 0.9319051262433052,
+      "grad_norm": 2.286133113069323,
+      "learning_rate": 1.1583066522649413e-05,
+      "loss": 0.7993,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9320581484315226,
+      "grad_norm": 2.1857984709529985,
+      "learning_rate": 1.1580619540911058e-05,
+      "loss": 0.7388,
+      "step": 6091
+    },
+    {
+      "epoch": 0.9322111706197399,
+      "grad_norm": 2.4184829856903547,
+      "learning_rate": 1.1578172462100676e-05,
+      "loss": 0.6769,
+      "step": 6092
+    },
+    {
+      "epoch": 0.9323641928079571,
+      "grad_norm": 2.2057962215002744,
+      "learning_rate": 1.157572528636855e-05,
+      "loss": 0.6771,
+      "step": 6093
+    },
+    {
+      "epoch": 0.9325172149961745,
+      "grad_norm": 1.9624533908826989,
+      "learning_rate": 1.1573278013864968e-05,
+      "loss": 0.5544,
+      "step": 6094
+    },
+    {
+      "epoch": 0.9326702371843917,
+      "grad_norm": 2.12121986199544,
+      "learning_rate": 1.1570830644740227e-05,
+      "loss": 0.6088,
+      "step": 6095
+    },
+    {
+      "epoch": 0.932823259372609,
+      "grad_norm": 2.2944618881122474,
+      "learning_rate": 1.1568383179144634e-05,
+      "loss": 0.7788,
+      "step": 6096
+    },
+    {
+      "epoch": 0.9329762815608263,
+      "grad_norm": 2.2593200428155256,
+      "learning_rate": 1.1565935617228492e-05,
+      "loss": 0.5957,
+      "step": 6097
+    },
+    {
+      "epoch": 0.9331293037490436,
+      "grad_norm": 2.252447232896904,
+      "learning_rate": 1.1563487959142121e-05,
+      "loss": 0.6521,
+      "step": 6098
+    },
+    {
+      "epoch": 0.9332823259372609,
+      "grad_norm": 2.5056513509402656,
+      "learning_rate": 1.1561040205035835e-05,
+      "loss": 0.8213,
+      "step": 6099
+    },
+    {
+      "epoch": 0.9334353481254782,
+      "grad_norm": 2.247790538197072,
+      "learning_rate": 1.1558592355059965e-05,
+      "loss": 0.6064,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9335883703136955,
+      "grad_norm": 2.1827896418062407,
+      "learning_rate": 1.1556144409364838e-05,
+      "loss": 0.6262,
+      "step": 6101
+    },
+    {
+      "epoch": 0.9337413925019128,
+      "grad_norm": 2.5213596631860202,
+      "learning_rate": 1.15536963681008e-05,
+      "loss": 0.714,
+      "step": 6102
+    },
+    {
+      "epoch": 0.93389441469013,
+      "grad_norm": 2.139941754037212,
+      "learning_rate": 1.1551248231418188e-05,
+      "loss": 0.6674,
+      "step": 6103
+    },
+    {
+      "epoch": 0.9340474368783473,
+      "grad_norm": 2.2803009133909513,
+      "learning_rate": 1.154879999946735e-05,
+      "loss": 0.6406,
+      "step": 6104
+    },
+    {
+      "epoch": 0.9342004590665647,
+      "grad_norm": 2.317103896037347,
+      "learning_rate": 1.154635167239865e-05,
+      "loss": 0.7685,
+      "step": 6105
+    },
+    {
+      "epoch": 0.9343534812547819,
+      "grad_norm": 2.1500102693630345,
+      "learning_rate": 1.1543903250362446e-05,
+      "loss": 0.6668,
+      "step": 6106
+    },
+    {
+      "epoch": 0.9345065034429992,
+      "grad_norm": 2.132728448753486,
+      "learning_rate": 1.1541454733509096e-05,
+      "loss": 0.6863,
+      "step": 6107
+    },
+    {
+      "epoch": 0.9346595256312166,
+      "grad_norm": 2.208420311449703,
+      "learning_rate": 1.1539006121988984e-05,
+      "loss": 0.751,
+      "step": 6108
+    },
+    {
+      "epoch": 0.9348125478194338,
+      "grad_norm": 2.24257468362804,
+      "learning_rate": 1.1536557415952488e-05,
+      "loss": 0.6207,
+      "step": 6109
+    },
+    {
+      "epoch": 0.9349655700076511,
+      "grad_norm": 2.504464488973117,
+      "learning_rate": 1.1534108615549988e-05,
+      "loss": 0.7274,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9351185921958683,
+      "grad_norm": 2.004035147085841,
+      "learning_rate": 1.1531659720931877e-05,
+      "loss": 0.6282,
+      "step": 6111
+    },
+    {
+      "epoch": 0.9352716143840857,
+      "grad_norm": 2.095751079721034,
+      "learning_rate": 1.152921073224855e-05,
+      "loss": 0.587,
+      "step": 6112
+    },
+    {
+      "epoch": 0.935424636572303,
+      "grad_norm": 2.0794742962572843,
+      "learning_rate": 1.152676164965041e-05,
+      "loss": 0.6929,
+      "step": 6113
+    },
+    {
+      "epoch": 0.9355776587605202,
+      "grad_norm": 2.2462164854122606,
+      "learning_rate": 1.1524312473287866e-05,
+      "loss": 0.6546,
+      "step": 6114
+    },
+    {
+      "epoch": 0.9357306809487376,
+      "grad_norm": 2.2331729878802467,
+      "learning_rate": 1.1521863203311327e-05,
+      "loss": 0.7246,
+      "step": 6115
+    },
+    {
+      "epoch": 0.9358837031369549,
+      "grad_norm": 2.2690525367182928,
+      "learning_rate": 1.1519413839871218e-05,
+      "loss": 0.7247,
+      "step": 6116
+    },
+    {
+      "epoch": 0.9360367253251721,
+      "grad_norm": 2.009458991477944,
+      "learning_rate": 1.1516964383117957e-05,
+      "loss": 0.6288,
+      "step": 6117
+    },
+    {
+      "epoch": 0.9361897475133895,
+      "grad_norm": 2.456698580139736,
+      "learning_rate": 1.1514514833201981e-05,
+      "loss": 0.7207,
+      "step": 6118
+    },
+    {
+      "epoch": 0.9363427697016067,
+      "grad_norm": 1.9635176193390491,
+      "learning_rate": 1.1512065190273728e-05,
+      "loss": 0.6222,
+      "step": 6119
+    },
+    {
+      "epoch": 0.936495791889824,
+      "grad_norm": 2.2822880914918287,
+      "learning_rate": 1.150961545448363e-05,
+      "loss": 0.697,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9366488140780413,
+      "grad_norm": 2.1295174618645745,
+      "learning_rate": 1.1507165625982144e-05,
+      "loss": 0.6507,
+      "step": 6121
+    },
+    {
+      "epoch": 0.9368018362662586,
+      "grad_norm": 2.1926894823041376,
+      "learning_rate": 1.1504715704919722e-05,
+      "loss": 0.6582,
+      "step": 6122
+    },
+    {
+      "epoch": 0.9369548584544759,
+      "grad_norm": 2.5892530861553373,
+      "learning_rate": 1.1502265691446821e-05,
+      "loss": 0.6738,
+      "step": 6123
+    },
+    {
+      "epoch": 0.9371078806426932,
+      "grad_norm": 2.068185855054658,
+      "learning_rate": 1.1499815585713909e-05,
+      "loss": 0.6418,
+      "step": 6124
+    },
+    {
+      "epoch": 0.9372609028309105,
+      "grad_norm": 2.2803267411959394,
+      "learning_rate": 1.1497365387871454e-05,
+      "loss": 0.5819,
+      "step": 6125
+    },
+    {
+      "epoch": 0.9374139250191278,
+      "grad_norm": 2.1349957197936713,
+      "learning_rate": 1.1494915098069927e-05,
+      "loss": 0.5879,
+      "step": 6126
+    },
+    {
+      "epoch": 0.937566947207345,
+      "grad_norm": 2.5188656753058507,
+      "learning_rate": 1.1492464716459824e-05,
+      "loss": 0.6856,
+      "step": 6127
+    },
+    {
+      "epoch": 0.9377199693955623,
+      "grad_norm": 2.0389553034383265,
+      "learning_rate": 1.1490014243191621e-05,
+      "loss": 0.5911,
+      "step": 6128
+    },
+    {
+      "epoch": 0.9378729915837797,
+      "grad_norm": 2.356383652380313,
+      "learning_rate": 1.1487563678415812e-05,
+      "loss": 0.7431,
+      "step": 6129
+    },
+    {
+      "epoch": 0.9380260137719969,
+      "grad_norm": 2.1593952450928033,
+      "learning_rate": 1.14851130222829e-05,
+      "loss": 0.6465,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9381790359602142,
+      "grad_norm": 2.4550090735416306,
+      "learning_rate": 1.148266227494339e-05,
+      "loss": 0.7766,
+      "step": 6131
+    },
+    {
+      "epoch": 0.9383320581484316,
+      "grad_norm": 2.4549971534102175,
+      "learning_rate": 1.1480211436547783e-05,
+      "loss": 0.68,
+      "step": 6132
+    },
+    {
+      "epoch": 0.9384850803366488,
+      "grad_norm": 2.3471098658149447,
+      "learning_rate": 1.1477760507246606e-05,
+      "loss": 0.6494,
+      "step": 6133
+    },
+    {
+      "epoch": 0.9386381025248661,
+      "grad_norm": 2.173375519361347,
+      "learning_rate": 1.1475309487190376e-05,
+      "loss": 0.6001,
+      "step": 6134
+    },
+    {
+      "epoch": 0.9387911247130833,
+      "grad_norm": 2.5263042796935973,
+      "learning_rate": 1.1472858376529616e-05,
+      "loss": 0.6864,
+      "step": 6135
+    },
+    {
+      "epoch": 0.9389441469013007,
+      "grad_norm": 2.177731124432879,
+      "learning_rate": 1.1470407175414864e-05,
+      "loss": 0.6324,
+      "step": 6136
+    },
+    {
+      "epoch": 0.939097169089518,
+      "grad_norm": 2.2483120867164,
+      "learning_rate": 1.1467955883996653e-05,
+      "loss": 0.7105,
+      "step": 6137
+    },
+    {
+      "epoch": 0.9392501912777352,
+      "grad_norm": 2.3912382526507656,
+      "learning_rate": 1.146550450242553e-05,
+      "loss": 0.6568,
+      "step": 6138
+    },
+    {
+      "epoch": 0.9394032134659526,
+      "grad_norm": 2.411089906961605,
+      "learning_rate": 1.146305303085204e-05,
+      "loss": 0.6886,
+      "step": 6139
+    },
+    {
+      "epoch": 0.9395562356541699,
+      "grad_norm": 2.2962550676687505,
+      "learning_rate": 1.1460601469426741e-05,
+      "loss": 0.7108,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9397092578423871,
+      "grad_norm": 2.2516999284059,
+      "learning_rate": 1.1458149818300192e-05,
+      "loss": 0.6853,
+      "step": 6141
+    },
+    {
+      "epoch": 0.9398622800306045,
+      "grad_norm": 2.3487910495210094,
+      "learning_rate": 1.1455698077622959e-05,
+      "loss": 0.6304,
+      "step": 6142
+    },
+    {
+      "epoch": 0.9400153022188217,
+      "grad_norm": 2.3253614074896682,
+      "learning_rate": 1.1453246247545612e-05,
+      "loss": 0.7826,
+      "step": 6143
+    },
+    {
+      "epoch": 0.940168324407039,
+      "grad_norm": 2.035455850686615,
+      "learning_rate": 1.145079432821873e-05,
+      "loss": 0.6026,
+      "step": 6144
+    },
+    {
+      "epoch": 0.9403213465952563,
+      "grad_norm": 2.0392874454271714,
+      "learning_rate": 1.1448342319792886e-05,
+      "loss": 0.7886,
+      "step": 6145
+    },
+    {
+      "epoch": 0.9404743687834736,
+      "grad_norm": 2.301481523837578,
+      "learning_rate": 1.1445890222418681e-05,
+      "loss": 0.6503,
+      "step": 6146
+    },
+    {
+      "epoch": 0.9406273909716909,
+      "grad_norm": 2.307110238176336,
+      "learning_rate": 1.1443438036246702e-05,
+      "loss": 0.7039,
+      "step": 6147
+    },
+    {
+      "epoch": 0.9407804131599082,
+      "grad_norm": 2.355842387023983,
+      "learning_rate": 1.144098576142754e-05,
+      "loss": 0.7409,
+      "step": 6148
+    },
+    {
+      "epoch": 0.9409334353481255,
+      "grad_norm": 2.5326182737200673,
+      "learning_rate": 1.1438533398111808e-05,
+      "loss": 0.8251,
+      "step": 6149
+    },
+    {
+      "epoch": 0.9410864575363428,
+      "grad_norm": 2.227199426196083,
+      "learning_rate": 1.1436080946450115e-05,
+      "loss": 0.6085,
+      "step": 6150
+    },
+    {
+      "epoch": 0.94123947972456,
+      "grad_norm": 2.2345241158679165,
+      "learning_rate": 1.1433628406593069e-05,
+      "loss": 0.6675,
+      "step": 6151
+    },
+    {
+      "epoch": 0.9413925019127773,
+      "grad_norm": 1.8715859371306767,
+      "learning_rate": 1.1431175778691297e-05,
+      "loss": 0.5941,
+      "step": 6152
+    },
+    {
+      "epoch": 0.9415455241009947,
+      "grad_norm": 2.2216163917317675,
+      "learning_rate": 1.1428723062895421e-05,
+      "loss": 0.6047,
+      "step": 6153
+    },
+    {
+      "epoch": 0.9416985462892119,
+      "grad_norm": 1.9294546956579186,
+      "learning_rate": 1.1426270259356071e-05,
+      "loss": 0.5409,
+      "step": 6154
+    },
+    {
+      "epoch": 0.9418515684774292,
+      "grad_norm": 2.1209551146155405,
+      "learning_rate": 1.1423817368223886e-05,
+      "loss": 0.6277,
+      "step": 6155
+    },
+    {
+      "epoch": 0.9420045906656466,
+      "grad_norm": 2.192111802183402,
+      "learning_rate": 1.1421364389649508e-05,
+      "loss": 0.5492,
+      "step": 6156
+    },
+    {
+      "epoch": 0.9421576128538638,
+      "grad_norm": 2.19448860446009,
+      "learning_rate": 1.141891132378358e-05,
+      "loss": 0.6536,
+      "step": 6157
+    },
+    {
+      "epoch": 0.9423106350420811,
+      "grad_norm": 2.071398629519424,
+      "learning_rate": 1.1416458170776757e-05,
+      "loss": 0.5741,
+      "step": 6158
+    },
+    {
+      "epoch": 0.9424636572302983,
+      "grad_norm": 2.138162941124051,
+      "learning_rate": 1.1414004930779697e-05,
+      "loss": 0.642,
+      "step": 6159
+    },
+    {
+      "epoch": 0.9426166794185157,
+      "grad_norm": 2.0968607838531805,
+      "learning_rate": 1.1411551603943065e-05,
+      "loss": 0.6482,
+      "step": 6160
+    },
+    {
+      "epoch": 0.942769701606733,
+      "grad_norm": 2.2631243499184577,
+      "learning_rate": 1.1409098190417523e-05,
+      "loss": 0.5649,
+      "step": 6161
+    },
+    {
+      "epoch": 0.9429227237949502,
+      "grad_norm": 2.2003646678049993,
+      "learning_rate": 1.1406644690353752e-05,
+      "loss": 0.6781,
+      "step": 6162
+    },
+    {
+      "epoch": 0.9430757459831676,
+      "grad_norm": 2.2561519198823246,
+      "learning_rate": 1.1404191103902425e-05,
+      "loss": 0.57,
+      "step": 6163
+    },
+    {
+      "epoch": 0.9432287681713849,
+      "grad_norm": 2.362404452628459,
+      "learning_rate": 1.140173743121423e-05,
+      "loss": 0.6441,
+      "step": 6164
+    },
+    {
+      "epoch": 0.9433817903596021,
+      "grad_norm": 2.2230712076613153,
+      "learning_rate": 1.1399283672439856e-05,
+      "loss": 0.6328,
+      "step": 6165
+    },
+    {
+      "epoch": 0.9435348125478195,
+      "grad_norm": 2.156629373945652,
+      "learning_rate": 1.1396829827729998e-05,
+      "loss": 0.5572,
+      "step": 6166
+    },
+    {
+      "epoch": 0.9436878347360367,
+      "grad_norm": 2.184268144546899,
+      "learning_rate": 1.1394375897235354e-05,
+      "loss": 0.6756,
+      "step": 6167
+    },
+    {
+      "epoch": 0.943840856924254,
+      "grad_norm": 2.2888378781567225,
+      "learning_rate": 1.1391921881106636e-05,
+      "loss": 0.7744,
+      "step": 6168
+    },
+    {
+      "epoch": 0.9439938791124713,
+      "grad_norm": 2.2626375112314685,
+      "learning_rate": 1.1389467779494545e-05,
+      "loss": 0.672,
+      "step": 6169
+    },
+    {
+      "epoch": 0.9441469013006886,
+      "grad_norm": 2.1640029277804635,
+      "learning_rate": 1.1387013592549803e-05,
+      "loss": 0.625,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9442999234889059,
+      "grad_norm": 2.37356580152094,
+      "learning_rate": 1.1384559320423132e-05,
+      "loss": 0.7834,
+      "step": 6171
+    },
+    {
+      "epoch": 0.9444529456771232,
+      "grad_norm": 2.1480740870003228,
+      "learning_rate": 1.1382104963265256e-05,
+      "loss": 0.6343,
+      "step": 6172
+    },
+    {
+      "epoch": 0.9446059678653405,
+      "grad_norm": 2.189878831615825,
+      "learning_rate": 1.1379650521226907e-05,
+      "loss": 0.7639,
+      "step": 6173
+    },
+    {
+      "epoch": 0.9447589900535578,
+      "grad_norm": 2.2193507647046324,
+      "learning_rate": 1.1377195994458823e-05,
+      "loss": 0.6586,
+      "step": 6174
+    },
+    {
+      "epoch": 0.944912012241775,
+      "grad_norm": 2.169829742772359,
+      "learning_rate": 1.1374741383111745e-05,
+      "loss": 0.6261,
+      "step": 6175
+    },
+    {
+      "epoch": 0.9450650344299923,
+      "grad_norm": 2.5015537311138467,
+      "learning_rate": 1.137228668733642e-05,
+      "loss": 0.6167,
+      "step": 6176
+    },
+    {
+      "epoch": 0.9452180566182097,
+      "grad_norm": 2.225124391565102,
+      "learning_rate": 1.13698319072836e-05,
+      "loss": 0.6276,
+      "step": 6177
+    },
+    {
+      "epoch": 0.9453710788064269,
+      "grad_norm": 2.218346428912096,
+      "learning_rate": 1.136737704310405e-05,
+      "loss": 0.6192,
+      "step": 6178
+    },
+    {
+      "epoch": 0.9455241009946442,
+      "grad_norm": 2.1116689466249743,
+      "learning_rate": 1.1364922094948521e-05,
+      "loss": 0.6749,
+      "step": 6179
+    },
+    {
+      "epoch": 0.9456771231828616,
+      "grad_norm": 2.5467465934570375,
+      "learning_rate": 1.1362467062967785e-05,
+      "loss": 0.733,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9458301453710788,
+      "grad_norm": 2.159692074442581,
+      "learning_rate": 1.1360011947312622e-05,
+      "loss": 0.5801,
+      "step": 6181
+    },
+    {
+      "epoch": 0.9459831675592961,
+      "grad_norm": 2.2593107197857982,
+      "learning_rate": 1.13575567481338e-05,
+      "loss": 0.7481,
+      "step": 6182
+    },
+    {
+      "epoch": 0.9461361897475133,
+      "grad_norm": 2.7468321108243154,
+      "learning_rate": 1.1355101465582109e-05,
+      "loss": 0.6763,
+      "step": 6183
+    },
+    {
+      "epoch": 0.9462892119357307,
+      "grad_norm": 2.4697235771987374,
+      "learning_rate": 1.1352646099808338e-05,
+      "loss": 0.7542,
+      "step": 6184
+    },
+    {
+      "epoch": 0.946442234123948,
+      "grad_norm": 2.6351990269079355,
+      "learning_rate": 1.1350190650963278e-05,
+      "loss": 0.6509,
+      "step": 6185
+    },
+    {
+      "epoch": 0.9465952563121652,
+      "grad_norm": 2.0847983344763823,
+      "learning_rate": 1.1347735119197724e-05,
+      "loss": 0.6195,
+      "step": 6186
+    },
+    {
+      "epoch": 0.9467482785003826,
+      "grad_norm": 2.2628289082127613,
+      "learning_rate": 1.1345279504662488e-05,
+      "loss": 0.6789,
+      "step": 6187
+    },
+    {
+      "epoch": 0.9469013006885999,
+      "grad_norm": 2.3024156670048446,
+      "learning_rate": 1.1342823807508371e-05,
+      "loss": 0.6697,
+      "step": 6188
+    },
+    {
+      "epoch": 0.9470543228768171,
+      "grad_norm": 2.2869700353035065,
+      "learning_rate": 1.1340368027886195e-05,
+      "loss": 0.6546,
+      "step": 6189
+    },
+    {
+      "epoch": 0.9472073450650345,
+      "grad_norm": 2.1321757720571153,
+      "learning_rate": 1.1337912165946773e-05,
+      "loss": 0.607,
+      "step": 6190
+    },
+    {
+      "epoch": 0.9473603672532517,
+      "grad_norm": 2.0630142586951647,
+      "learning_rate": 1.1335456221840932e-05,
+      "loss": 0.6203,
+      "step": 6191
+    },
+    {
+      "epoch": 0.947513389441469,
+      "grad_norm": 2.1348173216555995,
+      "learning_rate": 1.1333000195719498e-05,
+      "loss": 0.6734,
+      "step": 6192
+    },
+    {
+      "epoch": 0.9476664116296863,
+      "grad_norm": 2.5299071976646723,
+      "learning_rate": 1.1330544087733311e-05,
+      "loss": 0.7018,
+      "step": 6193
+    },
+    {
+      "epoch": 0.9478194338179036,
+      "grad_norm": 2.1658288076024137,
+      "learning_rate": 1.1328087898033204e-05,
+      "loss": 0.7224,
+      "step": 6194
+    },
+    {
+      "epoch": 0.9479724560061209,
+      "grad_norm": 2.03617237782173,
+      "learning_rate": 1.1325631626770024e-05,
+      "loss": 0.6653,
+      "step": 6195
+    },
+    {
+      "epoch": 0.9481254781943382,
+      "grad_norm": 2.28947007165544,
+      "learning_rate": 1.1323175274094615e-05,
+      "loss": 0.7274,
+      "step": 6196
+    },
+    {
+      "epoch": 0.9482785003825555,
+      "grad_norm": 2.2920800270128603,
+      "learning_rate": 1.1320718840157844e-05,
+      "loss": 0.6127,
+      "step": 6197
+    },
+    {
+      "epoch": 0.9484315225707728,
+      "grad_norm": 2.2807534856633374,
+      "learning_rate": 1.131826232511056e-05,
+      "loss": 0.7514,
+      "step": 6198
+    },
+    {
+      "epoch": 0.94858454475899,
+      "grad_norm": 2.277274760847133,
+      "learning_rate": 1.1315805729103626e-05,
+      "loss": 0.7422,
+      "step": 6199
+    },
+    {
+      "epoch": 0.9487375669472073,
+      "grad_norm": 2.3026912902212056,
+      "learning_rate": 1.1313349052287918e-05,
+      "loss": 0.7148,
+      "step": 6200
+    },
+    {
+      "epoch": 0.9488905891354247,
+      "grad_norm": 2.27866912590338,
+      "learning_rate": 1.1310892294814308e-05,
+      "loss": 0.6766,
+      "step": 6201
+    },
+    {
+      "epoch": 0.9490436113236419,
+      "grad_norm": 2.199628352082259,
+      "learning_rate": 1.130843545683367e-05,
+      "loss": 0.6551,
+      "step": 6202
+    },
+    {
+      "epoch": 0.9491966335118592,
+      "grad_norm": 2.3123931122807866,
+      "learning_rate": 1.1305978538496896e-05,
+      "loss": 0.7475,
+      "step": 6203
+    },
+    {
+      "epoch": 0.9493496557000765,
+      "grad_norm": 2.2427418927664564,
+      "learning_rate": 1.130352153995487e-05,
+      "loss": 0.7086,
+      "step": 6204
+    },
+    {
+      "epoch": 0.9495026778882938,
+      "grad_norm": 2.131171988564861,
+      "learning_rate": 1.1301064461358484e-05,
+      "loss": 0.6465,
+      "step": 6205
+    },
+    {
+      "epoch": 0.9496557000765111,
+      "grad_norm": 2.029476964393686,
+      "learning_rate": 1.1298607302858643e-05,
+      "loss": 0.5969,
+      "step": 6206
+    },
+    {
+      "epoch": 0.9498087222647283,
+      "grad_norm": 2.352972110165139,
+      "learning_rate": 1.1296150064606244e-05,
+      "loss": 0.7222,
+      "step": 6207
+    },
+    {
+      "epoch": 0.9499617444529457,
+      "grad_norm": 2.0387321968883874,
+      "learning_rate": 1.1293692746752201e-05,
+      "loss": 0.5474,
+      "step": 6208
+    },
+    {
+      "epoch": 0.950114766641163,
+      "grad_norm": 2.1671336258881344,
+      "learning_rate": 1.1291235349447427e-05,
+      "loss": 0.6044,
+      "step": 6209
+    },
+    {
+      "epoch": 0.9502677888293802,
+      "grad_norm": 2.11725262790491,
+      "learning_rate": 1.1288777872842837e-05,
+      "loss": 0.5895,
+      "step": 6210
+    },
+    {
+      "epoch": 0.9504208110175976,
+      "grad_norm": 2.175716482784178,
+      "learning_rate": 1.1286320317089354e-05,
+      "loss": 0.673,
+      "step": 6211
+    },
+    {
+      "epoch": 0.9505738332058148,
+      "grad_norm": 2.4334045623113507,
+      "learning_rate": 1.1283862682337909e-05,
+      "loss": 0.742,
+      "step": 6212
+    },
+    {
+      "epoch": 0.9507268553940321,
+      "grad_norm": 2.2530770860240295,
+      "learning_rate": 1.128140496873944e-05,
+      "loss": 0.6588,
+      "step": 6213
+    },
+    {
+      "epoch": 0.9508798775822495,
+      "grad_norm": 2.2343257508068772,
+      "learning_rate": 1.1278947176444872e-05,
+      "loss": 0.7206,
+      "step": 6214
+    },
+    {
+      "epoch": 0.9510328997704667,
+      "grad_norm": 2.3492826056594915,
+      "learning_rate": 1.1276489305605157e-05,
+      "loss": 0.681,
+      "step": 6215
+    },
+    {
+      "epoch": 0.951185921958684,
+      "grad_norm": 2.249954446637355,
+      "learning_rate": 1.127403135637124e-05,
+      "loss": 0.7103,
+      "step": 6216
+    },
+    {
+      "epoch": 0.9513389441469013,
+      "grad_norm": 2.106521792271321,
+      "learning_rate": 1.1271573328894074e-05,
+      "loss": 0.64,
+      "step": 6217
+    },
+    {
+      "epoch": 0.9514919663351186,
+      "grad_norm": 1.899570026742498,
+      "learning_rate": 1.1269115223324615e-05,
+      "loss": 0.6032,
+      "step": 6218
+    },
+    {
+      "epoch": 0.9516449885233359,
+      "grad_norm": 2.279531297155367,
+      "learning_rate": 1.1266657039813826e-05,
+      "loss": 0.6868,
+      "step": 6219
+    },
+    {
+      "epoch": 0.9517980107115531,
+      "grad_norm": 2.2086467285134965,
+      "learning_rate": 1.1264198778512674e-05,
+      "loss": 0.697,
+      "step": 6220
+    },
+    {
+      "epoch": 0.9519510328997705,
+      "grad_norm": 2.0313857844134904,
+      "learning_rate": 1.1261740439572126e-05,
+      "loss": 0.6574,
+      "step": 6221
+    },
+    {
+      "epoch": 0.9521040550879878,
+      "grad_norm": 1.930961524376918,
+      "learning_rate": 1.1259282023143166e-05,
+      "loss": 0.6359,
+      "step": 6222
+    },
+    {
+      "epoch": 0.952257077276205,
+      "grad_norm": 2.2039196753961376,
+      "learning_rate": 1.1256823529376768e-05,
+      "loss": 0.6681,
+      "step": 6223
+    },
+    {
+      "epoch": 0.9524100994644223,
+      "grad_norm": 2.455580268575347,
+      "learning_rate": 1.1254364958423917e-05,
+      "loss": 0.787,
+      "step": 6224
+    },
+    {
+      "epoch": 0.9525631216526397,
+      "grad_norm": 2.2229067809145566,
+      "learning_rate": 1.1251906310435611e-05,
+      "loss": 0.6966,
+      "step": 6225
+    },
+    {
+      "epoch": 0.9527161438408569,
+      "grad_norm": 2.233353532493785,
+      "learning_rate": 1.1249447585562843e-05,
+      "loss": 0.6934,
+      "step": 6226
+    },
+    {
+      "epoch": 0.9528691660290742,
+      "grad_norm": 2.2308301822804255,
+      "learning_rate": 1.1246988783956606e-05,
+      "loss": 0.6645,
+      "step": 6227
+    },
+    {
+      "epoch": 0.9530221882172915,
+      "grad_norm": 2.6020370006509546,
+      "learning_rate": 1.124452990576791e-05,
+      "loss": 0.7857,
+      "step": 6228
+    },
+    {
+      "epoch": 0.9531752104055088,
+      "grad_norm": 2.408349890753668,
+      "learning_rate": 1.1242070951147767e-05,
+      "loss": 0.6858,
+      "step": 6229
+    },
+    {
+      "epoch": 0.9533282325937261,
+      "grad_norm": 2.2188435241964544,
+      "learning_rate": 1.1239611920247187e-05,
+      "loss": 0.7378,
+      "step": 6230
+    },
+    {
+      "epoch": 0.9534812547819433,
+      "grad_norm": 2.3730115015136746,
+      "learning_rate": 1.123715281321719e-05,
+      "loss": 0.6563,
+      "step": 6231
+    },
+    {
+      "epoch": 0.9536342769701607,
+      "grad_norm": 2.2447044972291827,
+      "learning_rate": 1.1234693630208798e-05,
+      "loss": 0.7419,
+      "step": 6232
+    },
+    {
+      "epoch": 0.953787299158378,
+      "grad_norm": 2.1414125465799687,
+      "learning_rate": 1.1232234371373041e-05,
+      "loss": 0.6364,
+      "step": 6233
+    },
+    {
+      "epoch": 0.9539403213465952,
+      "grad_norm": 2.3749105034101388,
+      "learning_rate": 1.1229775036860948e-05,
+      "loss": 0.7654,
+      "step": 6234
+    },
+    {
+      "epoch": 0.9540933435348126,
+      "grad_norm": 2.0829661403047433,
+      "learning_rate": 1.1227315626823562e-05,
+      "loss": 0.6018,
+      "step": 6235
+    },
+    {
+      "epoch": 0.9542463657230298,
+      "grad_norm": 2.621822847733121,
+      "learning_rate": 1.1224856141411921e-05,
+      "loss": 0.652,
+      "step": 6236
+    },
+    {
+      "epoch": 0.9543993879112471,
+      "grad_norm": 1.8273526527697626,
+      "learning_rate": 1.122239658077707e-05,
+      "loss": 0.6322,
+      "step": 6237
+    },
+    {
+      "epoch": 0.9545524100994645,
+      "grad_norm": 2.0097539648147817,
+      "learning_rate": 1.121993694507007e-05,
+      "loss": 0.5612,
+      "step": 6238
+    },
+    {
+      "epoch": 0.9547054322876817,
+      "grad_norm": 2.5389843690982983,
+      "learning_rate": 1.1217477234441965e-05,
+      "loss": 0.7735,
+      "step": 6239
+    },
+    {
+      "epoch": 0.954858454475899,
+      "grad_norm": 2.1787684248203347,
+      "learning_rate": 1.1215017449043817e-05,
+      "loss": 0.7613,
+      "step": 6240
+    },
+    {
+      "epoch": 0.9550114766641163,
+      "grad_norm": 2.2525442589068376,
+      "learning_rate": 1.1212557589026699e-05,
+      "loss": 0.6428,
+      "step": 6241
+    },
+    {
+      "epoch": 0.9551644988523336,
+      "grad_norm": 2.3005374778193457,
+      "learning_rate": 1.1210097654541676e-05,
+      "loss": 0.7743,
+      "step": 6242
+    },
+    {
+      "epoch": 0.9553175210405509,
+      "grad_norm": 2.2746574255072023,
+      "learning_rate": 1.120763764573982e-05,
+      "loss": 0.6967,
+      "step": 6243
+    },
+    {
+      "epoch": 0.9554705432287681,
+      "grad_norm": 2.378397266502046,
+      "learning_rate": 1.1205177562772212e-05,
+      "loss": 0.7118,
+      "step": 6244
+    },
+    {
+      "epoch": 0.9556235654169855,
+      "grad_norm": 1.9416013991409358,
+      "learning_rate": 1.1202717405789936e-05,
+      "loss": 0.4577,
+      "step": 6245
+    },
+    {
+      "epoch": 0.9557765876052028,
+      "grad_norm": 2.3703928068791456,
+      "learning_rate": 1.120025717494408e-05,
+      "loss": 0.669,
+      "step": 6246
+    },
+    {
+      "epoch": 0.95592960979342,
+      "grad_norm": 2.1911621815779734,
+      "learning_rate": 1.1197796870385732e-05,
+      "loss": 0.6874,
+      "step": 6247
+    },
+    {
+      "epoch": 0.9560826319816373,
+      "grad_norm": 2.23570911339143,
+      "learning_rate": 1.1195336492265997e-05,
+      "loss": 0.5486,
+      "step": 6248
+    },
+    {
+      "epoch": 0.9562356541698547,
+      "grad_norm": 2.1892871871968955,
+      "learning_rate": 1.1192876040735968e-05,
+      "loss": 0.6156,
+      "step": 6249
+    },
+    {
+      "epoch": 0.9563886763580719,
+      "grad_norm": 2.5372633282546535,
+      "learning_rate": 1.1190415515946757e-05,
+      "loss": 0.7427,
+      "step": 6250
+    },
+    {
+      "epoch": 0.9565416985462892,
+      "grad_norm": 2.1746459071347517,
+      "learning_rate": 1.1187954918049473e-05,
+      "loss": 0.6861,
+      "step": 6251
+    },
+    {
+      "epoch": 0.9566947207345065,
+      "grad_norm": 2.3915477100098266,
+      "learning_rate": 1.1185494247195227e-05,
+      "loss": 0.6661,
+      "step": 6252
+    },
+    {
+      "epoch": 0.9568477429227238,
+      "grad_norm": 2.301774197826807,
+      "learning_rate": 1.1183033503535144e-05,
+      "loss": 0.7023,
+      "step": 6253
+    },
+    {
+      "epoch": 0.9570007651109411,
+      "grad_norm": 2.0690146112814616,
+      "learning_rate": 1.1180572687220349e-05,
+      "loss": 0.6382,
+      "step": 6254
+    },
+    {
+      "epoch": 0.9571537872991583,
+      "grad_norm": 2.0792422582969805,
+      "learning_rate": 1.1178111798401959e-05,
+      "loss": 0.6192,
+      "step": 6255
+    },
+    {
+      "epoch": 0.9573068094873757,
+      "grad_norm": 2.332152072423957,
+      "learning_rate": 1.1175650837231118e-05,
+      "loss": 0.7097,
+      "step": 6256
+    },
+    {
+      "epoch": 0.957459831675593,
+      "grad_norm": 2.252289252328118,
+      "learning_rate": 1.1173189803858961e-05,
+      "loss": 0.69,
+      "step": 6257
+    },
+    {
+      "epoch": 0.9576128538638102,
+      "grad_norm": 2.149486560835919,
+      "learning_rate": 1.1170728698436629e-05,
+      "loss": 0.6495,
+      "step": 6258
+    },
+    {
+      "epoch": 0.9577658760520276,
+      "grad_norm": 2.4865796283824535,
+      "learning_rate": 1.1168267521115266e-05,
+      "loss": 0.6406,
+      "step": 6259
+    },
+    {
+      "epoch": 0.9579188982402448,
+      "grad_norm": 2.562998136403146,
+      "learning_rate": 1.1165806272046024e-05,
+      "loss": 0.7459,
+      "step": 6260
+    },
+    {
+      "epoch": 0.9580719204284621,
+      "grad_norm": 2.244474523121574,
+      "learning_rate": 1.116334495138006e-05,
+      "loss": 0.7253,
+      "step": 6261
+    },
+    {
+      "epoch": 0.9582249426166795,
+      "grad_norm": 2.3237989302201343,
+      "learning_rate": 1.116088355926853e-05,
+      "loss": 0.7225,
+      "step": 6262
+    },
+    {
+      "epoch": 0.9583779648048967,
+      "grad_norm": 2.296054403182734,
+      "learning_rate": 1.11584220958626e-05,
+      "loss": 0.6612,
+      "step": 6263
+    },
+    {
+      "epoch": 0.958530986993114,
+      "grad_norm": 2.1853141461542034,
+      "learning_rate": 1.1155960561313437e-05,
+      "loss": 0.7537,
+      "step": 6264
+    },
+    {
+      "epoch": 0.9586840091813313,
+      "grad_norm": 2.2041375129714615,
+      "learning_rate": 1.1153498955772213e-05,
+      "loss": 0.5933,
+      "step": 6265
+    },
+    {
+      "epoch": 0.9588370313695486,
+      "grad_norm": 2.3673848542353024,
+      "learning_rate": 1.1151037279390106e-05,
+      "loss": 0.6677,
+      "step": 6266
+    },
+    {
+      "epoch": 0.9589900535577659,
+      "grad_norm": 2.1283912180610756,
+      "learning_rate": 1.11485755323183e-05,
+      "loss": 0.6213,
+      "step": 6267
+    },
+    {
+      "epoch": 0.9591430757459831,
+      "grad_norm": 2.104333165924923,
+      "learning_rate": 1.1146113714707973e-05,
+      "loss": 0.6682,
+      "step": 6268
+    },
+    {
+      "epoch": 0.9592960979342005,
+      "grad_norm": 2.0811203535808054,
+      "learning_rate": 1.114365182671032e-05,
+      "loss": 0.6221,
+      "step": 6269
+    },
+    {
+      "epoch": 0.9594491201224178,
+      "grad_norm": 2.1932532045760778,
+      "learning_rate": 1.1141189868476536e-05,
+      "loss": 0.7587,
+      "step": 6270
+    },
+    {
+      "epoch": 0.959602142310635,
+      "grad_norm": 2.489560695641433,
+      "learning_rate": 1.1138727840157817e-05,
+      "loss": 0.7461,
+      "step": 6271
+    },
+    {
+      "epoch": 0.9597551644988523,
+      "grad_norm": 2.2827480402995888,
+      "learning_rate": 1.1136265741905367e-05,
+      "loss": 0.6466,
+      "step": 6272
+    },
+    {
+      "epoch": 0.9599081866870697,
+      "grad_norm": 2.5641799903735167,
+      "learning_rate": 1.1133803573870392e-05,
+      "loss": 0.7205,
+      "step": 6273
+    },
+    {
+      "epoch": 0.9600612088752869,
+      "grad_norm": 2.242478983384006,
+      "learning_rate": 1.1131341336204104e-05,
+      "loss": 0.6067,
+      "step": 6274
+    },
+    {
+      "epoch": 0.9602142310635042,
+      "grad_norm": 2.3288654475119297,
+      "learning_rate": 1.112887902905772e-05,
+      "loss": 0.6887,
+      "step": 6275
+    },
+    {
+      "epoch": 0.9603672532517215,
+      "grad_norm": 2.304115755073521,
+      "learning_rate": 1.1126416652582456e-05,
+      "loss": 0.7624,
+      "step": 6276
+    },
+    {
+      "epoch": 0.9605202754399388,
+      "grad_norm": 2.279433405641376,
+      "learning_rate": 1.1123954206929542e-05,
+      "loss": 0.6743,
+      "step": 6277
+    },
+    {
+      "epoch": 0.9606732976281561,
+      "grad_norm": 2.331560845281564,
+      "learning_rate": 1.11214916922502e-05,
+      "loss": 0.6911,
+      "step": 6278
+    },
+    {
+      "epoch": 0.9608263198163733,
+      "grad_norm": 2.273458434140446,
+      "learning_rate": 1.111902910869567e-05,
+      "loss": 0.6622,
+      "step": 6279
+    },
+    {
+      "epoch": 0.9609793420045907,
+      "grad_norm": 2.154670876055646,
+      "learning_rate": 1.1116566456417181e-05,
+      "loss": 0.6577,
+      "step": 6280
+    },
+    {
+      "epoch": 0.961132364192808,
+      "grad_norm": 2.0581058974508832,
+      "learning_rate": 1.1114103735565977e-05,
+      "loss": 0.5898,
+      "step": 6281
+    },
+    {
+      "epoch": 0.9612853863810252,
+      "grad_norm": 2.2992654047403263,
+      "learning_rate": 1.1111640946293308e-05,
+      "loss": 0.7091,
+      "step": 6282
+    },
+    {
+      "epoch": 0.9614384085692426,
+      "grad_norm": 2.315602918362659,
+      "learning_rate": 1.1109178088750422e-05,
+      "loss": 0.5043,
+      "step": 6283
+    },
+    {
+      "epoch": 0.9615914307574598,
+      "grad_norm": 2.35266513608379,
+      "learning_rate": 1.1106715163088561e-05,
+      "loss": 0.6637,
+      "step": 6284
+    },
+    {
+      "epoch": 0.9617444529456771,
+      "grad_norm": 2.1833480375393672,
+      "learning_rate": 1.1104252169459e-05,
+      "loss": 0.6923,
+      "step": 6285
+    },
+    {
+      "epoch": 0.9618974751338945,
+      "grad_norm": 2.551652413206869,
+      "learning_rate": 1.110178910801299e-05,
+      "loss": 0.6136,
+      "step": 6286
+    },
+    {
+      "epoch": 0.9620504973221117,
+      "grad_norm": 2.42416461169718,
+      "learning_rate": 1.1099325978901802e-05,
+      "loss": 0.6589,
+      "step": 6287
+    },
+    {
+      "epoch": 0.962203519510329,
+      "grad_norm": 2.3115177880706854,
+      "learning_rate": 1.1096862782276707e-05,
+      "loss": 0.7251,
+      "step": 6288
+    },
+    {
+      "epoch": 0.9623565416985463,
+      "grad_norm": 2.4133212993849997,
+      "learning_rate": 1.1094399518288975e-05,
+      "loss": 0.6165,
+      "step": 6289
+    },
+    {
+      "epoch": 0.9625095638867636,
+      "grad_norm": 2.476384822568825,
+      "learning_rate": 1.1091936187089886e-05,
+      "loss": 0.7031,
+      "step": 6290
+    },
+    {
+      "epoch": 0.9626625860749809,
+      "grad_norm": 2.497069760113835,
+      "learning_rate": 1.1089472788830728e-05,
+      "loss": 0.8166,
+      "step": 6291
+    },
+    {
+      "epoch": 0.9628156082631981,
+      "grad_norm": 2.1440996425297296,
+      "learning_rate": 1.1087009323662784e-05,
+      "loss": 0.6813,
+      "step": 6292
+    },
+    {
+      "epoch": 0.9629686304514155,
+      "grad_norm": 2.2455927624657135,
+      "learning_rate": 1.108454579173734e-05,
+      "loss": 0.6084,
+      "step": 6293
+    },
+    {
+      "epoch": 0.9631216526396328,
+      "grad_norm": 2.240564503687054,
+      "learning_rate": 1.1082082193205699e-05,
+      "loss": 0.5922,
+      "step": 6294
+    },
+    {
+      "epoch": 0.96327467482785,
+      "grad_norm": 2.232747072090611,
+      "learning_rate": 1.1079618528219159e-05,
+      "loss": 0.6345,
+      "step": 6295
+    },
+    {
+      "epoch": 0.9634276970160673,
+      "grad_norm": 2.3703209346787446,
+      "learning_rate": 1.107715479692902e-05,
+      "loss": 0.6895,
+      "step": 6296
+    },
+    {
+      "epoch": 0.9635807192042847,
+      "grad_norm": 2.241003272516828,
+      "learning_rate": 1.1074690999486591e-05,
+      "loss": 0.7439,
+      "step": 6297
+    },
+    {
+      "epoch": 0.9637337413925019,
+      "grad_norm": 2.0491357691467074,
+      "learning_rate": 1.1072227136043182e-05,
+      "loss": 0.6752,
+      "step": 6298
+    },
+    {
+      "epoch": 0.9638867635807192,
+      "grad_norm": 2.236282969158411,
+      "learning_rate": 1.1069763206750113e-05,
+      "loss": 0.616,
+      "step": 6299
+    },
+    {
+      "epoch": 0.9640397857689365,
+      "grad_norm": 2.4976383991206865,
+      "learning_rate": 1.10672992117587e-05,
+      "loss": 0.6593,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9641928079571538,
+      "grad_norm": 2.2143008291617483,
+      "learning_rate": 1.1064835151220265e-05,
+      "loss": 0.6282,
+      "step": 6301
+    },
+    {
+      "epoch": 0.9643458301453711,
+      "grad_norm": 2.4819206347775427,
+      "learning_rate": 1.106237102528614e-05,
+      "loss": 0.6849,
+      "step": 6302
+    },
+    {
+      "epoch": 0.9644988523335883,
+      "grad_norm": 2.2845724829416985,
+      "learning_rate": 1.1059906834107652e-05,
+      "loss": 0.6554,
+      "step": 6303
+    },
+    {
+      "epoch": 0.9646518745218057,
+      "grad_norm": 2.536708963196747,
+      "learning_rate": 1.1057442577836141e-05,
+      "loss": 0.7574,
+      "step": 6304
+    },
+    {
+      "epoch": 0.964804896710023,
+      "grad_norm": 2.4234275083828565,
+      "learning_rate": 1.1054978256622946e-05,
+      "loss": 0.6173,
+      "step": 6305
+    },
+    {
+      "epoch": 0.9649579188982402,
+      "grad_norm": 2.3943143908981805,
+      "learning_rate": 1.1052513870619403e-05,
+      "loss": 0.7037,
+      "step": 6306
+    },
+    {
+      "epoch": 0.9651109410864576,
+      "grad_norm": 2.3606987033507507,
+      "learning_rate": 1.1050049419976872e-05,
+      "loss": 0.695,
+      "step": 6307
+    },
+    {
+      "epoch": 0.9652639632746748,
+      "grad_norm": 2.5633579786892686,
+      "learning_rate": 1.1047584904846697e-05,
+      "loss": 0.6554,
+      "step": 6308
+    },
+    {
+      "epoch": 0.9654169854628921,
+      "grad_norm": 2.2020850155001543,
+      "learning_rate": 1.1045120325380233e-05,
+      "loss": 0.6187,
+      "step": 6309
+    },
+    {
+      "epoch": 0.9655700076511095,
+      "grad_norm": 2.1260293430856634,
+      "learning_rate": 1.1042655681728842e-05,
+      "loss": 0.6128,
+      "step": 6310
+    },
+    {
+      "epoch": 0.9657230298393267,
+      "grad_norm": 2.3544765145165045,
+      "learning_rate": 1.1040190974043887e-05,
+      "loss": 0.6527,
+      "step": 6311
+    },
+    {
+      "epoch": 0.965876052027544,
+      "grad_norm": 2.218314723977588,
+      "learning_rate": 1.1037726202476735e-05,
+      "loss": 0.7008,
+      "step": 6312
+    },
+    {
+      "epoch": 0.9660290742157613,
+      "grad_norm": 2.403254705515555,
+      "learning_rate": 1.1035261367178758e-05,
+      "loss": 0.6956,
+      "step": 6313
+    },
+    {
+      "epoch": 0.9661820964039786,
+      "grad_norm": 2.0814432448403006,
+      "learning_rate": 1.103279646830133e-05,
+      "loss": 0.6016,
+      "step": 6314
+    },
+    {
+      "epoch": 0.9663351185921959,
+      "grad_norm": 2.379683650495674,
+      "learning_rate": 1.103033150599583e-05,
+      "loss": 0.6538,
+      "step": 6315
+    },
+    {
+      "epoch": 0.9664881407804131,
+      "grad_norm": 2.286013236090055,
+      "learning_rate": 1.1027866480413642e-05,
+      "loss": 0.6498,
+      "step": 6316
+    },
+    {
+      "epoch": 0.9666411629686305,
+      "grad_norm": 2.3171666595250646,
+      "learning_rate": 1.1025401391706152e-05,
+      "loss": 0.6939,
+      "step": 6317
+    },
+    {
+      "epoch": 0.9667941851568478,
+      "grad_norm": 2.1385639093202724,
+      "learning_rate": 1.1022936240024754e-05,
+      "loss": 0.6075,
+      "step": 6318
+    },
+    {
+      "epoch": 0.966947207345065,
+      "grad_norm": 2.038129807160247,
+      "learning_rate": 1.1020471025520833e-05,
+      "loss": 0.5521,
+      "step": 6319
+    },
+    {
+      "epoch": 0.9671002295332823,
+      "grad_norm": 2.782764068823436,
+      "learning_rate": 1.1018005748345802e-05,
+      "loss": 0.7901,
+      "step": 6320
+    },
+    {
+      "epoch": 0.9672532517214996,
+      "grad_norm": 1.9353389662910547,
+      "learning_rate": 1.1015540408651051e-05,
+      "loss": 0.5659,
+      "step": 6321
+    },
+    {
+      "epoch": 0.9674062739097169,
+      "grad_norm": 2.364548658211014,
+      "learning_rate": 1.101307500658799e-05,
+      "loss": 0.6433,
+      "step": 6322
+    },
+    {
+      "epoch": 0.9675592960979342,
+      "grad_norm": 2.14311294534737,
+      "learning_rate": 1.101060954230803e-05,
+      "loss": 0.6272,
+      "step": 6323
+    },
+    {
+      "epoch": 0.9677123182861515,
+      "grad_norm": 2.3056064170443893,
+      "learning_rate": 1.100814401596259e-05,
+      "loss": 0.6459,
+      "step": 6324
+    },
+    {
+      "epoch": 0.9678653404743688,
+      "grad_norm": 2.02182565029233,
+      "learning_rate": 1.1005678427703075e-05,
+      "loss": 0.6518,
+      "step": 6325
+    },
+    {
+      "epoch": 0.9680183626625861,
+      "grad_norm": 2.3303131018908547,
+      "learning_rate": 1.1003212777680916e-05,
+      "loss": 0.6813,
+      "step": 6326
+    },
+    {
+      "epoch": 0.9681713848508033,
+      "grad_norm": 2.114523748506382,
+      "learning_rate": 1.1000747066047536e-05,
+      "loss": 0.5194,
+      "step": 6327
+    },
+    {
+      "epoch": 0.9683244070390207,
+      "grad_norm": 2.265418197230814,
+      "learning_rate": 1.0998281292954362e-05,
+      "loss": 0.6422,
+      "step": 6328
+    },
+    {
+      "epoch": 0.9684774292272379,
+      "grad_norm": 2.338030038622006,
+      "learning_rate": 1.0995815458552833e-05,
+      "loss": 0.7219,
+      "step": 6329
+    },
+    {
+      "epoch": 0.9686304514154552,
+      "grad_norm": 2.521459284136043,
+      "learning_rate": 1.0993349562994378e-05,
+      "loss": 0.6203,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9687834736036726,
+      "grad_norm": 2.0493537019726564,
+      "learning_rate": 1.0990883606430439e-05,
+      "loss": 0.593,
+      "step": 6331
+    },
+    {
+      "epoch": 0.9689364957918898,
+      "grad_norm": 2.6405871157081777,
+      "learning_rate": 1.0988417589012464e-05,
+      "loss": 0.79,
+      "step": 6332
+    },
+    {
+      "epoch": 0.9690895179801071,
+      "grad_norm": 2.096405636694104,
+      "learning_rate": 1.0985951510891899e-05,
+      "loss": 0.7241,
+      "step": 6333
+    },
+    {
+      "epoch": 0.9692425401683245,
+      "grad_norm": 2.088604644544387,
+      "learning_rate": 1.098348537222019e-05,
+      "loss": 0.6886,
+      "step": 6334
+    },
+    {
+      "epoch": 0.9693955623565417,
+      "grad_norm": 1.9904441066693555,
+      "learning_rate": 1.0981019173148802e-05,
+      "loss": 0.5716,
+      "step": 6335
+    },
+    {
+      "epoch": 0.969548584544759,
+      "grad_norm": 2.062780568286232,
+      "learning_rate": 1.0978552913829185e-05,
+      "loss": 0.5858,
+      "step": 6336
+    },
+    {
+      "epoch": 0.9697016067329762,
+      "grad_norm": 2.2337258488936538,
+      "learning_rate": 1.0976086594412808e-05,
+      "loss": 0.6003,
+      "step": 6337
+    },
+    {
+      "epoch": 0.9698546289211936,
+      "grad_norm": 2.372510122203268,
+      "learning_rate": 1.0973620215051132e-05,
+      "loss": 0.7456,
+      "step": 6338
+    },
+    {
+      "epoch": 0.9700076511094109,
+      "grad_norm": 2.2495942249697083,
+      "learning_rate": 1.0971153775895633e-05,
+      "loss": 0.6003,
+      "step": 6339
+    },
+    {
+      "epoch": 0.9701606732976281,
+      "grad_norm": 2.239803426785077,
+      "learning_rate": 1.096868727709778e-05,
+      "loss": 0.5947,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9703136954858455,
+      "grad_norm": 2.0374315924280446,
+      "learning_rate": 1.0966220718809054e-05,
+      "loss": 0.6142,
+      "step": 6341
+    },
+    {
+      "epoch": 0.9704667176740628,
+      "grad_norm": 2.1103707648171106,
+      "learning_rate": 1.096375410118093e-05,
+      "loss": 0.5306,
+      "step": 6342
+    },
+    {
+      "epoch": 0.97061973986228,
+      "grad_norm": 2.07129788644162,
+      "learning_rate": 1.0961287424364896e-05,
+      "loss": 0.6902,
+      "step": 6343
+    },
+    {
+      "epoch": 0.9707727620504973,
+      "grad_norm": 2.3083558399211164,
+      "learning_rate": 1.0958820688512441e-05,
+      "loss": 0.6831,
+      "step": 6344
+    },
+    {
+      "epoch": 0.9709257842387146,
+      "grad_norm": 2.426716481951558,
+      "learning_rate": 1.0956353893775062e-05,
+      "loss": 0.7141,
+      "step": 6345
+    },
+    {
+      "epoch": 0.9710788064269319,
+      "grad_norm": 2.162494654252418,
+      "learning_rate": 1.0953887040304245e-05,
+      "loss": 0.5416,
+      "step": 6346
+    },
+    {
+      "epoch": 0.9712318286151492,
+      "grad_norm": 2.1205750952379447,
+      "learning_rate": 1.0951420128251491e-05,
+      "loss": 0.6947,
+      "step": 6347
+    },
+    {
+      "epoch": 0.9713848508033665,
+      "grad_norm": 2.4172902478803358,
+      "learning_rate": 1.0948953157768309e-05,
+      "loss": 0.667,
+      "step": 6348
+    },
+    {
+      "epoch": 0.9715378729915838,
+      "grad_norm": 2.238302293539362,
+      "learning_rate": 1.0946486129006202e-05,
+      "loss": 0.666,
+      "step": 6349
+    },
+    {
+      "epoch": 0.9716908951798011,
+      "grad_norm": 2.2943963497156576,
+      "learning_rate": 1.0944019042116673e-05,
+      "loss": 0.6383,
+      "step": 6350
+    },
+    {
+      "epoch": 0.9718439173680183,
+      "grad_norm": 2.0424734728227354,
+      "learning_rate": 1.0941551897251248e-05,
+      "loss": 0.6146,
+      "step": 6351
+    },
+    {
+      "epoch": 0.9719969395562357,
+      "grad_norm": 2.348686525361142,
+      "learning_rate": 1.0939084694561434e-05,
+      "loss": 0.6622,
+      "step": 6352
+    },
+    {
+      "epoch": 0.9721499617444529,
+      "grad_norm": 1.9904593663363055,
+      "learning_rate": 1.0936617434198758e-05,
+      "loss": 0.5622,
+      "step": 6353
+    },
+    {
+      "epoch": 0.9723029839326702,
+      "grad_norm": 2.2619272391880254,
+      "learning_rate": 1.093415011631474e-05,
+      "loss": 0.7357,
+      "step": 6354
+    },
+    {
+      "epoch": 0.9724560061208876,
+      "grad_norm": 2.142159801287082,
+      "learning_rate": 1.0931682741060907e-05,
+      "loss": 0.6831,
+      "step": 6355
+    },
+    {
+      "epoch": 0.9726090283091048,
+      "grad_norm": 2.4643906335625196,
+      "learning_rate": 1.0929215308588795e-05,
+      "loss": 0.715,
+      "step": 6356
+    },
+    {
+      "epoch": 0.9727620504973221,
+      "grad_norm": 2.0428082848603037,
+      "learning_rate": 1.0926747819049935e-05,
+      "loss": 0.6328,
+      "step": 6357
+    },
+    {
+      "epoch": 0.9729150726855395,
+      "grad_norm": 2.253982118695505,
+      "learning_rate": 1.0924280272595863e-05,
+      "loss": 0.7491,
+      "step": 6358
+    },
+    {
+      "epoch": 0.9730680948737567,
+      "grad_norm": 2.1914156685495443,
+      "learning_rate": 1.0921812669378126e-05,
+      "loss": 0.6725,
+      "step": 6359
+    },
+    {
+      "epoch": 0.973221117061974,
+      "grad_norm": 2.26967414402112,
+      "learning_rate": 1.0919345009548266e-05,
+      "loss": 0.7652,
+      "step": 6360
+    },
+    {
+      "epoch": 0.9733741392501912,
+      "grad_norm": 2.246918731454627,
+      "learning_rate": 1.0916877293257837e-05,
+      "loss": 0.635,
+      "step": 6361
+    },
+    {
+      "epoch": 0.9735271614384086,
+      "grad_norm": 2.237823796273232,
+      "learning_rate": 1.0914409520658382e-05,
+      "loss": 0.6303,
+      "step": 6362
+    },
+    {
+      "epoch": 0.9736801836266259,
+      "grad_norm": 2.3232073625174903,
+      "learning_rate": 1.091194169190146e-05,
+      "loss": 0.5327,
+      "step": 6363
+    },
+    {
+      "epoch": 0.9738332058148431,
+      "grad_norm": 2.387396773346526,
+      "learning_rate": 1.0909473807138633e-05,
+      "loss": 0.6605,
+      "step": 6364
+    },
+    {
+      "epoch": 0.9739862280030605,
+      "grad_norm": 2.3255204349093996,
+      "learning_rate": 1.0907005866521462e-05,
+      "loss": 0.6902,
+      "step": 6365
+    },
+    {
+      "epoch": 0.9741392501912778,
+      "grad_norm": 2.350835243846951,
+      "learning_rate": 1.0904537870201512e-05,
+      "loss": 0.6583,
+      "step": 6366
+    },
+    {
+      "epoch": 0.974292272379495,
+      "grad_norm": 2.0874447124765667,
+      "learning_rate": 1.0902069818330354e-05,
+      "loss": 0.6323,
+      "step": 6367
+    },
+    {
+      "epoch": 0.9744452945677123,
+      "grad_norm": 2.4843168652978216,
+      "learning_rate": 1.089960171105956e-05,
+      "loss": 0.7759,
+      "step": 6368
+    },
+    {
+      "epoch": 0.9745983167559296,
+      "grad_norm": 2.375106997044347,
+      "learning_rate": 1.0897133548540701e-05,
+      "loss": 0.6102,
+      "step": 6369
+    },
+    {
+      "epoch": 0.9747513389441469,
+      "grad_norm": 2.343614659524406,
+      "learning_rate": 1.0894665330925368e-05,
+      "loss": 0.6207,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9749043611323642,
+      "grad_norm": 2.2805865460307237,
+      "learning_rate": 1.0892197058365135e-05,
+      "loss": 0.6677,
+      "step": 6371
+    },
+    {
+      "epoch": 0.9750573833205815,
+      "grad_norm": 2.235287175901108,
+      "learning_rate": 1.0889728731011587e-05,
+      "loss": 0.7031,
+      "step": 6372
+    },
+    {
+      "epoch": 0.9752104055087988,
+      "grad_norm": 2.104978197412232,
+      "learning_rate": 1.0887260349016318e-05,
+      "loss": 0.5624,
+      "step": 6373
+    },
+    {
+      "epoch": 0.9753634276970161,
+      "grad_norm": 2.3101831680546323,
+      "learning_rate": 1.0884791912530923e-05,
+      "loss": 0.5689,
+      "step": 6374
+    },
+    {
+      "epoch": 0.9755164498852333,
+      "grad_norm": 2.3371258325914255,
+      "learning_rate": 1.0882323421706991e-05,
+      "loss": 0.7777,
+      "step": 6375
+    },
+    {
+      "epoch": 0.9756694720734507,
+      "grad_norm": 2.7997626109199745,
+      "learning_rate": 1.087985487669613e-05,
+      "loss": 0.8104,
+      "step": 6376
+    },
+    {
+      "epoch": 0.9758224942616679,
+      "grad_norm": 2.412073839993447,
+      "learning_rate": 1.0877386277649938e-05,
+      "loss": 0.7437,
+      "step": 6377
+    },
+    {
+      "epoch": 0.9759755164498852,
+      "grad_norm": 2.316799724033941,
+      "learning_rate": 1.087491762472002e-05,
+      "loss": 0.725,
+      "step": 6378
+    },
+    {
+      "epoch": 0.9761285386381026,
+      "grad_norm": 2.2241974714922486,
+      "learning_rate": 1.0872448918057989e-05,
+      "loss": 0.768,
+      "step": 6379
+    },
+    {
+      "epoch": 0.9762815608263198,
+      "grad_norm": 2.0678765281664386,
+      "learning_rate": 1.0869980157815458e-05,
+      "loss": 0.5852,
+      "step": 6380
+    },
+    {
+      "epoch": 0.9764345830145371,
+      "grad_norm": 2.211865767184328,
+      "learning_rate": 1.0867511344144042e-05,
+      "loss": 0.6851,
+      "step": 6381
+    },
+    {
+      "epoch": 0.9765876052027544,
+      "grad_norm": 2.113420351755564,
+      "learning_rate": 1.0865042477195362e-05,
+      "loss": 0.6036,
+      "step": 6382
+    },
+    {
+      "epoch": 0.9767406273909717,
+      "grad_norm": 2.238136535568473,
+      "learning_rate": 1.0862573557121036e-05,
+      "loss": 0.6267,
+      "step": 6383
+    },
+    {
+      "epoch": 0.976893649579189,
+      "grad_norm": 2.2475432663562107,
+      "learning_rate": 1.0860104584072698e-05,
+      "loss": 0.7343,
+      "step": 6384
+    },
+    {
+      "epoch": 0.9770466717674062,
+      "grad_norm": 2.332824885622173,
+      "learning_rate": 1.0857635558201966e-05,
+      "loss": 0.6669,
+      "step": 6385
+    },
+    {
+      "epoch": 0.9771996939556236,
+      "grad_norm": 2.1477623949536624,
+      "learning_rate": 1.0855166479660487e-05,
+      "loss": 0.5838,
+      "step": 6386
+    },
+    {
+      "epoch": 0.9773527161438409,
+      "grad_norm": 2.2525788281704773,
+      "learning_rate": 1.0852697348599883e-05,
+      "loss": 0.7541,
+      "step": 6387
+    },
+    {
+      "epoch": 0.9775057383320581,
+      "grad_norm": 2.2180595252269955,
+      "learning_rate": 1.08502281651718e-05,
+      "loss": 0.6829,
+      "step": 6388
+    },
+    {
+      "epoch": 0.9776587605202754,
+      "grad_norm": 2.216679839529683,
+      "learning_rate": 1.0847758929527881e-05,
+      "loss": 0.6328,
+      "step": 6389
+    },
+    {
+      "epoch": 0.9778117827084928,
+      "grad_norm": 2.094749687529566,
+      "learning_rate": 1.0845289641819772e-05,
+      "loss": 0.5704,
+      "step": 6390
+    },
+    {
+      "epoch": 0.97796480489671,
+      "grad_norm": 2.158104895092101,
+      "learning_rate": 1.0842820302199115e-05,
+      "loss": 0.6678,
+      "step": 6391
+    },
+    {
+      "epoch": 0.9781178270849273,
+      "grad_norm": 2.5011576786078495,
+      "learning_rate": 1.0840350910817568e-05,
+      "loss": 0.6058,
+      "step": 6392
+    },
+    {
+      "epoch": 0.9782708492731446,
+      "grad_norm": 2.1857763444580067,
+      "learning_rate": 1.0837881467826783e-05,
+      "loss": 0.6703,
+      "step": 6393
+    },
+    {
+      "epoch": 0.9784238714613619,
+      "grad_norm": 2.2003107696201027,
+      "learning_rate": 1.0835411973378421e-05,
+      "loss": 0.7169,
+      "step": 6394
+    },
+    {
+      "epoch": 0.9785768936495792,
+      "grad_norm": 2.0337957252981096,
+      "learning_rate": 1.083294242762414e-05,
+      "loss": 0.5438,
+      "step": 6395
+    },
+    {
+      "epoch": 0.9787299158377964,
+      "grad_norm": 2.150512225972815,
+      "learning_rate": 1.0830472830715605e-05,
+      "loss": 0.6302,
+      "step": 6396
+    },
+    {
+      "epoch": 0.9788829380260138,
+      "grad_norm": 2.484740579652343,
+      "learning_rate": 1.0828003182804488e-05,
+      "loss": 0.6283,
+      "step": 6397
+    },
+    {
+      "epoch": 0.9790359602142311,
+      "grad_norm": 2.5080302326286668,
+      "learning_rate": 1.0825533484042451e-05,
+      "loss": 0.7363,
+      "step": 6398
+    },
+    {
+      "epoch": 0.9791889824024483,
+      "grad_norm": 2.4527364589897527,
+      "learning_rate": 1.082306373458118e-05,
+      "loss": 0.6692,
+      "step": 6399
+    },
+    {
+      "epoch": 0.9793420045906657,
+      "grad_norm": 2.5425671821277587,
+      "learning_rate": 1.0820593934572344e-05,
+      "loss": 0.5979,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9794950267788829,
+      "grad_norm": 2.3776505208003607,
+      "learning_rate": 1.0818124084167619e-05,
+      "loss": 0.6738,
+      "step": 6401
+    },
+    {
+      "epoch": 0.9796480489671002,
+      "grad_norm": 2.2778353809646887,
+      "learning_rate": 1.0815654183518698e-05,
+      "loss": 0.7114,
+      "step": 6402
+    },
+    {
+      "epoch": 0.9798010711553176,
+      "grad_norm": 2.193755629471049,
+      "learning_rate": 1.0813184232777262e-05,
+      "loss": 0.6681,
+      "step": 6403
+    },
+    {
+      "epoch": 0.9799540933435348,
+      "grad_norm": 1.9695219026147595,
+      "learning_rate": 1.0810714232095001e-05,
+      "loss": 0.6318,
+      "step": 6404
+    },
+    {
+      "epoch": 0.9801071155317521,
+      "grad_norm": 2.279528997785273,
+      "learning_rate": 1.0808244181623606e-05,
+      "loss": 0.6502,
+      "step": 6405
+    },
+    {
+      "epoch": 0.9802601377199694,
+      "grad_norm": 2.3554587060275716,
+      "learning_rate": 1.0805774081514774e-05,
+      "loss": 0.7645,
+      "step": 6406
+    },
+    {
+      "epoch": 0.9804131599081867,
+      "grad_norm": 2.178746017369633,
+      "learning_rate": 1.0803303931920206e-05,
+      "loss": 0.5972,
+      "step": 6407
+    },
+    {
+      "epoch": 0.980566182096404,
+      "grad_norm": 2.200882154882437,
+      "learning_rate": 1.08008337329916e-05,
+      "loss": 0.6289,
+      "step": 6408
+    },
+    {
+      "epoch": 0.9807192042846212,
+      "grad_norm": 2.1322827299103064,
+      "learning_rate": 1.079836348488066e-05,
+      "loss": 0.5765,
+      "step": 6409
+    },
+    {
+      "epoch": 0.9808722264728386,
+      "grad_norm": 2.264692330253768,
+      "learning_rate": 1.0795893187739095e-05,
+      "loss": 0.728,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9810252486610559,
+      "grad_norm": 2.2316755970527984,
+      "learning_rate": 1.079342284171862e-05,
+      "loss": 0.6295,
+      "step": 6411
+    },
+    {
+      "epoch": 0.9811782708492731,
+      "grad_norm": 2.2053391147834756,
+      "learning_rate": 1.0790952446970942e-05,
+      "loss": 0.6607,
+      "step": 6412
+    },
+    {
+      "epoch": 0.9813312930374904,
+      "grad_norm": 2.1109423143732897,
+      "learning_rate": 1.0788482003647775e-05,
+      "loss": 0.6516,
+      "step": 6413
+    },
+    {
+      "epoch": 0.9814843152257078,
+      "grad_norm": 2.268297928669402,
+      "learning_rate": 1.078601151190085e-05,
+      "loss": 0.7403,
+      "step": 6414
+    },
+    {
+      "epoch": 0.981637337413925,
+      "grad_norm": 2.2137529570759122,
+      "learning_rate": 1.0783540971881882e-05,
+      "loss": 0.6743,
+      "step": 6415
+    },
+    {
+      "epoch": 0.9817903596021423,
+      "grad_norm": 2.0507644073612137,
+      "learning_rate": 1.0781070383742595e-05,
+      "loss": 0.4971,
+      "step": 6416
+    },
+    {
+      "epoch": 0.9819433817903596,
+      "grad_norm": 2.088792875167074,
+      "learning_rate": 1.0778599747634723e-05,
+      "loss": 0.6713,
+      "step": 6417
+    },
+    {
+      "epoch": 0.9820964039785769,
+      "grad_norm": 2.3516082787434063,
+      "learning_rate": 1.0776129063709994e-05,
+      "loss": 0.6598,
+      "step": 6418
+    },
+    {
+      "epoch": 0.9822494261667942,
+      "grad_norm": 2.6734111206794915,
+      "learning_rate": 1.0773658332120143e-05,
+      "loss": 0.6645,
+      "step": 6419
+    },
+    {
+      "epoch": 0.9824024483550114,
+      "grad_norm": 2.3219289364900053,
+      "learning_rate": 1.0771187553016907e-05,
+      "loss": 0.6809,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9825554705432288,
+      "grad_norm": 2.295162797643992,
+      "learning_rate": 1.0768716726552027e-05,
+      "loss": 0.6209,
+      "step": 6421
+    },
+    {
+      "epoch": 0.9827084927314461,
+      "grad_norm": 2.1896216122076155,
+      "learning_rate": 1.0766245852877244e-05,
+      "loss": 0.633,
+      "step": 6422
+    },
+    {
+      "epoch": 0.9828615149196633,
+      "grad_norm": 2.423235832841934,
+      "learning_rate": 1.0763774932144306e-05,
+      "loss": 0.634,
+      "step": 6423
+    },
+    {
+      "epoch": 0.9830145371078807,
+      "grad_norm": 2.352333616434348,
+      "learning_rate": 1.0761303964504963e-05,
+      "loss": 0.6987,
+      "step": 6424
+    },
+    {
+      "epoch": 0.9831675592960979,
+      "grad_norm": 2.223375288114711,
+      "learning_rate": 1.0758832950110965e-05,
+      "loss": 0.5861,
+      "step": 6425
+    },
+    {
+      "epoch": 0.9833205814843152,
+      "grad_norm": 2.2183349950150646,
+      "learning_rate": 1.0756361889114065e-05,
+      "loss": 0.6365,
+      "step": 6426
+    },
+    {
+      "epoch": 0.9834736036725326,
+      "grad_norm": 1.9636195416029705,
+      "learning_rate": 1.0753890781666025e-05,
+      "loss": 0.6346,
+      "step": 6427
+    },
+    {
+      "epoch": 0.9836266258607498,
+      "grad_norm": 2.2297263596321404,
+      "learning_rate": 1.0751419627918602e-05,
+      "loss": 0.666,
+      "step": 6428
+    },
+    {
+      "epoch": 0.9837796480489671,
+      "grad_norm": 2.4350159907453146,
+      "learning_rate": 1.0748948428023557e-05,
+      "loss": 0.6205,
+      "step": 6429
+    },
+    {
+      "epoch": 0.9839326702371843,
+      "grad_norm": 2.181247738797785,
+      "learning_rate": 1.0746477182132665e-05,
+      "loss": 0.6318,
+      "step": 6430
+    },
+    {
+      "epoch": 0.9840856924254017,
+      "grad_norm": 2.646466395978632,
+      "learning_rate": 1.0744005890397686e-05,
+      "loss": 0.6104,
+      "step": 6431
+    },
+    {
+      "epoch": 0.984238714613619,
+      "grad_norm": 2.2751928031916804,
+      "learning_rate": 1.0741534552970393e-05,
+      "loss": 0.6352,
+      "step": 6432
+    },
+    {
+      "epoch": 0.9843917368018362,
+      "grad_norm": 19.9658720243908,
+      "learning_rate": 1.0739063170002564e-05,
+      "loss": 0.6467,
+      "step": 6433
+    },
+    {
+      "epoch": 0.9845447589900536,
+      "grad_norm": 2.084851787789992,
+      "learning_rate": 1.0736591741645974e-05,
+      "loss": 0.4974,
+      "step": 6434
+    },
+    {
+      "epoch": 0.9846977811782709,
+      "grad_norm": 2.0840003737302095,
+      "learning_rate": 1.0734120268052403e-05,
+      "loss": 0.6906,
+      "step": 6435
+    },
+    {
+      "epoch": 0.9848508033664881,
+      "grad_norm": 2.2573509449638096,
+      "learning_rate": 1.0731648749373637e-05,
+      "loss": 0.6702,
+      "step": 6436
+    },
+    {
+      "epoch": 0.9850038255547054,
+      "grad_norm": 2.105354426708069,
+      "learning_rate": 1.0729177185761457e-05,
+      "loss": 0.571,
+      "step": 6437
+    },
+    {
+      "epoch": 0.9851568477429227,
+      "grad_norm": 2.288305079827146,
+      "learning_rate": 1.0726705577367652e-05,
+      "loss": 0.7155,
+      "step": 6438
+    },
+    {
+      "epoch": 0.98530986993114,
+      "grad_norm": 2.2850243056547734,
+      "learning_rate": 1.0724233924344016e-05,
+      "loss": 0.6159,
+      "step": 6439
+    },
+    {
+      "epoch": 0.9854628921193573,
+      "grad_norm": 1.9649566640224054,
+      "learning_rate": 1.0721762226842344e-05,
+      "loss": 0.5237,
+      "step": 6440
+    },
+    {
+      "epoch": 0.9856159143075746,
+      "grad_norm": 1.9893769537451282,
+      "learning_rate": 1.0719290485014428e-05,
+      "loss": 0.5814,
+      "step": 6441
+    },
+    {
+      "epoch": 0.9857689364957919,
+      "grad_norm": 2.564034080142649,
+      "learning_rate": 1.0716818699012067e-05,
+      "loss": 0.7568,
+      "step": 6442
+    },
+    {
+      "epoch": 0.9859219586840092,
+      "grad_norm": 2.295834576636577,
+      "learning_rate": 1.0714346868987068e-05,
+      "loss": 0.7564,
+      "step": 6443
+    },
+    {
+      "epoch": 0.9860749808722264,
+      "grad_norm": 2.0971214868536894,
+      "learning_rate": 1.0711874995091239e-05,
+      "loss": 0.6822,
+      "step": 6444
+    },
+    {
+      "epoch": 0.9862280030604438,
+      "grad_norm": 2.21508227273722,
+      "learning_rate": 1.0709403077476372e-05,
+      "loss": 0.6229,
+      "step": 6445
+    },
+    {
+      "epoch": 0.986381025248661,
+      "grad_norm": 2.043342706115372,
+      "learning_rate": 1.070693111629429e-05,
+      "loss": 0.5323,
+      "step": 6446
+    },
+    {
+      "epoch": 0.9865340474368783,
+      "grad_norm": 2.2277205626682277,
+      "learning_rate": 1.0704459111696805e-05,
+      "loss": 0.6203,
+      "step": 6447
+    },
+    {
+      "epoch": 0.9866870696250957,
+      "grad_norm": 2.5719639860424133,
+      "learning_rate": 1.0701987063835728e-05,
+      "loss": 0.784,
+      "step": 6448
+    },
+    {
+      "epoch": 0.9868400918133129,
+      "grad_norm": 2.4296363914023047,
+      "learning_rate": 1.069951497286288e-05,
+      "loss": 0.6575,
+      "step": 6449
+    },
+    {
+      "epoch": 0.9869931140015302,
+      "grad_norm": 2.7202807521864147,
+      "learning_rate": 1.069704283893008e-05,
+      "loss": 0.6461,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9871461361897476,
+      "grad_norm": 2.143158373415149,
+      "learning_rate": 1.0694570662189152e-05,
+      "loss": 0.6412,
+      "step": 6451
+    },
+    {
+      "epoch": 0.9872991583779648,
+      "grad_norm": 2.3625004784021257,
+      "learning_rate": 1.0692098442791925e-05,
+      "loss": 0.6861,
+      "step": 6452
+    },
+    {
+      "epoch": 0.9874521805661821,
+      "grad_norm": 2.5257222065048013,
+      "learning_rate": 1.0689626180890225e-05,
+      "loss": 0.7539,
+      "step": 6453
+    },
+    {
+      "epoch": 0.9876052027543993,
+      "grad_norm": 2.1250124084674864,
+      "learning_rate": 1.0687153876635878e-05,
+      "loss": 0.7075,
+      "step": 6454
+    },
+    {
+      "epoch": 0.9877582249426167,
+      "grad_norm": 2.409259571306803,
+      "learning_rate": 1.068468153018073e-05,
+      "loss": 0.6375,
+      "step": 6455
+    },
+    {
+      "epoch": 0.987911247130834,
+      "grad_norm": 2.32874679222682,
+      "learning_rate": 1.068220914167661e-05,
+      "loss": 0.6387,
+      "step": 6456
+    },
+    {
+      "epoch": 0.9880642693190512,
+      "grad_norm": 2.2644969331295517,
+      "learning_rate": 1.0679736711275352e-05,
+      "loss": 0.6875,
+      "step": 6457
+    },
+    {
+      "epoch": 0.9882172915072686,
+      "grad_norm": 1.9025109306992631,
+      "learning_rate": 1.0677264239128809e-05,
+      "loss": 0.5653,
+      "step": 6458
+    },
+    {
+      "epoch": 0.9883703136954859,
+      "grad_norm": 2.1561627079866503,
+      "learning_rate": 1.0674791725388818e-05,
+      "loss": 0.6281,
+      "step": 6459
+    },
+    {
+      "epoch": 0.9885233358837031,
+      "grad_norm": 2.061516117776862,
+      "learning_rate": 1.0672319170207227e-05,
+      "loss": 0.6177,
+      "step": 6460
+    },
+    {
+      "epoch": 0.9886763580719204,
+      "grad_norm": 2.2521575998416066,
+      "learning_rate": 1.0669846573735885e-05,
+      "loss": 0.6979,
+      "step": 6461
+    },
+    {
+      "epoch": 0.9888293802601377,
+      "grad_norm": 2.1209188840044573,
+      "learning_rate": 1.0667373936126646e-05,
+      "loss": 0.7254,
+      "step": 6462
+    },
+    {
+      "epoch": 0.988982402448355,
+      "grad_norm": 2.4464366863064515,
+      "learning_rate": 1.0664901257531362e-05,
+      "loss": 0.6411,
+      "step": 6463
+    },
+    {
+      "epoch": 0.9891354246365723,
+      "grad_norm": 2.2092005460424895,
+      "learning_rate": 1.0662428538101893e-05,
+      "loss": 0.6615,
+      "step": 6464
+    },
+    {
+      "epoch": 0.9892884468247896,
+      "grad_norm": 2.298112700131103,
+      "learning_rate": 1.0659955777990092e-05,
+      "loss": 0.6517,
+      "step": 6465
+    },
+    {
+      "epoch": 0.9894414690130069,
+      "grad_norm": 2.51742007501707,
+      "learning_rate": 1.0657482977347826e-05,
+      "loss": 0.6929,
+      "step": 6466
+    },
+    {
+      "epoch": 0.9895944912012242,
+      "grad_norm": 2.2902881205759833,
+      "learning_rate": 1.0655010136326957e-05,
+      "loss": 0.6197,
+      "step": 6467
+    },
+    {
+      "epoch": 0.9897475133894414,
+      "grad_norm": 2.0662944894518684,
+      "learning_rate": 1.0652537255079359e-05,
+      "loss": 0.6396,
+      "step": 6468
+    },
+    {
+      "epoch": 0.9899005355776588,
+      "grad_norm": 2.3815178778296033,
+      "learning_rate": 1.0650064333756892e-05,
+      "loss": 0.7265,
+      "step": 6469
+    },
+    {
+      "epoch": 0.990053557765876,
+      "grad_norm": 2.332863841758195,
+      "learning_rate": 1.0647591372511427e-05,
+      "loss": 0.6633,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9902065799540933,
+      "grad_norm": 2.289934714096384,
+      "learning_rate": 1.0645118371494847e-05,
+      "loss": 0.6172,
+      "step": 6471
+    },
+    {
+      "epoch": 0.9903596021423107,
+      "grad_norm": 2.2808303258671625,
+      "learning_rate": 1.0642645330859023e-05,
+      "loss": 0.647,
+      "step": 6472
+    },
+    {
+      "epoch": 0.9905126243305279,
+      "grad_norm": 2.001627220573405,
+      "learning_rate": 1.0640172250755836e-05,
+      "loss": 0.6933,
+      "step": 6473
+    },
+    {
+      "epoch": 0.9906656465187452,
+      "grad_norm": 2.133895070860213,
+      "learning_rate": 1.0637699131337167e-05,
+      "loss": 0.6485,
+      "step": 6474
+    },
+    {
+      "epoch": 0.9908186687069626,
+      "grad_norm": 2.306956253050283,
+      "learning_rate": 1.06352259727549e-05,
+      "loss": 0.6595,
+      "step": 6475
+    },
+    {
+      "epoch": 0.9909716908951798,
+      "grad_norm": 2.1617438230826407,
+      "learning_rate": 1.0632752775160917e-05,
+      "loss": 0.5378,
+      "step": 6476
+    },
+    {
+      "epoch": 0.9911247130833971,
+      "grad_norm": 2.5429299085792296,
+      "learning_rate": 1.0630279538707116e-05,
+      "loss": 0.7616,
+      "step": 6477
+    },
+    {
+      "epoch": 0.9912777352716143,
+      "grad_norm": 2.252167453852857,
+      "learning_rate": 1.0627806263545379e-05,
+      "loss": 0.6494,
+      "step": 6478
+    },
+    {
+      "epoch": 0.9914307574598317,
+      "grad_norm": 2.2880745559100775,
+      "learning_rate": 1.0625332949827604e-05,
+      "loss": 0.6506,
+      "step": 6479
+    },
+    {
+      "epoch": 0.991583779648049,
+      "grad_norm": 2.476915922169972,
+      "learning_rate": 1.0622859597705686e-05,
+      "loss": 0.7807,
+      "step": 6480
+    },
+    {
+      "epoch": 0.9917368018362662,
+      "grad_norm": 2.4263691944067034,
+      "learning_rate": 1.0620386207331528e-05,
+      "loss": 0.6673,
+      "step": 6481
+    },
+    {
+      "epoch": 0.9918898240244836,
+      "grad_norm": 2.288249879251959,
+      "learning_rate": 1.0617912778857022e-05,
+      "loss": 0.5908,
+      "step": 6482
+    },
+    {
+      "epoch": 0.9920428462127009,
+      "grad_norm": 2.1329413418096625,
+      "learning_rate": 1.0615439312434073e-05,
+      "loss": 0.6263,
+      "step": 6483
+    },
+    {
+      "epoch": 0.9921958684009181,
+      "grad_norm": 2.3367729594747804,
+      "learning_rate": 1.061296580821459e-05,
+      "loss": 0.6701,
+      "step": 6484
+    },
+    {
+      "epoch": 0.9923488905891354,
+      "grad_norm": 2.0871602655046506,
+      "learning_rate": 1.0610492266350484e-05,
+      "loss": 0.5942,
+      "step": 6485
+    },
+    {
+      "epoch": 0.9925019127773527,
+      "grad_norm": 2.1715923797764236,
+      "learning_rate": 1.0608018686993651e-05,
+      "loss": 0.592,
+      "step": 6486
+    },
+    {
+      "epoch": 0.99265493496557,
+      "grad_norm": 2.1720410847528266,
+      "learning_rate": 1.0605545070296017e-05,
+      "loss": 0.6236,
+      "step": 6487
+    },
+    {
+      "epoch": 0.9928079571537873,
+      "grad_norm": 2.199533292871803,
+      "learning_rate": 1.060307141640949e-05,
+      "loss": 0.5939,
+      "step": 6488
+    },
+    {
+      "epoch": 0.9929609793420046,
+      "grad_norm": 2.314051801230297,
+      "learning_rate": 1.0600597725485988e-05,
+      "loss": 0.6454,
+      "step": 6489
+    },
+    {
+      "epoch": 0.9931140015302219,
+      "grad_norm": 2.414577584604856,
+      "learning_rate": 1.0598123997677432e-05,
+      "loss": 0.5892,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9932670237184392,
+      "grad_norm": 2.220440454756989,
+      "learning_rate": 1.059565023313574e-05,
+      "loss": 0.6345,
+      "step": 6491
+    },
+    {
+      "epoch": 0.9934200459066564,
+      "grad_norm": 2.359228123391197,
+      "learning_rate": 1.0593176432012836e-05,
+      "loss": 0.6315,
+      "step": 6492
+    },
+    {
+      "epoch": 0.9935730680948738,
+      "grad_norm": 2.2007100542708797,
+      "learning_rate": 1.0590702594460649e-05,
+      "loss": 0.672,
+      "step": 6493
+    },
+    {
+      "epoch": 0.993726090283091,
+      "grad_norm": 2.162309831580927,
+      "learning_rate": 1.0588228720631104e-05,
+      "loss": 0.6506,
+      "step": 6494
+    },
+    {
+      "epoch": 0.9938791124713083,
+      "grad_norm": 2.5394139530332698,
+      "learning_rate": 1.0585754810676127e-05,
+      "loss": 0.8196,
+      "step": 6495
+    },
+    {
+      "epoch": 0.9940321346595257,
+      "grad_norm": 2.1750634220281104,
+      "learning_rate": 1.0583280864747663e-05,
+      "loss": 0.5415,
+      "step": 6496
+    },
+    {
+      "epoch": 0.9941851568477429,
+      "grad_norm": 2.056184817961102,
+      "learning_rate": 1.0580806882997638e-05,
+      "loss": 0.5811,
+      "step": 6497
+    },
+    {
+      "epoch": 0.9943381790359602,
+      "grad_norm": 1.9628993588402721,
+      "learning_rate": 1.0578332865577987e-05,
+      "loss": 0.5865,
+      "step": 6498
+    },
+    {
+      "epoch": 0.9944912012241776,
+      "grad_norm": 2.0909315968389572,
+      "learning_rate": 1.0575858812640653e-05,
+      "loss": 0.6559,
+      "step": 6499
+    },
+    {
+      "epoch": 0.9946442234123948,
+      "grad_norm": 2.1491534931389547,
+      "learning_rate": 1.057338472433758e-05,
+      "loss": 0.5262,
+      "step": 6500
+    },
+    {
+      "epoch": 0.9947972456006121,
+      "grad_norm": 2.25180793663901,
+      "learning_rate": 1.0570910600820703e-05,
+      "loss": 0.7352,
+      "step": 6501
+    },
+    {
+      "epoch": 0.9949502677888293,
+      "grad_norm": 2.1499742816028613,
+      "learning_rate": 1.0568436442241975e-05,
+      "loss": 0.6087,
+      "step": 6502
+    },
+    {
+      "epoch": 0.9951032899770467,
+      "grad_norm": 2.202787694955412,
+      "learning_rate": 1.0565962248753344e-05,
+      "loss": 0.6215,
+      "step": 6503
+    },
+    {
+      "epoch": 0.995256312165264,
+      "grad_norm": 2.225784673710967,
+      "learning_rate": 1.0563488020506754e-05,
+      "loss": 0.7141,
+      "step": 6504
+    },
+    {
+      "epoch": 0.9954093343534812,
+      "grad_norm": 2.3329954360329843,
+      "learning_rate": 1.056101375765416e-05,
+      "loss": 0.7444,
+      "step": 6505
+    },
+    {
+      "epoch": 0.9955623565416986,
+      "grad_norm": 2.0139835559989505,
+      "learning_rate": 1.0558539460347518e-05,
+      "loss": 0.5935,
+      "step": 6506
+    },
+    {
+      "epoch": 0.9957153787299159,
+      "grad_norm": 2.263146610045272,
+      "learning_rate": 1.0556065128738782e-05,
+      "loss": 0.6425,
+      "step": 6507
+    },
+    {
+      "epoch": 0.9958684009181331,
+      "grad_norm": 2.2635096597755426,
+      "learning_rate": 1.055359076297991e-05,
+      "loss": 0.6783,
+      "step": 6508
+    },
+    {
+      "epoch": 0.9960214231063504,
+      "grad_norm": 2.2452785958982764,
+      "learning_rate": 1.0551116363222864e-05,
+      "loss": 0.8419,
+      "step": 6509
+    },
+    {
+      "epoch": 0.9961744452945677,
+      "grad_norm": 2.447402852546272,
+      "learning_rate": 1.054864192961961e-05,
+      "loss": 0.6773,
+      "step": 6510
+    },
+    {
+      "epoch": 0.996327467482785,
+      "grad_norm": 1.9661543729169815,
+      "learning_rate": 1.0546167462322103e-05,
+      "loss": 0.5886,
+      "step": 6511
+    },
+    {
+      "epoch": 0.9964804896710023,
+      "grad_norm": 2.0809576482884693,
+      "learning_rate": 1.054369296148232e-05,
+      "loss": 0.6409,
+      "step": 6512
+    },
+    {
+      "epoch": 0.9966335118592196,
+      "grad_norm": 1.9729856131001111,
+      "learning_rate": 1.0541218427252222e-05,
+      "loss": 0.6166,
+      "step": 6513
+    },
+    {
+      "epoch": 0.9967865340474369,
+      "grad_norm": 2.206970740539544,
+      "learning_rate": 1.0538743859783787e-05,
+      "loss": 0.6654,
+      "step": 6514
+    },
+    {
+      "epoch": 0.9969395562356542,
+      "grad_norm": 2.0815380641893793,
+      "learning_rate": 1.053626925922898e-05,
+      "loss": 0.564,
+      "step": 6515
+    },
+    {
+      "epoch": 0.9970925784238714,
+      "grad_norm": 2.043032926313095,
+      "learning_rate": 1.0533794625739782e-05,
+      "loss": 0.6125,
+      "step": 6516
+    },
+    {
+      "epoch": 0.9972456006120888,
+      "grad_norm": 2.08135336365385,
+      "learning_rate": 1.0531319959468167e-05,
+      "loss": 0.5659,
+      "step": 6517
+    },
+    {
+      "epoch": 0.997398622800306,
+      "grad_norm": 2.0383774964273287,
+      "learning_rate": 1.0528845260566116e-05,
+      "loss": 0.5637,
+      "step": 6518
+    },
+    {
+      "epoch": 0.9975516449885233,
+      "grad_norm": 2.066268610798444,
+      "learning_rate": 1.0526370529185608e-05,
+      "loss": 0.6478,
+      "step": 6519
+    },
+    {
+      "epoch": 0.9977046671767407,
+      "grad_norm": 2.5050073606635817,
+      "learning_rate": 1.0523895765478627e-05,
+      "loss": 0.6566,
+      "step": 6520
+    },
+    {
+      "epoch": 0.9978576893649579,
+      "grad_norm": 2.5001747882359084,
+      "learning_rate": 1.0521420969597156e-05,
+      "loss": 0.6971,
+      "step": 6521
+    },
+    {
+      "epoch": 0.9980107115531752,
+      "grad_norm": 2.21713217988516,
+      "learning_rate": 1.0518946141693185e-05,
+      "loss": 0.5671,
+      "step": 6522
+    },
+    {
+      "epoch": 0.9981637337413926,
+      "grad_norm": 2.216049429630443,
+      "learning_rate": 1.0516471281918702e-05,
+      "loss": 0.6914,
+      "step": 6523
+    },
+    {
+      "epoch": 0.9983167559296098,
+      "grad_norm": 2.2994572356248106,
+      "learning_rate": 1.0513996390425694e-05,
+      "loss": 0.6335,
+      "step": 6524
+    },
+    {
+      "epoch": 0.9984697781178271,
+      "grad_norm": 2.1282739516172238,
+      "learning_rate": 1.051152146736616e-05,
+      "loss": 0.5977,
+      "step": 6525
+    },
+    {
+      "epoch": 0.9986228003060443,
+      "grad_norm": 2.4591726916231615,
+      "learning_rate": 1.0509046512892092e-05,
+      "loss": 0.684,
+      "step": 6526
+    },
+    {
+      "epoch": 0.9987758224942617,
+      "grad_norm": 2.64549225493789,
+      "learning_rate": 1.0506571527155482e-05,
+      "loss": 0.6533,
+      "step": 6527
+    },
+    {
+      "epoch": 0.998928844682479,
+      "grad_norm": 2.3069809469331384,
+      "learning_rate": 1.0504096510308335e-05,
+      "loss": 0.601,
+      "step": 6528
+    },
+    {
+      "epoch": 0.9990818668706962,
+      "grad_norm": 2.144064608208212,
+      "learning_rate": 1.0501621462502651e-05,
+      "loss": 0.6259,
+      "step": 6529
+    },
+    {
+      "epoch": 0.9992348890589136,
+      "grad_norm": 2.4104646760425394,
+      "learning_rate": 1.049914638389043e-05,
+      "loss": 0.7642,
+      "step": 6530
+    },
+    {
+      "epoch": 0.9993879112471309,
+      "grad_norm": 2.048851027289111,
+      "learning_rate": 1.0496671274623675e-05,
+      "loss": 0.6663,
+      "step": 6531
+    },
+    {
+      "epoch": 0.9995409334353481,
+      "grad_norm": 2.531917512682566,
+      "learning_rate": 1.0494196134854395e-05,
+      "loss": 0.8121,
+      "step": 6532
+    },
+    {
+      "epoch": 0.9996939556235654,
+      "grad_norm": 2.7016640424743534,
+      "learning_rate": 1.0491720964734595e-05,
+      "loss": 0.7759,
+      "step": 6533
+    },
+    {
+      "epoch": 0.9998469778117827,
+      "grad_norm": 2.227959878473994,
+      "learning_rate": 1.0489245764416294e-05,
+      "loss": 0.6892,
+      "step": 6534
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2604243565912374,
+      "learning_rate": 1.0486770534051492e-05,
+      "loss": 0.7176,
+      "step": 6535
+    },
+    {
+      "epoch": 1.0001530221882173,
+      "grad_norm": 2.033100428541894,
+      "learning_rate": 1.0484295273792204e-05,
+      "loss": 0.6903,
+      "step": 6536
+    },
+    {
+      "epoch": 1.0003060443764347,
+      "grad_norm": 2.1314696539623275,
+      "learning_rate": 1.0481819983790454e-05,
+      "loss": 0.5955,
+      "step": 6537
+    },
+    {
+      "epoch": 1.0004590665646518,
+      "grad_norm": 2.5135837640358405,
+      "learning_rate": 1.0479344664198251e-05,
+      "loss": 0.6917,
+      "step": 6538
+    },
+    {
+      "epoch": 1.000612088752869,
+      "grad_norm": 2.133925287881583,
+      "learning_rate": 1.0476869315167617e-05,
+      "loss": 0.616,
+      "step": 6539
+    },
+    {
+      "epoch": 1.0007651109410864,
+      "grad_norm": 2.1679524566869377,
+      "learning_rate": 1.0474393936850573e-05,
+      "loss": 0.7227,
+      "step": 6540
+    },
+    {
+      "epoch": 1.0009181331293038,
+      "grad_norm": 2.3980955664439105,
+      "learning_rate": 1.0471918529399143e-05,
+      "loss": 0.7499,
+      "step": 6541
+    },
+    {
+      "epoch": 1.001071155317521,
+      "grad_norm": 2.3699351005706255,
+      "learning_rate": 1.046944309296535e-05,
+      "loss": 0.6967,
+      "step": 6542
+    },
+    {
+      "epoch": 1.0012241775057382,
+      "grad_norm": 2.647323810066508,
+      "learning_rate": 1.0466967627701219e-05,
+      "loss": 0.7663,
+      "step": 6543
+    },
+    {
+      "epoch": 1.0013771996939556,
+      "grad_norm": 2.137859552412195,
+      "learning_rate": 1.0464492133758779e-05,
+      "loss": 0.5863,
+      "step": 6544
+    },
+    {
+      "epoch": 1.0015302218821729,
+      "grad_norm": 2.0219765558236285,
+      "learning_rate": 1.046201661129006e-05,
+      "loss": 0.6693,
+      "step": 6545
+    },
+    {
+      "epoch": 1.0016832440703902,
+      "grad_norm": 2.4366531420683137,
+      "learning_rate": 1.0459541060447091e-05,
+      "loss": 0.703,
+      "step": 6546
+    },
+    {
+      "epoch": 1.0018362662586076,
+      "grad_norm": 1.9828507027857656,
+      "learning_rate": 1.0457065481381913e-05,
+      "loss": 0.582,
+      "step": 6547
+    },
+    {
+      "epoch": 1.0019892884468249,
+      "grad_norm": 2.0692223786471624,
+      "learning_rate": 1.0454589874246555e-05,
+      "loss": 0.5675,
+      "step": 6548
+    },
+    {
+      "epoch": 1.002142310635042,
+      "grad_norm": 1.9859837618858298,
+      "learning_rate": 1.0452114239193049e-05,
+      "loss": 0.5558,
+      "step": 6549
+    },
+    {
+      "epoch": 1.0022953328232593,
+      "grad_norm": 2.272954348610601,
+      "learning_rate": 1.0449638576373445e-05,
+      "loss": 0.6769,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0024483550114767,
+      "grad_norm": 2.420602351894704,
+      "learning_rate": 1.0447162885939776e-05,
+      "loss": 0.6877,
+      "step": 6551
+    },
+    {
+      "epoch": 1.002601377199694,
+      "grad_norm": 2.494825351434444,
+      "learning_rate": 1.0444687168044082e-05,
+      "loss": 0.7181,
+      "step": 6552
+    },
+    {
+      "epoch": 1.0027543993879113,
+      "grad_norm": 2.285517189923609,
+      "learning_rate": 1.044221142283841e-05,
+      "loss": 0.6558,
+      "step": 6553
+    },
+    {
+      "epoch": 1.0029074215761284,
+      "grad_norm": 2.044150801571359,
+      "learning_rate": 1.0439735650474807e-05,
+      "loss": 0.5108,
+      "step": 6554
+    },
+    {
+      "epoch": 1.0030604437643458,
+      "grad_norm": 2.26517016730885,
+      "learning_rate": 1.0437259851105318e-05,
+      "loss": 0.6558,
+      "step": 6555
+    },
+    {
+      "epoch": 1.003213465952563,
+      "grad_norm": 2.24291348116444,
+      "learning_rate": 1.0434784024881988e-05,
+      "loss": 0.6491,
+      "step": 6556
+    },
+    {
+      "epoch": 1.0033664881407804,
+      "grad_norm": 2.187640461685535,
+      "learning_rate": 1.043230817195687e-05,
+      "loss": 0.6019,
+      "step": 6557
+    },
+    {
+      "epoch": 1.0035195103289978,
+      "grad_norm": 2.3819794343778278,
+      "learning_rate": 1.0429832292482019e-05,
+      "loss": 0.7496,
+      "step": 6558
+    },
+    {
+      "epoch": 1.0036725325172149,
+      "grad_norm": 2.1489073395127356,
+      "learning_rate": 1.0427356386609482e-05,
+      "loss": 0.6125,
+      "step": 6559
+    },
+    {
+      "epoch": 1.0038255547054322,
+      "grad_norm": 2.0082206501261686,
+      "learning_rate": 1.042488045449132e-05,
+      "loss": 0.5703,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0039785768936496,
+      "grad_norm": 2.142001534829241,
+      "learning_rate": 1.0422404496279585e-05,
+      "loss": 0.5988,
+      "step": 6561
+    },
+    {
+      "epoch": 1.0041315990818669,
+      "grad_norm": 2.1385641923885648,
+      "learning_rate": 1.0419928512126337e-05,
+      "loss": 0.6956,
+      "step": 6562
+    },
+    {
+      "epoch": 1.0042846212700842,
+      "grad_norm": 2.3824901717985436,
+      "learning_rate": 1.041745250218364e-05,
+      "loss": 0.6451,
+      "step": 6563
+    },
+    {
+      "epoch": 1.0044376434583016,
+      "grad_norm": 2.4015296436629776,
+      "learning_rate": 1.0414976466603551e-05,
+      "loss": 0.7278,
+      "step": 6564
+    },
+    {
+      "epoch": 1.0045906656465187,
+      "grad_norm": 2.1534435564300347,
+      "learning_rate": 1.0412500405538129e-05,
+      "loss": 0.6363,
+      "step": 6565
+    },
+    {
+      "epoch": 1.004743687834736,
+      "grad_norm": 2.48153539686987,
+      "learning_rate": 1.0410024319139447e-05,
+      "loss": 0.6363,
+      "step": 6566
+    },
+    {
+      "epoch": 1.0048967100229533,
+      "grad_norm": 2.172810737488868,
+      "learning_rate": 1.040754820755957e-05,
+      "loss": 0.6484,
+      "step": 6567
+    },
+    {
+      "epoch": 1.0050497322111707,
+      "grad_norm": 2.5460489768001144,
+      "learning_rate": 1.040507207095056e-05,
+      "loss": 0.8215,
+      "step": 6568
+    },
+    {
+      "epoch": 1.005202754399388,
+      "grad_norm": 2.16079862891005,
+      "learning_rate": 1.0402595909464489e-05,
+      "loss": 0.5477,
+      "step": 6569
+    },
+    {
+      "epoch": 1.005355776587605,
+      "grad_norm": 2.202814891087014,
+      "learning_rate": 1.0400119723253428e-05,
+      "loss": 0.5666,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0055087987758224,
+      "grad_norm": 2.212828269877084,
+      "learning_rate": 1.039764351246945e-05,
+      "loss": 0.5264,
+      "step": 6571
+    },
+    {
+      "epoch": 1.0056618209640398,
+      "grad_norm": 2.117103171907091,
+      "learning_rate": 1.039516727726463e-05,
+      "loss": 0.5496,
+      "step": 6572
+    },
+    {
+      "epoch": 1.005814843152257,
+      "grad_norm": 2.2084826072189996,
+      "learning_rate": 1.0392691017791041e-05,
+      "loss": 0.5141,
+      "step": 6573
+    },
+    {
+      "epoch": 1.0059678653404744,
+      "grad_norm": 2.32130852843778,
+      "learning_rate": 1.0390214734200758e-05,
+      "loss": 0.6334,
+      "step": 6574
+    },
+    {
+      "epoch": 1.0061208875286916,
+      "grad_norm": 2.2964584356167284,
+      "learning_rate": 1.0387738426645865e-05,
+      "loss": 0.6497,
+      "step": 6575
+    },
+    {
+      "epoch": 1.0062739097169089,
+      "grad_norm": 2.4158826799520434,
+      "learning_rate": 1.0385262095278438e-05,
+      "loss": 0.6641,
+      "step": 6576
+    },
+    {
+      "epoch": 1.0064269319051262,
+      "grad_norm": 2.167843641637399,
+      "learning_rate": 1.0382785740250557e-05,
+      "loss": 0.5314,
+      "step": 6577
+    },
+    {
+      "epoch": 1.0065799540933436,
+      "grad_norm": 2.283231101433695,
+      "learning_rate": 1.0380309361714306e-05,
+      "loss": 0.6207,
+      "step": 6578
+    },
+    {
+      "epoch": 1.0067329762815609,
+      "grad_norm": 2.536910953794421,
+      "learning_rate": 1.0377832959821773e-05,
+      "loss": 0.7732,
+      "step": 6579
+    },
+    {
+      "epoch": 1.0068859984697782,
+      "grad_norm": 2.194479016477561,
+      "learning_rate": 1.0375356534725036e-05,
+      "loss": 0.6016,
+      "step": 6580
+    },
+    {
+      "epoch": 1.0070390206579953,
+      "grad_norm": 2.568553979829469,
+      "learning_rate": 1.0372880086576187e-05,
+      "loss": 0.5968,
+      "step": 6581
+    },
+    {
+      "epoch": 1.0071920428462127,
+      "grad_norm": 2.2861022154819017,
+      "learning_rate": 1.0370403615527311e-05,
+      "loss": 0.6048,
+      "step": 6582
+    },
+    {
+      "epoch": 1.00734506503443,
+      "grad_norm": 2.302797866266313,
+      "learning_rate": 1.0367927121730503e-05,
+      "loss": 0.622,
+      "step": 6583
+    },
+    {
+      "epoch": 1.0074980872226473,
+      "grad_norm": 2.0879458077730715,
+      "learning_rate": 1.036545060533785e-05,
+      "loss": 0.5533,
+      "step": 6584
+    },
+    {
+      "epoch": 1.0076511094108647,
+      "grad_norm": 2.557603243001996,
+      "learning_rate": 1.0362974066501445e-05,
+      "loss": 0.7397,
+      "step": 6585
+    },
+    {
+      "epoch": 1.0078041315990818,
+      "grad_norm": 2.312252735085079,
+      "learning_rate": 1.0360497505373386e-05,
+      "loss": 0.7411,
+      "step": 6586
+    },
+    {
+      "epoch": 1.007957153787299,
+      "grad_norm": 2.3173357806675736,
+      "learning_rate": 1.035802092210576e-05,
+      "loss": 0.6168,
+      "step": 6587
+    },
+    {
+      "epoch": 1.0081101759755164,
+      "grad_norm": 2.2189725542214678,
+      "learning_rate": 1.0355544316850671e-05,
+      "loss": 0.6329,
+      "step": 6588
+    },
+    {
+      "epoch": 1.0082631981637338,
+      "grad_norm": 2.204599066701292,
+      "learning_rate": 1.0353067689760217e-05,
+      "loss": 0.5969,
+      "step": 6589
+    },
+    {
+      "epoch": 1.008416220351951,
+      "grad_norm": 2.6794556741279414,
+      "learning_rate": 1.0350591040986489e-05,
+      "loss": 0.6986,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0085692425401682,
+      "grad_norm": 2.3559628832205797,
+      "learning_rate": 1.03481143706816e-05,
+      "loss": 0.6133,
+      "step": 6591
+    },
+    {
+      "epoch": 1.0087222647283856,
+      "grad_norm": 2.759684199213884,
+      "learning_rate": 1.0345637678997642e-05,
+      "loss": 0.6948,
+      "step": 6592
+    },
+    {
+      "epoch": 1.0088752869166029,
+      "grad_norm": 2.051050778484884,
+      "learning_rate": 1.0343160966086723e-05,
+      "loss": 0.6322,
+      "step": 6593
+    },
+    {
+      "epoch": 1.0090283091048202,
+      "grad_norm": 2.375738470682142,
+      "learning_rate": 1.0340684232100946e-05,
+      "loss": 0.6488,
+      "step": 6594
+    },
+    {
+      "epoch": 1.0091813312930376,
+      "grad_norm": 2.2479561575153966,
+      "learning_rate": 1.0338207477192417e-05,
+      "loss": 0.6915,
+      "step": 6595
+    },
+    {
+      "epoch": 1.0093343534812549,
+      "grad_norm": 2.1508005625889686,
+      "learning_rate": 1.0335730701513245e-05,
+      "loss": 0.5837,
+      "step": 6596
+    },
+    {
+      "epoch": 1.009487375669472,
+      "grad_norm": 2.4929163407137604,
+      "learning_rate": 1.0333253905215537e-05,
+      "loss": 0.7049,
+      "step": 6597
+    },
+    {
+      "epoch": 1.0096403978576893,
+      "grad_norm": 2.322330911445932,
+      "learning_rate": 1.0330777088451402e-05,
+      "loss": 0.6233,
+      "step": 6598
+    },
+    {
+      "epoch": 1.0097934200459067,
+      "grad_norm": 2.283433120066431,
+      "learning_rate": 1.0328300251372954e-05,
+      "loss": 0.5305,
+      "step": 6599
+    },
+    {
+      "epoch": 1.009946442234124,
+      "grad_norm": 2.17015984089956,
+      "learning_rate": 1.0325823394132301e-05,
+      "loss": 0.5958,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0100994644223413,
+      "grad_norm": 2.436120696696377,
+      "learning_rate": 1.032334651688156e-05,
+      "loss": 0.6444,
+      "step": 6601
+    },
+    {
+      "epoch": 1.0102524866105584,
+      "grad_norm": 2.3607087470720938,
+      "learning_rate": 1.0320869619772847e-05,
+      "loss": 0.6319,
+      "step": 6602
+    },
+    {
+      "epoch": 1.0104055087987758,
+      "grad_norm": 2.2905924990988273,
+      "learning_rate": 1.0318392702958268e-05,
+      "loss": 0.6301,
+      "step": 6603
+    },
+    {
+      "epoch": 1.010558530986993,
+      "grad_norm": 2.44559146610296,
+      "learning_rate": 1.0315915766589957e-05,
+      "loss": 0.6883,
+      "step": 6604
+    },
+    {
+      "epoch": 1.0107115531752104,
+      "grad_norm": 2.4650413110292977,
+      "learning_rate": 1.0313438810820018e-05,
+      "loss": 0.5647,
+      "step": 6605
+    },
+    {
+      "epoch": 1.0108645753634278,
+      "grad_norm": 2.1910260377940567,
+      "learning_rate": 1.0310961835800572e-05,
+      "loss": 0.577,
+      "step": 6606
+    },
+    {
+      "epoch": 1.0110175975516449,
+      "grad_norm": 2.38892339041073,
+      "learning_rate": 1.0308484841683749e-05,
+      "loss": 0.5909,
+      "step": 6607
+    },
+    {
+      "epoch": 1.0111706197398622,
+      "grad_norm": 2.4041673025346832,
+      "learning_rate": 1.0306007828621665e-05,
+      "loss": 0.6755,
+      "step": 6608
+    },
+    {
+      "epoch": 1.0113236419280796,
+      "grad_norm": 2.2844546507484287,
+      "learning_rate": 1.0303530796766439e-05,
+      "loss": 0.5774,
+      "step": 6609
+    },
+    {
+      "epoch": 1.0114766641162969,
+      "grad_norm": 2.116403540663527,
+      "learning_rate": 1.03010537462702e-05,
+      "loss": 0.515,
+      "step": 6610
+    },
+    {
+      "epoch": 1.0116296863045142,
+      "grad_norm": 2.1836805111004374,
+      "learning_rate": 1.0298576677285076e-05,
+      "loss": 0.5105,
+      "step": 6611
+    },
+    {
+      "epoch": 1.0117827084927316,
+      "grad_norm": 2.507796655419912,
+      "learning_rate": 1.0296099589963185e-05,
+      "loss": 0.6309,
+      "step": 6612
+    },
+    {
+      "epoch": 1.0119357306809487,
+      "grad_norm": 2.0527464335910413,
+      "learning_rate": 1.0293622484456665e-05,
+      "loss": 0.6004,
+      "step": 6613
+    },
+    {
+      "epoch": 1.012088752869166,
+      "grad_norm": 2.334226311405419,
+      "learning_rate": 1.0291145360917639e-05,
+      "loss": 0.5387,
+      "step": 6614
+    },
+    {
+      "epoch": 1.0122417750573833,
+      "grad_norm": 2.204170110208716,
+      "learning_rate": 1.0288668219498232e-05,
+      "loss": 0.6759,
+      "step": 6615
+    },
+    {
+      "epoch": 1.0123947972456007,
+      "grad_norm": 2.3108595745215776,
+      "learning_rate": 1.0286191060350586e-05,
+      "loss": 0.6272,
+      "step": 6616
+    },
+    {
+      "epoch": 1.012547819433818,
+      "grad_norm": 2.2782166898829077,
+      "learning_rate": 1.0283713883626829e-05,
+      "loss": 0.6998,
+      "step": 6617
+    },
+    {
+      "epoch": 1.012700841622035,
+      "grad_norm": 2.2993109109204903,
+      "learning_rate": 1.0281236689479086e-05,
+      "loss": 0.6204,
+      "step": 6618
+    },
+    {
+      "epoch": 1.0128538638102524,
+      "grad_norm": 2.3404054714266582,
+      "learning_rate": 1.0278759478059502e-05,
+      "loss": 0.6709,
+      "step": 6619
+    },
+    {
+      "epoch": 1.0130068859984698,
+      "grad_norm": 2.5884362282825637,
+      "learning_rate": 1.0276282249520207e-05,
+      "loss": 0.7326,
+      "step": 6620
+    },
+    {
+      "epoch": 1.013159908186687,
+      "grad_norm": 1.9300570634216025,
+      "learning_rate": 1.0273805004013337e-05,
+      "loss": 0.5012,
+      "step": 6621
+    },
+    {
+      "epoch": 1.0133129303749044,
+      "grad_norm": 2.377275021865992,
+      "learning_rate": 1.0271327741691032e-05,
+      "loss": 0.6814,
+      "step": 6622
+    },
+    {
+      "epoch": 1.0134659525631216,
+      "grad_norm": 2.182477995481934,
+      "learning_rate": 1.0268850462705431e-05,
+      "loss": 0.5598,
+      "step": 6623
+    },
+    {
+      "epoch": 1.0136189747513389,
+      "grad_norm": 2.3066689578762207,
+      "learning_rate": 1.0266373167208668e-05,
+      "loss": 0.6129,
+      "step": 6624
+    },
+    {
+      "epoch": 1.0137719969395562,
+      "grad_norm": 2.307693232138134,
+      "learning_rate": 1.0263895855352887e-05,
+      "loss": 0.5754,
+      "step": 6625
+    },
+    {
+      "epoch": 1.0139250191277736,
+      "grad_norm": 2.292009668838204,
+      "learning_rate": 1.0261418527290233e-05,
+      "loss": 0.6121,
+      "step": 6626
+    },
+    {
+      "epoch": 1.0140780413159909,
+      "grad_norm": 2.492568937733668,
+      "learning_rate": 1.025894118317284e-05,
+      "loss": 0.6629,
+      "step": 6627
+    },
+    {
+      "epoch": 1.0142310635042082,
+      "grad_norm": 2.2340661216832096,
+      "learning_rate": 1.0256463823152855e-05,
+      "loss": 0.5866,
+      "step": 6628
+    },
+    {
+      "epoch": 1.0143840856924253,
+      "grad_norm": 2.110334526976937,
+      "learning_rate": 1.0253986447382429e-05,
+      "loss": 0.5203,
+      "step": 6629
+    },
+    {
+      "epoch": 1.0145371078806427,
+      "grad_norm": 2.692658979279218,
+      "learning_rate": 1.0251509056013697e-05,
+      "loss": 0.5705,
+      "step": 6630
+    },
+    {
+      "epoch": 1.01469013006886,
+      "grad_norm": 2.428970298715567,
+      "learning_rate": 1.024903164919881e-05,
+      "loss": 0.6109,
+      "step": 6631
+    },
+    {
+      "epoch": 1.0148431522570773,
+      "grad_norm": 2.238146074078266,
+      "learning_rate": 1.0246554227089918e-05,
+      "loss": 0.6261,
+      "step": 6632
+    },
+    {
+      "epoch": 1.0149961744452947,
+      "grad_norm": 2.573325893114674,
+      "learning_rate": 1.0244076789839169e-05,
+      "loss": 0.6369,
+      "step": 6633
+    },
+    {
+      "epoch": 1.0151491966335118,
+      "grad_norm": 2.3159256199731053,
+      "learning_rate": 1.0241599337598702e-05,
+      "loss": 0.5916,
+      "step": 6634
+    },
+    {
+      "epoch": 1.015302218821729,
+      "grad_norm": 2.2308322352798444,
+      "learning_rate": 1.0239121870520679e-05,
+      "loss": 0.5703,
+      "step": 6635
+    },
+    {
+      "epoch": 1.0154552410099464,
+      "grad_norm": 2.329808400919156,
+      "learning_rate": 1.0236644388757245e-05,
+      "loss": 0.6378,
+      "step": 6636
+    },
+    {
+      "epoch": 1.0156082631981638,
+      "grad_norm": 2.435679508309255,
+      "learning_rate": 1.0234166892460554e-05,
+      "loss": 0.6254,
+      "step": 6637
+    },
+    {
+      "epoch": 1.015761285386381,
+      "grad_norm": 2.4224361592010295,
+      "learning_rate": 1.0231689381782756e-05,
+      "loss": 0.5599,
+      "step": 6638
+    },
+    {
+      "epoch": 1.0159143075745982,
+      "grad_norm": 2.1885969206768032,
+      "learning_rate": 1.0229211856876011e-05,
+      "loss": 0.5952,
+      "step": 6639
+    },
+    {
+      "epoch": 1.0160673297628156,
+      "grad_norm": 2.3897857156470255,
+      "learning_rate": 1.0226734317892466e-05,
+      "loss": 0.5323,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0162203519510329,
+      "grad_norm": 2.092213659667365,
+      "learning_rate": 1.0224256764984282e-05,
+      "loss": 0.4703,
+      "step": 6641
+    },
+    {
+      "epoch": 1.0163733741392502,
+      "grad_norm": 2.1854404332671185,
+      "learning_rate": 1.022177919830361e-05,
+      "loss": 0.6041,
+      "step": 6642
+    },
+    {
+      "epoch": 1.0165263963274676,
+      "grad_norm": 2.4022816997942593,
+      "learning_rate": 1.0219301618002611e-05,
+      "loss": 0.64,
+      "step": 6643
+    },
+    {
+      "epoch": 1.0166794185156847,
+      "grad_norm": 2.3801526159134982,
+      "learning_rate": 1.021682402423344e-05,
+      "loss": 0.5524,
+      "step": 6644
+    },
+    {
+      "epoch": 1.016832440703902,
+      "grad_norm": 2.218696255320501,
+      "learning_rate": 1.021434641714826e-05,
+      "loss": 0.6162,
+      "step": 6645
+    },
+    {
+      "epoch": 1.0169854628921193,
+      "grad_norm": 2.7939201783520056,
+      "learning_rate": 1.0211868796899229e-05,
+      "loss": 0.6066,
+      "step": 6646
+    },
+    {
+      "epoch": 1.0171384850803367,
+      "grad_norm": 2.264811330544672,
+      "learning_rate": 1.0209391163638503e-05,
+      "loss": 0.539,
+      "step": 6647
+    },
+    {
+      "epoch": 1.017291507268554,
+      "grad_norm": 2.318419121835664,
+      "learning_rate": 1.0206913517518246e-05,
+      "loss": 0.6371,
+      "step": 6648
+    },
+    {
+      "epoch": 1.0174445294567713,
+      "grad_norm": 2.206422409859083,
+      "learning_rate": 1.0204435858690625e-05,
+      "loss": 0.5619,
+      "step": 6649
+    },
+    {
+      "epoch": 1.0175975516449884,
+      "grad_norm": 2.626006555589893,
+      "learning_rate": 1.0201958187307794e-05,
+      "loss": 0.5531,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0177505738332058,
+      "grad_norm": 2.344140763614317,
+      "learning_rate": 1.0199480503521924e-05,
+      "loss": 0.6345,
+      "step": 6651
+    },
+    {
+      "epoch": 1.017903596021423,
+      "grad_norm": 2.2889519250935177,
+      "learning_rate": 1.0197002807485175e-05,
+      "loss": 0.5799,
+      "step": 6652
+    },
+    {
+      "epoch": 1.0180566182096404,
+      "grad_norm": 2.6253194916546154,
+      "learning_rate": 1.0194525099349708e-05,
+      "loss": 0.6793,
+      "step": 6653
+    },
+    {
+      "epoch": 1.0182096403978578,
+      "grad_norm": 2.3472924349764344,
+      "learning_rate": 1.01920473792677e-05,
+      "loss": 0.5215,
+      "step": 6654
+    },
+    {
+      "epoch": 1.0183626625860749,
+      "grad_norm": 2.22200896592304,
+      "learning_rate": 1.0189569647391308e-05,
+      "loss": 0.5543,
+      "step": 6655
+    },
+    {
+      "epoch": 1.0185156847742922,
+      "grad_norm": 2.0095456114314585,
+      "learning_rate": 1.0187091903872703e-05,
+      "loss": 0.5121,
+      "step": 6656
+    },
+    {
+      "epoch": 1.0186687069625096,
+      "grad_norm": 2.297998178865159,
+      "learning_rate": 1.0184614148864052e-05,
+      "loss": 0.584,
+      "step": 6657
+    },
+    {
+      "epoch": 1.0188217291507269,
+      "grad_norm": 2.332497200625412,
+      "learning_rate": 1.0182136382517526e-05,
+      "loss": 0.5356,
+      "step": 6658
+    },
+    {
+      "epoch": 1.0189747513389442,
+      "grad_norm": 2.4417067324261064,
+      "learning_rate": 1.017965860498529e-05,
+      "loss": 0.543,
+      "step": 6659
+    },
+    {
+      "epoch": 1.0191277735271613,
+      "grad_norm": 2.15499036744487,
+      "learning_rate": 1.0177180816419516e-05,
+      "loss": 0.5874,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0192807957153787,
+      "grad_norm": 2.108874428758097,
+      "learning_rate": 1.0174703016972376e-05,
+      "loss": 0.4502,
+      "step": 6661
+    },
+    {
+      "epoch": 1.019433817903596,
+      "grad_norm": 2.548469773857961,
+      "learning_rate": 1.017222520679604e-05,
+      "loss": 0.5865,
+      "step": 6662
+    },
+    {
+      "epoch": 1.0195868400918133,
+      "grad_norm": 2.513582400490902,
+      "learning_rate": 1.0169747386042681e-05,
+      "loss": 0.653,
+      "step": 6663
+    },
+    {
+      "epoch": 1.0197398622800307,
+      "grad_norm": 2.492061054460402,
+      "learning_rate": 1.0167269554864472e-05,
+      "loss": 0.6406,
+      "step": 6664
+    },
+    {
+      "epoch": 1.019892884468248,
+      "grad_norm": 2.180717896254185,
+      "learning_rate": 1.0164791713413583e-05,
+      "loss": 0.5883,
+      "step": 6665
+    },
+    {
+      "epoch": 1.020045906656465,
+      "grad_norm": 2.2006717979763786,
+      "learning_rate": 1.016231386184219e-05,
+      "loss": 0.5509,
+      "step": 6666
+    },
+    {
+      "epoch": 1.0201989288446824,
+      "grad_norm": 2.3252357391678586,
+      "learning_rate": 1.015983600030247e-05,
+      "loss": 0.6702,
+      "step": 6667
+    },
+    {
+      "epoch": 1.0203519510328998,
+      "grad_norm": 2.14665930839897,
+      "learning_rate": 1.0157358128946596e-05,
+      "loss": 0.5438,
+      "step": 6668
+    },
+    {
+      "epoch": 1.020504973221117,
+      "grad_norm": 2.3069969574975002,
+      "learning_rate": 1.0154880247926739e-05,
+      "loss": 0.5291,
+      "step": 6669
+    },
+    {
+      "epoch": 1.0206579954093344,
+      "grad_norm": 2.413058002612041,
+      "learning_rate": 1.0152402357395086e-05,
+      "loss": 0.5105,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0208110175975516,
+      "grad_norm": 2.272849768241214,
+      "learning_rate": 1.0149924457503806e-05,
+      "loss": 0.6272,
+      "step": 6671
+    },
+    {
+      "epoch": 1.0209640397857689,
+      "grad_norm": 2.122826286435337,
+      "learning_rate": 1.0147446548405076e-05,
+      "loss": 0.516,
+      "step": 6672
+    },
+    {
+      "epoch": 1.0211170619739862,
+      "grad_norm": 2.329647854250823,
+      "learning_rate": 1.0144968630251078e-05,
+      "loss": 0.5394,
+      "step": 6673
+    },
+    {
+      "epoch": 1.0212700841622036,
+      "grad_norm": 2.1569395923518138,
+      "learning_rate": 1.0142490703193992e-05,
+      "loss": 0.6182,
+      "step": 6674
+    },
+    {
+      "epoch": 1.0214231063504209,
+      "grad_norm": 2.241545575430639,
+      "learning_rate": 1.014001276738599e-05,
+      "loss": 0.5988,
+      "step": 6675
+    },
+    {
+      "epoch": 1.021576128538638,
+      "grad_norm": 2.129116623058144,
+      "learning_rate": 1.0137534822979258e-05,
+      "loss": 0.539,
+      "step": 6676
+    },
+    {
+      "epoch": 1.0217291507268553,
+      "grad_norm": 2.256824403314073,
+      "learning_rate": 1.0135056870125976e-05,
+      "loss": 0.6206,
+      "step": 6677
+    },
+    {
+      "epoch": 1.0218821729150727,
+      "grad_norm": 2.354857212906966,
+      "learning_rate": 1.0132578908978323e-05,
+      "loss": 0.6072,
+      "step": 6678
+    },
+    {
+      "epoch": 1.02203519510329,
+      "grad_norm": 2.2770225052597994,
+      "learning_rate": 1.0130100939688478e-05,
+      "loss": 0.6071,
+      "step": 6679
+    },
+    {
+      "epoch": 1.0221882172915073,
+      "grad_norm": 2.1691193316787336,
+      "learning_rate": 1.012762296240863e-05,
+      "loss": 0.5668,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0223412394797247,
+      "grad_norm": 2.138908409021962,
+      "learning_rate": 1.0125144977290952e-05,
+      "loss": 0.5591,
+      "step": 6681
+    },
+    {
+      "epoch": 1.0224942616679418,
+      "grad_norm": 2.3379186543767854,
+      "learning_rate": 1.0122666984487632e-05,
+      "loss": 0.5591,
+      "step": 6682
+    },
+    {
+      "epoch": 1.022647283856159,
+      "grad_norm": 2.480112583086363,
+      "learning_rate": 1.0120188984150857e-05,
+      "loss": 0.6374,
+      "step": 6683
+    },
+    {
+      "epoch": 1.0228003060443764,
+      "grad_norm": 2.2457431345148935,
+      "learning_rate": 1.0117710976432802e-05,
+      "loss": 0.603,
+      "step": 6684
+    },
+    {
+      "epoch": 1.0229533282325938,
+      "grad_norm": 2.3766699997888945,
+      "learning_rate": 1.0115232961485655e-05,
+      "loss": 0.553,
+      "step": 6685
+    },
+    {
+      "epoch": 1.023106350420811,
+      "grad_norm": 2.1244860650080652,
+      "learning_rate": 1.0112754939461603e-05,
+      "loss": 0.5574,
+      "step": 6686
+    },
+    {
+      "epoch": 1.0232593726090282,
+      "grad_norm": 2.5816730909218815,
+      "learning_rate": 1.011027691051283e-05,
+      "loss": 0.6624,
+      "step": 6687
+    },
+    {
+      "epoch": 1.0234123947972456,
+      "grad_norm": 2.2465134024423703,
+      "learning_rate": 1.010779887479152e-05,
+      "loss": 0.6497,
+      "step": 6688
+    },
+    {
+      "epoch": 1.0235654169854629,
+      "grad_norm": 2.193620461985959,
+      "learning_rate": 1.0105320832449856e-05,
+      "loss": 0.5496,
+      "step": 6689
+    },
+    {
+      "epoch": 1.0237184391736802,
+      "grad_norm": 2.3333263624843212,
+      "learning_rate": 1.0102842783640032e-05,
+      "loss": 0.6004,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0238714613618976,
+      "grad_norm": 2.3841598412894442,
+      "learning_rate": 1.0100364728514228e-05,
+      "loss": 0.6371,
+      "step": 6691
+    },
+    {
+      "epoch": 1.0240244835501147,
+      "grad_norm": 2.223066765452285,
+      "learning_rate": 1.0097886667224634e-05,
+      "loss": 0.5576,
+      "step": 6692
+    },
+    {
+      "epoch": 1.024177505738332,
+      "grad_norm": 2.5087968900231012,
+      "learning_rate": 1.0095408599923438e-05,
+      "loss": 0.5812,
+      "step": 6693
+    },
+    {
+      "epoch": 1.0243305279265493,
+      "grad_norm": 2.57215618551181,
+      "learning_rate": 1.0092930526762824e-05,
+      "loss": 0.6023,
+      "step": 6694
+    },
+    {
+      "epoch": 1.0244835501147667,
+      "grad_norm": 2.402925023331007,
+      "learning_rate": 1.0090452447894985e-05,
+      "loss": 0.5092,
+      "step": 6695
+    },
+    {
+      "epoch": 1.024636572302984,
+      "grad_norm": 2.2994948689718275,
+      "learning_rate": 1.0087974363472107e-05,
+      "loss": 0.6011,
+      "step": 6696
+    },
+    {
+      "epoch": 1.0247895944912013,
+      "grad_norm": 2.220827506450029,
+      "learning_rate": 1.0085496273646377e-05,
+      "loss": 0.5272,
+      "step": 6697
+    },
+    {
+      "epoch": 1.0249426166794184,
+      "grad_norm": 2.337313218700674,
+      "learning_rate": 1.0083018178569987e-05,
+      "loss": 0.5645,
+      "step": 6698
+    },
+    {
+      "epoch": 1.0250956388676358,
+      "grad_norm": 2.3675969691011254,
+      "learning_rate": 1.008054007839513e-05,
+      "loss": 0.6337,
+      "step": 6699
+    },
+    {
+      "epoch": 1.025248661055853,
+      "grad_norm": 2.150314678408855,
+      "learning_rate": 1.0078061973273986e-05,
+      "loss": 0.5149,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0254016832440704,
+      "grad_norm": 2.5780589470023942,
+      "learning_rate": 1.0075583863358756e-05,
+      "loss": 0.608,
+      "step": 6701
+    },
+    {
+      "epoch": 1.0255547054322878,
+      "grad_norm": 2.147503852334118,
+      "learning_rate": 1.0073105748801622e-05,
+      "loss": 0.4818,
+      "step": 6702
+    },
+    {
+      "epoch": 1.0257077276205049,
+      "grad_norm": 2.4672879810343504,
+      "learning_rate": 1.0070627629754778e-05,
+      "loss": 0.4973,
+      "step": 6703
+    },
+    {
+      "epoch": 1.0258607498087222,
+      "grad_norm": 2.2872325728792804,
+      "learning_rate": 1.0068149506370418e-05,
+      "loss": 0.4786,
+      "step": 6704
+    },
+    {
+      "epoch": 1.0260137719969395,
+      "grad_norm": 2.8109264658399375,
+      "learning_rate": 1.0065671378800725e-05,
+      "loss": 0.5947,
+      "step": 6705
+    },
+    {
+      "epoch": 1.0261667941851569,
+      "grad_norm": 2.12874627307908,
+      "learning_rate": 1.00631932471979e-05,
+      "loss": 0.475,
+      "step": 6706
+    },
+    {
+      "epoch": 1.0263198163733742,
+      "grad_norm": 2.3004995092458516,
+      "learning_rate": 1.006071511171413e-05,
+      "loss": 0.577,
+      "step": 6707
+    },
+    {
+      "epoch": 1.0264728385615913,
+      "grad_norm": 2.530976996073714,
+      "learning_rate": 1.0058236972501607e-05,
+      "loss": 0.7009,
+      "step": 6708
+    },
+    {
+      "epoch": 1.0266258607498087,
+      "grad_norm": 2.3849208819910586,
+      "learning_rate": 1.0055758829712522e-05,
+      "loss": 0.5976,
+      "step": 6709
+    },
+    {
+      "epoch": 1.026778882938026,
+      "grad_norm": 2.141201510237259,
+      "learning_rate": 1.0053280683499069e-05,
+      "loss": 0.5255,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0269319051262433,
+      "grad_norm": 2.3014505047534985,
+      "learning_rate": 1.0050802534013444e-05,
+      "loss": 0.5084,
+      "step": 6711
+    },
+    {
+      "epoch": 1.0270849273144607,
+      "grad_norm": 2.6769632245570394,
+      "learning_rate": 1.0048324381407837e-05,
+      "loss": 0.6121,
+      "step": 6712
+    },
+    {
+      "epoch": 1.027237949502678,
+      "grad_norm": 2.2107203457710756,
+      "learning_rate": 1.0045846225834434e-05,
+      "loss": 0.5219,
+      "step": 6713
+    },
+    {
+      "epoch": 1.027390971690895,
+      "grad_norm": 2.2878697948454403,
+      "learning_rate": 1.004336806744544e-05,
+      "loss": 0.5553,
+      "step": 6714
+    },
+    {
+      "epoch": 1.0275439938791124,
+      "grad_norm": 2.1499839688324327,
+      "learning_rate": 1.0040889906393044e-05,
+      "loss": 0.5307,
+      "step": 6715
+    },
+    {
+      "epoch": 1.0276970160673298,
+      "grad_norm": 2.3165989907378037,
+      "learning_rate": 1.0038411742829437e-05,
+      "loss": 0.5045,
+      "step": 6716
+    },
+    {
+      "epoch": 1.027850038255547,
+      "grad_norm": 2.1737988536205797,
+      "learning_rate": 1.0035933576906815e-05,
+      "loss": 0.5147,
+      "step": 6717
+    },
+    {
+      "epoch": 1.0280030604437644,
+      "grad_norm": 2.489338280509698,
+      "learning_rate": 1.0033455408777374e-05,
+      "loss": 0.5733,
+      "step": 6718
+    },
+    {
+      "epoch": 1.0281560826319815,
+      "grad_norm": 2.3878592780046497,
+      "learning_rate": 1.0030977238593303e-05,
+      "loss": 0.5415,
+      "step": 6719
+    },
+    {
+      "epoch": 1.0283091048201989,
+      "grad_norm": 2.4479340330867587,
+      "learning_rate": 1.0028499066506799e-05,
+      "loss": 0.6103,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0284621270084162,
+      "grad_norm": 2.3518568222133633,
+      "learning_rate": 1.0026020892670056e-05,
+      "loss": 0.5775,
+      "step": 6721
+    },
+    {
+      "epoch": 1.0286151491966335,
+      "grad_norm": 2.938195775116116,
+      "learning_rate": 1.0023542717235268e-05,
+      "loss": 0.5382,
+      "step": 6722
+    },
+    {
+      "epoch": 1.0287681713848509,
+      "grad_norm": 2.3623126529241554,
+      "learning_rate": 1.0021064540354627e-05,
+      "loss": 0.5309,
+      "step": 6723
+    },
+    {
+      "epoch": 1.028921193573068,
+      "grad_norm": 2.4496299533549166,
+      "learning_rate": 1.0018586362180335e-05,
+      "loss": 0.6377,
+      "step": 6724
+    },
+    {
+      "epoch": 1.0290742157612853,
+      "grad_norm": 2.573262792528309,
+      "learning_rate": 1.0016108182864578e-05,
+      "loss": 0.5198,
+      "step": 6725
+    },
+    {
+      "epoch": 1.0292272379495027,
+      "grad_norm": 2.29374203363862,
+      "learning_rate": 1.0013630002559556e-05,
+      "loss": 0.5467,
+      "step": 6726
+    },
+    {
+      "epoch": 1.02938026013772,
+      "grad_norm": 2.544986422031994,
+      "learning_rate": 1.0011151821417462e-05,
+      "loss": 0.6458,
+      "step": 6727
+    },
+    {
+      "epoch": 1.0295332823259373,
+      "grad_norm": 2.355349246426756,
+      "learning_rate": 1.0008673639590493e-05,
+      "loss": 0.533,
+      "step": 6728
+    },
+    {
+      "epoch": 1.0296863045141547,
+      "grad_norm": 2.4051355343169605,
+      "learning_rate": 1.0006195457230838e-05,
+      "loss": 0.5813,
+      "step": 6729
+    },
+    {
+      "epoch": 1.0298393267023718,
+      "grad_norm": 2.2601662708897052,
+      "learning_rate": 1.0003717274490698e-05,
+      "loss": 0.503,
+      "step": 6730
+    },
+    {
+      "epoch": 1.029992348890589,
+      "grad_norm": 2.6026331972423145,
+      "learning_rate": 1.0001239091522266e-05,
+      "loss": 0.6445,
+      "step": 6731
+    },
+    {
+      "epoch": 1.0301453710788064,
+      "grad_norm": 2.4502665700638664,
+      "learning_rate": 9.998760908477734e-06,
+      "loss": 0.6071,
+      "step": 6732
+    },
+    {
+      "epoch": 1.0302983932670238,
+      "grad_norm": 2.313243201112418,
+      "learning_rate": 9.996282725509305e-06,
+      "loss": 0.6863,
+      "step": 6733
+    },
+    {
+      "epoch": 1.030451415455241,
+      "grad_norm": 2.1587677649613357,
+      "learning_rate": 9.993804542769167e-06,
+      "loss": 0.5032,
+      "step": 6734
+    },
+    {
+      "epoch": 1.0306044376434582,
+      "grad_norm": 2.4333278255649176,
+      "learning_rate": 9.991326360409509e-06,
+      "loss": 0.5529,
+      "step": 6735
+    },
+    {
+      "epoch": 1.0307574598316755,
+      "grad_norm": 2.4443357441123292,
+      "learning_rate": 9.988848178582541e-06,
+      "loss": 0.5748,
+      "step": 6736
+    },
+    {
+      "epoch": 1.0309104820198929,
+      "grad_norm": 2.190378352896407,
+      "learning_rate": 9.986369997440445e-06,
+      "loss": 0.5912,
+      "step": 6737
+    },
+    {
+      "epoch": 1.0310635042081102,
+      "grad_norm": 2.2856778913735143,
+      "learning_rate": 9.983891817135423e-06,
+      "loss": 0.6169,
+      "step": 6738
+    },
+    {
+      "epoch": 1.0312165263963275,
+      "grad_norm": 2.3383660191650604,
+      "learning_rate": 9.98141363781967e-06,
+      "loss": 0.5175,
+      "step": 6739
+    },
+    {
+      "epoch": 1.0313695485845447,
+      "grad_norm": 2.4669727398099264,
+      "learning_rate": 9.978935459645374e-06,
+      "loss": 0.5764,
+      "step": 6740
+    },
+    {
+      "epoch": 1.031522570772762,
+      "grad_norm": 2.454183738970548,
+      "learning_rate": 9.976457282764735e-06,
+      "loss": 0.5029,
+      "step": 6741
+    },
+    {
+      "epoch": 1.0316755929609793,
+      "grad_norm": 2.146600754225679,
+      "learning_rate": 9.97397910732995e-06,
+      "loss": 0.4988,
+      "step": 6742
+    },
+    {
+      "epoch": 1.0318286151491967,
+      "grad_norm": 2.544317459747779,
+      "learning_rate": 9.971500933493203e-06,
+      "loss": 0.5676,
+      "step": 6743
+    },
+    {
+      "epoch": 1.031981637337414,
+      "grad_norm": 2.129239502795901,
+      "learning_rate": 9.9690227614067e-06,
+      "loss": 0.5151,
+      "step": 6744
+    },
+    {
+      "epoch": 1.032134659525631,
+      "grad_norm": 2.5537722341810354,
+      "learning_rate": 9.966544591222626e-06,
+      "loss": 0.515,
+      "step": 6745
+    },
+    {
+      "epoch": 1.0322876817138484,
+      "grad_norm": 2.4961159403188464,
+      "learning_rate": 9.964066423093186e-06,
+      "loss": 0.6098,
+      "step": 6746
+    },
+    {
+      "epoch": 1.0324407039020658,
+      "grad_norm": 2.26264700212384,
+      "learning_rate": 9.961588257170565e-06,
+      "loss": 0.4847,
+      "step": 6747
+    },
+    {
+      "epoch": 1.032593726090283,
+      "grad_norm": 2.3122842418111667,
+      "learning_rate": 9.959110093606956e-06,
+      "loss": 0.5332,
+      "step": 6748
+    },
+    {
+      "epoch": 1.0327467482785004,
+      "grad_norm": 2.3473237403635676,
+      "learning_rate": 9.95663193255456e-06,
+      "loss": 0.4923,
+      "step": 6749
+    },
+    {
+      "epoch": 1.0328997704667178,
+      "grad_norm": 2.493445708562281,
+      "learning_rate": 9.954153774165564e-06,
+      "loss": 0.6176,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0330527926549349,
+      "grad_norm": 2.0928945035756823,
+      "learning_rate": 9.951675618592168e-06,
+      "loss": 0.4642,
+      "step": 6751
+    },
+    {
+      "epoch": 1.0332058148431522,
+      "grad_norm": 2.3615436131724716,
+      "learning_rate": 9.94919746598656e-06,
+      "loss": 0.4763,
+      "step": 6752
+    },
+    {
+      "epoch": 1.0333588370313695,
+      "grad_norm": 2.4754545731602575,
+      "learning_rate": 9.946719316500931e-06,
+      "loss": 0.5336,
+      "step": 6753
+    },
+    {
+      "epoch": 1.0335118592195869,
+      "grad_norm": 2.442943084544504,
+      "learning_rate": 9.94424117028748e-06,
+      "loss": 0.5174,
+      "step": 6754
+    },
+    {
+      "epoch": 1.0336648814078042,
+      "grad_norm": 2.682564815574864,
+      "learning_rate": 9.941763027498398e-06,
+      "loss": 0.6083,
+      "step": 6755
+    },
+    {
+      "epoch": 1.0338179035960213,
+      "grad_norm": 2.6595816044960965,
+      "learning_rate": 9.939284888285872e-06,
+      "loss": 0.5766,
+      "step": 6756
+    },
+    {
+      "epoch": 1.0339709257842387,
+      "grad_norm": 2.5083499459558656,
+      "learning_rate": 9.936806752802103e-06,
+      "loss": 0.5509,
+      "step": 6757
+    },
+    {
+      "epoch": 1.034123947972456,
+      "grad_norm": 2.275555021513342,
+      "learning_rate": 9.934328621199273e-06,
+      "loss": 0.5047,
+      "step": 6758
+    },
+    {
+      "epoch": 1.0342769701606733,
+      "grad_norm": 2.028981076722719,
+      "learning_rate": 9.931850493629587e-06,
+      "loss": 0.4789,
+      "step": 6759
+    },
+    {
+      "epoch": 1.0344299923488907,
+      "grad_norm": 2.0211581388756037,
+      "learning_rate": 9.929372370245225e-06,
+      "loss": 0.5024,
+      "step": 6760
+    },
+    {
+      "epoch": 1.034583014537108,
+      "grad_norm": 2.324035265435848,
+      "learning_rate": 9.92689425119838e-06,
+      "loss": 0.5728,
+      "step": 6761
+    },
+    {
+      "epoch": 1.034736036725325,
+      "grad_norm": 2.242721972544215,
+      "learning_rate": 9.924416136641249e-06,
+      "loss": 0.4953,
+      "step": 6762
+    },
+    {
+      "epoch": 1.0348890589135424,
+      "grad_norm": 2.432779427177171,
+      "learning_rate": 9.921938026726015e-06,
+      "loss": 0.6302,
+      "step": 6763
+    },
+    {
+      "epoch": 1.0350420811017598,
+      "grad_norm": 2.2598617265081153,
+      "learning_rate": 9.919459921604872e-06,
+      "loss": 0.5825,
+      "step": 6764
+    },
+    {
+      "epoch": 1.035195103289977,
+      "grad_norm": 2.5631858944269412,
+      "learning_rate": 9.916981821430016e-06,
+      "loss": 0.6006,
+      "step": 6765
+    },
+    {
+      "epoch": 1.0353481254781944,
+      "grad_norm": 2.50597329793168,
+      "learning_rate": 9.914503726353623e-06,
+      "loss": 0.5106,
+      "step": 6766
+    },
+    {
+      "epoch": 1.0355011476664115,
+      "grad_norm": 2.5551518698686464,
+      "learning_rate": 9.912025636527897e-06,
+      "loss": 0.5799,
+      "step": 6767
+    },
+    {
+      "epoch": 1.0356541698546289,
+      "grad_norm": 2.716633047615582,
+      "learning_rate": 9.90954755210502e-06,
+      "loss": 0.5963,
+      "step": 6768
+    },
+    {
+      "epoch": 1.0358071920428462,
+      "grad_norm": 2.187872786560717,
+      "learning_rate": 9.907069473237178e-06,
+      "loss": 0.5286,
+      "step": 6769
+    },
+    {
+      "epoch": 1.0359602142310635,
+      "grad_norm": 2.27583781573365,
+      "learning_rate": 9.904591400076567e-06,
+      "loss": 0.5259,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0361132364192809,
+      "grad_norm": 2.2572498780404486,
+      "learning_rate": 9.902113332775372e-06,
+      "loss": 0.507,
+      "step": 6771
+    },
+    {
+      "epoch": 1.036266258607498,
+      "grad_norm": 2.270017680287196,
+      "learning_rate": 9.899635271485774e-06,
+      "loss": 0.5607,
+      "step": 6772
+    },
+    {
+      "epoch": 1.0364192807957153,
+      "grad_norm": 2.23521431485531,
+      "learning_rate": 9.897157216359972e-06,
+      "loss": 0.5248,
+      "step": 6773
+    },
+    {
+      "epoch": 1.0365723029839327,
+      "grad_norm": 2.3389129894008693,
+      "learning_rate": 9.894679167550143e-06,
+      "loss": 0.4359,
+      "step": 6774
+    },
+    {
+      "epoch": 1.03672532517215,
+      "grad_norm": 2.2620396345267344,
+      "learning_rate": 9.892201125208484e-06,
+      "loss": 0.5479,
+      "step": 6775
+    },
+    {
+      "epoch": 1.0368783473603673,
+      "grad_norm": 2.728001867195493,
+      "learning_rate": 9.889723089487175e-06,
+      "loss": 0.6709,
+      "step": 6776
+    },
+    {
+      "epoch": 1.0370313695485844,
+      "grad_norm": 2.1827657385278445,
+      "learning_rate": 9.887245060538397e-06,
+      "loss": 0.496,
+      "step": 6777
+    },
+    {
+      "epoch": 1.0371843917368018,
+      "grad_norm": 2.3223435649613973,
+      "learning_rate": 9.884767038514348e-06,
+      "loss": 0.5999,
+      "step": 6778
+    },
+    {
+      "epoch": 1.037337413925019,
+      "grad_norm": 2.2614536615688663,
+      "learning_rate": 9.882289023567203e-06,
+      "loss": 0.5056,
+      "step": 6779
+    },
+    {
+      "epoch": 1.0374904361132364,
+      "grad_norm": 2.73646991422452,
+      "learning_rate": 9.879811015849147e-06,
+      "loss": 0.6103,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0376434583014538,
+      "grad_norm": 2.5056718112340093,
+      "learning_rate": 9.87733301551237e-06,
+      "loss": 0.5605,
+      "step": 6781
+    },
+    {
+      "epoch": 1.037796480489671,
+      "grad_norm": 2.48718682079311,
+      "learning_rate": 9.87485502270905e-06,
+      "loss": 0.5004,
+      "step": 6782
+    },
+    {
+      "epoch": 1.0379495026778882,
+      "grad_norm": 2.581830328744179,
+      "learning_rate": 9.872377037591374e-06,
+      "loss": 0.5789,
+      "step": 6783
+    },
+    {
+      "epoch": 1.0381025248661055,
+      "grad_norm": 2.368386825181525,
+      "learning_rate": 9.869899060311525e-06,
+      "loss": 0.5313,
+      "step": 6784
+    },
+    {
+      "epoch": 1.0382555470543229,
+      "grad_norm": 2.4010158149306884,
+      "learning_rate": 9.86742109102168e-06,
+      "loss": 0.5985,
+      "step": 6785
+    },
+    {
+      "epoch": 1.0384085692425402,
+      "grad_norm": 2.483515581888754,
+      "learning_rate": 9.864943129874027e-06,
+      "loss": 0.517,
+      "step": 6786
+    },
+    {
+      "epoch": 1.0385615914307575,
+      "grad_norm": 2.694645694193147,
+      "learning_rate": 9.862465177020742e-06,
+      "loss": 0.5725,
+      "step": 6787
+    },
+    {
+      "epoch": 1.0387146136189747,
+      "grad_norm": 2.00193207158381,
+      "learning_rate": 9.859987232614012e-06,
+      "loss": 0.4655,
+      "step": 6788
+    },
+    {
+      "epoch": 1.038867635807192,
+      "grad_norm": 2.297472219311569,
+      "learning_rate": 9.857509296806014e-06,
+      "loss": 0.5012,
+      "step": 6789
+    },
+    {
+      "epoch": 1.0390206579954093,
+      "grad_norm": 2.0400785317490127,
+      "learning_rate": 9.855031369748922e-06,
+      "loss": 0.5425,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0391736801836267,
+      "grad_norm": 2.162087427618444,
+      "learning_rate": 9.852553451594929e-06,
+      "loss": 0.4972,
+      "step": 6791
+    },
+    {
+      "epoch": 1.039326702371844,
+      "grad_norm": 2.297269429343117,
+      "learning_rate": 9.8500755424962e-06,
+      "loss": 0.4935,
+      "step": 6792
+    },
+    {
+      "epoch": 1.039479724560061,
+      "grad_norm": 2.4070600512021785,
+      "learning_rate": 9.847597642604917e-06,
+      "loss": 0.5847,
+      "step": 6793
+    },
+    {
+      "epoch": 1.0396327467482784,
+      "grad_norm": 2.517284455670916,
+      "learning_rate": 9.845119752073265e-06,
+      "loss": 0.5392,
+      "step": 6794
+    },
+    {
+      "epoch": 1.0397857689364958,
+      "grad_norm": 2.3382151130233817,
+      "learning_rate": 9.84264187105341e-06,
+      "loss": 0.5164,
+      "step": 6795
+    },
+    {
+      "epoch": 1.039938791124713,
+      "grad_norm": 2.2612802000483216,
+      "learning_rate": 9.840163999697532e-06,
+      "loss": 0.5151,
+      "step": 6796
+    },
+    {
+      "epoch": 1.0400918133129304,
+      "grad_norm": 2.2601432005493263,
+      "learning_rate": 9.837686138157813e-06,
+      "loss": 0.5168,
+      "step": 6797
+    },
+    {
+      "epoch": 1.0402448355011478,
+      "grad_norm": 2.6100201009422235,
+      "learning_rate": 9.835208286586419e-06,
+      "loss": 0.6259,
+      "step": 6798
+    },
+    {
+      "epoch": 1.0403978576893649,
+      "grad_norm": 2.4883767574053706,
+      "learning_rate": 9.832730445135531e-06,
+      "loss": 0.5042,
+      "step": 6799
+    },
+    {
+      "epoch": 1.0405508798775822,
+      "grad_norm": 2.2820139860370183,
+      "learning_rate": 9.830252613957322e-06,
+      "loss": 0.4631,
+      "step": 6800
+    },
+    {
+      "epoch": 1.0407039020657995,
+      "grad_norm": 2.25483309950987,
+      "learning_rate": 9.827774793203961e-06,
+      "loss": 0.5455,
+      "step": 6801
+    },
+    {
+      "epoch": 1.0408569242540169,
+      "grad_norm": 2.6447960053774993,
+      "learning_rate": 9.825296983027625e-06,
+      "loss": 0.6592,
+      "step": 6802
+    },
+    {
+      "epoch": 1.0410099464422342,
+      "grad_norm": 2.053680781948206,
+      "learning_rate": 9.822819183580484e-06,
+      "loss": 0.5091,
+      "step": 6803
+    },
+    {
+      "epoch": 1.0411629686304513,
+      "grad_norm": 2.395301936062382,
+      "learning_rate": 9.820341395014713e-06,
+      "loss": 0.546,
+      "step": 6804
+    },
+    {
+      "epoch": 1.0413159908186687,
+      "grad_norm": 2.033876449030659,
+      "learning_rate": 9.817863617482479e-06,
+      "loss": 0.4092,
+      "step": 6805
+    },
+    {
+      "epoch": 1.041469013006886,
+      "grad_norm": 2.413056030738432,
+      "learning_rate": 9.815385851135948e-06,
+      "loss": 0.652,
+      "step": 6806
+    },
+    {
+      "epoch": 1.0416220351951033,
+      "grad_norm": 2.448201678783148,
+      "learning_rate": 9.8129080961273e-06,
+      "loss": 0.4885,
+      "step": 6807
+    },
+    {
+      "epoch": 1.0417750573833207,
+      "grad_norm": 2.518063374444106,
+      "learning_rate": 9.810430352608695e-06,
+      "loss": 0.5058,
+      "step": 6808
+    },
+    {
+      "epoch": 1.0419280795715378,
+      "grad_norm": 2.000437078764586,
+      "learning_rate": 9.807952620732302e-06,
+      "loss": 0.4609,
+      "step": 6809
+    },
+    {
+      "epoch": 1.042081101759755,
+      "grad_norm": 2.602756111222112,
+      "learning_rate": 9.805474900650296e-06,
+      "loss": 0.6179,
+      "step": 6810
+    },
+    {
+      "epoch": 1.0422341239479724,
+      "grad_norm": 2.4146641274588654,
+      "learning_rate": 9.802997192514827e-06,
+      "loss": 0.5391,
+      "step": 6811
+    },
+    {
+      "epoch": 1.0423871461361898,
+      "grad_norm": 2.4073284310419356,
+      "learning_rate": 9.80051949647808e-06,
+      "loss": 0.5885,
+      "step": 6812
+    },
+    {
+      "epoch": 1.042540168324407,
+      "grad_norm": 2.3612873008392166,
+      "learning_rate": 9.798041812692211e-06,
+      "loss": 0.5162,
+      "step": 6813
+    },
+    {
+      "epoch": 1.0426931905126244,
+      "grad_norm": 2.5148254857651042,
+      "learning_rate": 9.795564141309376e-06,
+      "loss": 0.5554,
+      "step": 6814
+    },
+    {
+      "epoch": 1.0428462127008415,
+      "grad_norm": 2.193824725601136,
+      "learning_rate": 9.793086482481755e-06,
+      "loss": 0.4637,
+      "step": 6815
+    },
+    {
+      "epoch": 1.0429992348890589,
+      "grad_norm": 2.195282093201084,
+      "learning_rate": 9.790608836361502e-06,
+      "loss": 0.5601,
+      "step": 6816
+    },
+    {
+      "epoch": 1.0431522570772762,
+      "grad_norm": 2.3815272787115522,
+      "learning_rate": 9.788131203100774e-06,
+      "loss": 0.5558,
+      "step": 6817
+    },
+    {
+      "epoch": 1.0433052792654935,
+      "grad_norm": 2.579512220117735,
+      "learning_rate": 9.785653582851745e-06,
+      "loss": 0.5961,
+      "step": 6818
+    },
+    {
+      "epoch": 1.0434583014537109,
+      "grad_norm": 2.3883347162224244,
+      "learning_rate": 9.783175975766561e-06,
+      "loss": 0.6108,
+      "step": 6819
+    },
+    {
+      "epoch": 1.043611323641928,
+      "grad_norm": 2.3174436373500686,
+      "learning_rate": 9.78069838199739e-06,
+      "loss": 0.5733,
+      "step": 6820
+    },
+    {
+      "epoch": 1.0437643458301453,
+      "grad_norm": 2.2060498412069744,
+      "learning_rate": 9.778220801696395e-06,
+      "loss": 0.5214,
+      "step": 6821
+    },
+    {
+      "epoch": 1.0439173680183627,
+      "grad_norm": 2.2232999074708797,
+      "learning_rate": 9.775743235015721e-06,
+      "loss": 0.5198,
+      "step": 6822
+    },
+    {
+      "epoch": 1.04407039020658,
+      "grad_norm": 2.3554090047539535,
+      "learning_rate": 9.773265682107538e-06,
+      "loss": 0.5541,
+      "step": 6823
+    },
+    {
+      "epoch": 1.0442234123947973,
+      "grad_norm": 2.16785446117587,
+      "learning_rate": 9.77078814312399e-06,
+      "loss": 0.4758,
+      "step": 6824
+    },
+    {
+      "epoch": 1.0443764345830144,
+      "grad_norm": 2.454126851389489,
+      "learning_rate": 9.768310618217246e-06,
+      "loss": 0.5185,
+      "step": 6825
+    },
+    {
+      "epoch": 1.0445294567712318,
+      "grad_norm": 2.6924414731654864,
+      "learning_rate": 9.76583310753945e-06,
+      "loss": 0.639,
+      "step": 6826
+    },
+    {
+      "epoch": 1.044682478959449,
+      "grad_norm": 2.265111260087062,
+      "learning_rate": 9.763355611242757e-06,
+      "loss": 0.4323,
+      "step": 6827
+    },
+    {
+      "epoch": 1.0448355011476664,
+      "grad_norm": 2.277809428494371,
+      "learning_rate": 9.760878129479325e-06,
+      "loss": 0.4809,
+      "step": 6828
+    },
+    {
+      "epoch": 1.0449885233358838,
+      "grad_norm": 2.2456569046235453,
+      "learning_rate": 9.758400662401301e-06,
+      "loss": 0.4522,
+      "step": 6829
+    },
+    {
+      "epoch": 1.045141545524101,
+      "grad_norm": 2.453577541760022,
+      "learning_rate": 9.755923210160836e-06,
+      "loss": 0.4833,
+      "step": 6830
+    },
+    {
+      "epoch": 1.0452945677123182,
+      "grad_norm": 2.482612822718974,
+      "learning_rate": 9.753445772910085e-06,
+      "loss": 0.5288,
+      "step": 6831
+    },
+    {
+      "epoch": 1.0454475899005355,
+      "grad_norm": 2.2688767121063216,
+      "learning_rate": 9.75096835080119e-06,
+      "loss": 0.5265,
+      "step": 6832
+    },
+    {
+      "epoch": 1.0456006120887529,
+      "grad_norm": 2.3742525332643485,
+      "learning_rate": 9.748490943986304e-06,
+      "loss": 0.5401,
+      "step": 6833
+    },
+    {
+      "epoch": 1.0457536342769702,
+      "grad_norm": 2.297535035475942,
+      "learning_rate": 9.746013552617576e-06,
+      "loss": 0.5176,
+      "step": 6834
+    },
+    {
+      "epoch": 1.0459066564651875,
+      "grad_norm": 2.2986799616130615,
+      "learning_rate": 9.743536176847145e-06,
+      "loss": 0.6058,
+      "step": 6835
+    },
+    {
+      "epoch": 1.0460596786534047,
+      "grad_norm": 2.6040700168531097,
+      "learning_rate": 9.741058816827162e-06,
+      "loss": 0.4995,
+      "step": 6836
+    },
+    {
+      "epoch": 1.046212700841622,
+      "grad_norm": 2.305543948337454,
+      "learning_rate": 9.738581472709774e-06,
+      "loss": 0.5135,
+      "step": 6837
+    },
+    {
+      "epoch": 1.0463657230298393,
+      "grad_norm": 2.4772310516819034,
+      "learning_rate": 9.736104144647114e-06,
+      "loss": 0.5416,
+      "step": 6838
+    },
+    {
+      "epoch": 1.0465187452180567,
+      "grad_norm": 2.292649813878011,
+      "learning_rate": 9.733626832791336e-06,
+      "loss": 0.4853,
+      "step": 6839
+    },
+    {
+      "epoch": 1.046671767406274,
+      "grad_norm": 2.7654251480866803,
+      "learning_rate": 9.73114953729457e-06,
+      "loss": 0.4503,
+      "step": 6840
+    },
+    {
+      "epoch": 1.046824789594491,
+      "grad_norm": 2.4129007595804945,
+      "learning_rate": 9.72867225830897e-06,
+      "loss": 0.547,
+      "step": 6841
+    },
+    {
+      "epoch": 1.0469778117827084,
+      "grad_norm": 2.565444877722573,
+      "learning_rate": 9.726194995986665e-06,
+      "loss": 0.5448,
+      "step": 6842
+    },
+    {
+      "epoch": 1.0471308339709258,
+      "grad_norm": 2.3819084893973965,
+      "learning_rate": 9.723717750479793e-06,
+      "loss": 0.4662,
+      "step": 6843
+    },
+    {
+      "epoch": 1.047283856159143,
+      "grad_norm": 1.9489195524522251,
+      "learning_rate": 9.721240521940501e-06,
+      "loss": 0.5165,
+      "step": 6844
+    },
+    {
+      "epoch": 1.0474368783473604,
+      "grad_norm": 2.5377098649664727,
+      "learning_rate": 9.718763310520916e-06,
+      "loss": 0.6242,
+      "step": 6845
+    },
+    {
+      "epoch": 1.0475899005355775,
+      "grad_norm": 2.280532226310244,
+      "learning_rate": 9.716286116373174e-06,
+      "loss": 0.6094,
+      "step": 6846
+    },
+    {
+      "epoch": 1.0477429227237949,
+      "grad_norm": 2.3174754567622773,
+      "learning_rate": 9.713808939649417e-06,
+      "loss": 0.4831,
+      "step": 6847
+    },
+    {
+      "epoch": 1.0478959449120122,
+      "grad_norm": 2.5142215573367515,
+      "learning_rate": 9.711331780501766e-06,
+      "loss": 0.5593,
+      "step": 6848
+    },
+    {
+      "epoch": 1.0480489671002295,
+      "grad_norm": 2.1926272912275264,
+      "learning_rate": 9.708854639082364e-06,
+      "loss": 0.4583,
+      "step": 6849
+    },
+    {
+      "epoch": 1.0482019892884469,
+      "grad_norm": 2.226115977484892,
+      "learning_rate": 9.70637751554334e-06,
+      "loss": 0.537,
+      "step": 6850
+    },
+    {
+      "epoch": 1.0483550114766642,
+      "grad_norm": 2.115691597008675,
+      "learning_rate": 9.703900410036815e-06,
+      "loss": 0.5448,
+      "step": 6851
+    },
+    {
+      "epoch": 1.0485080336648813,
+      "grad_norm": 2.180567669515745,
+      "learning_rate": 9.701423322714928e-06,
+      "loss": 0.5617,
+      "step": 6852
+    },
+    {
+      "epoch": 1.0486610558530987,
+      "grad_norm": 2.382574902525598,
+      "learning_rate": 9.698946253729804e-06,
+      "loss": 0.5341,
+      "step": 6853
+    },
+    {
+      "epoch": 1.048814078041316,
+      "grad_norm": 2.239884675264611,
+      "learning_rate": 9.696469203233565e-06,
+      "loss": 0.534,
+      "step": 6854
+    },
+    {
+      "epoch": 1.0489671002295333,
+      "grad_norm": 2.2210027907755334,
+      "learning_rate": 9.693992171378342e-06,
+      "loss": 0.4105,
+      "step": 6855
+    },
+    {
+      "epoch": 1.0491201224177507,
+      "grad_norm": 2.1536780424328894,
+      "learning_rate": 9.691515158316253e-06,
+      "loss": 0.4926,
+      "step": 6856
+    },
+    {
+      "epoch": 1.0492731446059678,
+      "grad_norm": 2.533943071647297,
+      "learning_rate": 9.68903816419943e-06,
+      "loss": 0.6514,
+      "step": 6857
+    },
+    {
+      "epoch": 1.049426166794185,
+      "grad_norm": 2.292108221676427,
+      "learning_rate": 9.686561189179989e-06,
+      "loss": 0.4834,
+      "step": 6858
+    },
+    {
+      "epoch": 1.0495791889824024,
+      "grad_norm": 2.1828211174386403,
+      "learning_rate": 9.684084233410048e-06,
+      "loss": 0.4351,
+      "step": 6859
+    },
+    {
+      "epoch": 1.0497322111706198,
+      "grad_norm": 2.3255510612548007,
+      "learning_rate": 9.681607297041734e-06,
+      "loss": 0.4607,
+      "step": 6860
+    },
+    {
+      "epoch": 1.049885233358837,
+      "grad_norm": 2.1176499482167475,
+      "learning_rate": 9.679130380227158e-06,
+      "loss": 0.4311,
+      "step": 6861
+    },
+    {
+      "epoch": 1.0500382555470544,
+      "grad_norm": 2.317475873274802,
+      "learning_rate": 9.676653483118441e-06,
+      "loss": 0.4604,
+      "step": 6862
+    },
+    {
+      "epoch": 1.0501912777352715,
+      "grad_norm": 2.855097863082061,
+      "learning_rate": 9.674176605867702e-06,
+      "loss": 0.5651,
+      "step": 6863
+    },
+    {
+      "epoch": 1.0503442999234889,
+      "grad_norm": 2.2833665485609074,
+      "learning_rate": 9.67169974862705e-06,
+      "loss": 0.4677,
+      "step": 6864
+    },
+    {
+      "epoch": 1.0504973221117062,
+      "grad_norm": 2.606929688063134,
+      "learning_rate": 9.6692229115486e-06,
+      "loss": 0.597,
+      "step": 6865
+    },
+    {
+      "epoch": 1.0506503442999235,
+      "grad_norm": 2.390847436864242,
+      "learning_rate": 9.666746094784468e-06,
+      "loss": 0.5273,
+      "step": 6866
+    },
+    {
+      "epoch": 1.0508033664881409,
+      "grad_norm": 2.5605618344036545,
+      "learning_rate": 9.664269298486759e-06,
+      "loss": 0.572,
+      "step": 6867
+    },
+    {
+      "epoch": 1.050956388676358,
+      "grad_norm": 2.254858739297128,
+      "learning_rate": 9.661792522807586e-06,
+      "loss": 0.4767,
+      "step": 6868
+    },
+    {
+      "epoch": 1.0511094108645753,
+      "grad_norm": 2.4423773067977423,
+      "learning_rate": 9.659315767899055e-06,
+      "loss": 0.5312,
+      "step": 6869
+    },
+    {
+      "epoch": 1.0512624330527927,
+      "grad_norm": 2.5114329776973423,
+      "learning_rate": 9.656839033913282e-06,
+      "loss": 0.5757,
+      "step": 6870
+    },
+    {
+      "epoch": 1.05141545524101,
+      "grad_norm": 2.17181345130201,
+      "learning_rate": 9.654362321002363e-06,
+      "loss": 0.4097,
+      "step": 6871
+    },
+    {
+      "epoch": 1.0515684774292273,
+      "grad_norm": 2.0419482932100363,
+      "learning_rate": 9.651885629318402e-06,
+      "loss": 0.4301,
+      "step": 6872
+    },
+    {
+      "epoch": 1.0517214996174444,
+      "grad_norm": 2.266374883329687,
+      "learning_rate": 9.649408959013513e-06,
+      "loss": 0.4822,
+      "step": 6873
+    },
+    {
+      "epoch": 1.0518745218056618,
+      "grad_norm": 2.205778807974861,
+      "learning_rate": 9.64693231023979e-06,
+      "loss": 0.5209,
+      "step": 6874
+    },
+    {
+      "epoch": 1.052027543993879,
+      "grad_norm": 2.260631995522916,
+      "learning_rate": 9.64445568314933e-06,
+      "loss": 0.44,
+      "step": 6875
+    },
+    {
+      "epoch": 1.0521805661820964,
+      "grad_norm": 2.3307479914634466,
+      "learning_rate": 9.641979077894244e-06,
+      "loss": 0.4535,
+      "step": 6876
+    },
+    {
+      "epoch": 1.0523335883703138,
+      "grad_norm": 2.2111017557168826,
+      "learning_rate": 9.639502494626618e-06,
+      "loss": 0.5096,
+      "step": 6877
+    },
+    {
+      "epoch": 1.0524866105585309,
+      "grad_norm": 2.366404180462465,
+      "learning_rate": 9.637025933498556e-06,
+      "loss": 0.5082,
+      "step": 6878
+    },
+    {
+      "epoch": 1.0526396327467482,
+      "grad_norm": 2.51983682850274,
+      "learning_rate": 9.634549394662154e-06,
+      "loss": 0.5402,
+      "step": 6879
+    },
+    {
+      "epoch": 1.0527926549349655,
+      "grad_norm": 2.605006859877165,
+      "learning_rate": 9.6320728782695e-06,
+      "loss": 0.5005,
+      "step": 6880
+    },
+    {
+      "epoch": 1.0529456771231829,
+      "grad_norm": 1.9584831153342839,
+      "learning_rate": 9.62959638447269e-06,
+      "loss": 0.4544,
+      "step": 6881
+    },
+    {
+      "epoch": 1.0530986993114002,
+      "grad_norm": 2.503851041495755,
+      "learning_rate": 9.62711991342382e-06,
+      "loss": 0.4845,
+      "step": 6882
+    },
+    {
+      "epoch": 1.0532517214996175,
+      "grad_norm": 2.197647946951353,
+      "learning_rate": 9.624643465274968e-06,
+      "loss": 0.5008,
+      "step": 6883
+    },
+    {
+      "epoch": 1.0534047436878347,
+      "grad_norm": 2.6562728932565185,
+      "learning_rate": 9.622167040178233e-06,
+      "loss": 0.5155,
+      "step": 6884
+    },
+    {
+      "epoch": 1.053557765876052,
+      "grad_norm": 2.2358372881033945,
+      "learning_rate": 9.619690638285694e-06,
+      "loss": 0.526,
+      "step": 6885
+    },
+    {
+      "epoch": 1.0537107880642693,
+      "grad_norm": 2.278041438282626,
+      "learning_rate": 9.617214259749445e-06,
+      "loss": 0.4483,
+      "step": 6886
+    },
+    {
+      "epoch": 1.0538638102524867,
+      "grad_norm": 3.765678149655499,
+      "learning_rate": 9.614737904721567e-06,
+      "loss": 0.4127,
+      "step": 6887
+    },
+    {
+      "epoch": 1.054016832440704,
+      "grad_norm": 2.345689494157314,
+      "learning_rate": 9.612261573354137e-06,
+      "loss": 0.4669,
+      "step": 6888
+    },
+    {
+      "epoch": 1.054169854628921,
+      "grad_norm": 2.215666024931825,
+      "learning_rate": 9.609785265799243e-06,
+      "loss": 0.4277,
+      "step": 6889
+    },
+    {
+      "epoch": 1.0543228768171384,
+      "grad_norm": 2.1371736190516355,
+      "learning_rate": 9.607308982208959e-06,
+      "loss": 0.4995,
+      "step": 6890
+    },
+    {
+      "epoch": 1.0544758990053558,
+      "grad_norm": 2.5607618798633167,
+      "learning_rate": 9.604832722735373e-06,
+      "loss": 0.4405,
+      "step": 6891
+    },
+    {
+      "epoch": 1.054628921193573,
+      "grad_norm": 2.2072868289383214,
+      "learning_rate": 9.602356487530553e-06,
+      "loss": 0.4679,
+      "step": 6892
+    },
+    {
+      "epoch": 1.0547819433817904,
+      "grad_norm": 2.417589786036523,
+      "learning_rate": 9.599880276746572e-06,
+      "loss": 0.4911,
+      "step": 6893
+    },
+    {
+      "epoch": 1.0549349655700078,
+      "grad_norm": 2.0872106304582982,
+      "learning_rate": 9.597404090535515e-06,
+      "loss": 0.4545,
+      "step": 6894
+    },
+    {
+      "epoch": 1.0550879877582249,
+      "grad_norm": 2.46713381700791,
+      "learning_rate": 9.594927929049447e-06,
+      "loss": 0.4741,
+      "step": 6895
+    },
+    {
+      "epoch": 1.0552410099464422,
+      "grad_norm": 2.306058536221254,
+      "learning_rate": 9.592451792440433e-06,
+      "loss": 0.4439,
+      "step": 6896
+    },
+    {
+      "epoch": 1.0553940321346595,
+      "grad_norm": 3.022745301253467,
+      "learning_rate": 9.589975680860556e-06,
+      "loss": 0.5254,
+      "step": 6897
+    },
+    {
+      "epoch": 1.0555470543228769,
+      "grad_norm": 2.345844867262732,
+      "learning_rate": 9.587499594461871e-06,
+      "loss": 0.3766,
+      "step": 6898
+    },
+    {
+      "epoch": 1.0557000765110942,
+      "grad_norm": 2.223361710672595,
+      "learning_rate": 9.585023533396452e-06,
+      "loss": 0.4829,
+      "step": 6899
+    },
+    {
+      "epoch": 1.0558530986993113,
+      "grad_norm": 2.328724388731237,
+      "learning_rate": 9.582547497816364e-06,
+      "loss": 0.5028,
+      "step": 6900
+    },
+    {
+      "epoch": 1.0560061208875287,
+      "grad_norm": 2.2646982308630363,
+      "learning_rate": 9.580071487873663e-06,
+      "loss": 0.4627,
+      "step": 6901
+    },
+    {
+      "epoch": 1.056159143075746,
+      "grad_norm": 2.2351291324961093,
+      "learning_rate": 9.577595503720417e-06,
+      "loss": 0.5294,
+      "step": 6902
+    },
+    {
+      "epoch": 1.0563121652639633,
+      "grad_norm": 2.317554741459491,
+      "learning_rate": 9.575119545508686e-06,
+      "loss": 0.4855,
+      "step": 6903
+    },
+    {
+      "epoch": 1.0564651874521807,
+      "grad_norm": 2.335702932508031,
+      "learning_rate": 9.572643613390521e-06,
+      "loss": 0.4449,
+      "step": 6904
+    },
+    {
+      "epoch": 1.0566182096403978,
+      "grad_norm": 2.3390630669388472,
+      "learning_rate": 9.570167707517986e-06,
+      "loss": 0.497,
+      "step": 6905
+    },
+    {
+      "epoch": 1.056771231828615,
+      "grad_norm": 2.0915049224848734,
+      "learning_rate": 9.567691828043131e-06,
+      "loss": 0.4223,
+      "step": 6906
+    },
+    {
+      "epoch": 1.0569242540168324,
+      "grad_norm": 2.4297434477300177,
+      "learning_rate": 9.565215975118016e-06,
+      "loss": 0.5876,
+      "step": 6907
+    },
+    {
+      "epoch": 1.0570772762050498,
+      "grad_norm": 2.4562615398492014,
+      "learning_rate": 9.562740148894687e-06,
+      "loss": 0.4997,
+      "step": 6908
+    },
+    {
+      "epoch": 1.057230298393267,
+      "grad_norm": 2.5933852297859024,
+      "learning_rate": 9.560264349525193e-06,
+      "loss": 0.5457,
+      "step": 6909
+    },
+    {
+      "epoch": 1.0573833205814842,
+      "grad_norm": 2.107241488019224,
+      "learning_rate": 9.557788577161592e-06,
+      "loss": 0.4718,
+      "step": 6910
+    },
+    {
+      "epoch": 1.0575363427697015,
+      "grad_norm": 2.3210855880243195,
+      "learning_rate": 9.555312831955921e-06,
+      "loss": 0.4619,
+      "step": 6911
+    },
+    {
+      "epoch": 1.0576893649579189,
+      "grad_norm": 2.2978307140280405,
+      "learning_rate": 9.552837114060226e-06,
+      "loss": 0.5951,
+      "step": 6912
+    },
+    {
+      "epoch": 1.0578423871461362,
+      "grad_norm": 2.2115010649647653,
+      "learning_rate": 9.550361423626558e-06,
+      "loss": 0.484,
+      "step": 6913
+    },
+    {
+      "epoch": 1.0579954093343535,
+      "grad_norm": 2.003534556604302,
+      "learning_rate": 9.54788576080695e-06,
+      "loss": 0.3852,
+      "step": 6914
+    },
+    {
+      "epoch": 1.0581484315225709,
+      "grad_norm": 2.429515724313838,
+      "learning_rate": 9.545410125753448e-06,
+      "loss": 0.5315,
+      "step": 6915
+    },
+    {
+      "epoch": 1.058301453710788,
+      "grad_norm": 2.173108303541178,
+      "learning_rate": 9.542934518618092e-06,
+      "loss": 0.4607,
+      "step": 6916
+    },
+    {
+      "epoch": 1.0584544758990053,
+      "grad_norm": 2.232859421941343,
+      "learning_rate": 9.540458939552907e-06,
+      "loss": 0.4832,
+      "step": 6917
+    },
+    {
+      "epoch": 1.0586074980872227,
+      "grad_norm": 2.1284743617910205,
+      "learning_rate": 9.537983388709943e-06,
+      "loss": 0.4607,
+      "step": 6918
+    },
+    {
+      "epoch": 1.05876052027544,
+      "grad_norm": 2.1196258230083673,
+      "learning_rate": 9.535507866241226e-06,
+      "loss": 0.4231,
+      "step": 6919
+    },
+    {
+      "epoch": 1.0589135424636573,
+      "grad_norm": 2.532512044150864,
+      "learning_rate": 9.533032372298784e-06,
+      "loss": 0.5098,
+      "step": 6920
+    },
+    {
+      "epoch": 1.0590665646518744,
+      "grad_norm": 2.2109378116606604,
+      "learning_rate": 9.530556907034653e-06,
+      "loss": 0.4378,
+      "step": 6921
+    },
+    {
+      "epoch": 1.0592195868400918,
+      "grad_norm": 2.3043607556194154,
+      "learning_rate": 9.528081470600857e-06,
+      "loss": 0.4155,
+      "step": 6922
+    },
+    {
+      "epoch": 1.059372609028309,
+      "grad_norm": 2.319007189741688,
+      "learning_rate": 9.525606063149429e-06,
+      "loss": 0.4975,
+      "step": 6923
+    },
+    {
+      "epoch": 1.0595256312165264,
+      "grad_norm": 2.2928485267167447,
+      "learning_rate": 9.523130684832386e-06,
+      "loss": 0.5141,
+      "step": 6924
+    },
+    {
+      "epoch": 1.0596786534047438,
+      "grad_norm": 2.4235326737855254,
+      "learning_rate": 9.52065533580175e-06,
+      "loss": 0.4867,
+      "step": 6925
+    },
+    {
+      "epoch": 1.0598316755929609,
+      "grad_norm": 2.168516218109533,
+      "learning_rate": 9.518180016209551e-06,
+      "loss": 0.4407,
+      "step": 6926
+    },
+    {
+      "epoch": 1.0599846977811782,
+      "grad_norm": 2.069744482004005,
+      "learning_rate": 9.515704726207796e-06,
+      "loss": 0.3991,
+      "step": 6927
+    },
+    {
+      "epoch": 1.0601377199693955,
+      "grad_norm": 2.1506333934930257,
+      "learning_rate": 9.513229465948511e-06,
+      "loss": 0.4787,
+      "step": 6928
+    },
+    {
+      "epoch": 1.0602907421576129,
+      "grad_norm": 2.2749822448358814,
+      "learning_rate": 9.510754235583713e-06,
+      "loss": 0.4889,
+      "step": 6929
+    },
+    {
+      "epoch": 1.0604437643458302,
+      "grad_norm": 2.018532232312824,
+      "learning_rate": 9.508279035265405e-06,
+      "loss": 0.4167,
+      "step": 6930
+    },
+    {
+      "epoch": 1.0605967865340475,
+      "grad_norm": 2.120989225905039,
+      "learning_rate": 9.505803865145606e-06,
+      "loss": 0.4536,
+      "step": 6931
+    },
+    {
+      "epoch": 1.0607498087222647,
+      "grad_norm": 2.1292285782630396,
+      "learning_rate": 9.50332872537633e-06,
+      "loss": 0.4711,
+      "step": 6932
+    },
+    {
+      "epoch": 1.060902830910482,
+      "grad_norm": 2.5222440705575466,
+      "learning_rate": 9.500853616109572e-06,
+      "loss": 0.536,
+      "step": 6933
+    },
+    {
+      "epoch": 1.0610558530986993,
+      "grad_norm": 2.396454824637159,
+      "learning_rate": 9.498378537497352e-06,
+      "loss": 0.5248,
+      "step": 6934
+    },
+    {
+      "epoch": 1.0612088752869167,
+      "grad_norm": 2.42334812785671,
+      "learning_rate": 9.495903489691665e-06,
+      "loss": 0.5531,
+      "step": 6935
+    },
+    {
+      "epoch": 1.061361897475134,
+      "grad_norm": 2.217976890998022,
+      "learning_rate": 9.493428472844521e-06,
+      "loss": 0.4391,
+      "step": 6936
+    },
+    {
+      "epoch": 1.061514919663351,
+      "grad_norm": 2.170116082357399,
+      "learning_rate": 9.490953487107913e-06,
+      "loss": 0.3839,
+      "step": 6937
+    },
+    {
+      "epoch": 1.0616679418515684,
+      "grad_norm": 2.2091807909239263,
+      "learning_rate": 9.488478532633842e-06,
+      "loss": 0.4498,
+      "step": 6938
+    },
+    {
+      "epoch": 1.0618209640397858,
+      "grad_norm": 2.455234760203659,
+      "learning_rate": 9.48600360957431e-06,
+      "loss": 0.5034,
+      "step": 6939
+    },
+    {
+      "epoch": 1.061973986228003,
+      "grad_norm": 2.6446464434159487,
+      "learning_rate": 9.483528718081303e-06,
+      "loss": 0.5489,
+      "step": 6940
+    },
+    {
+      "epoch": 1.0621270084162204,
+      "grad_norm": 2.493535788047184,
+      "learning_rate": 9.481053858306816e-06,
+      "loss": 0.5001,
+      "step": 6941
+    },
+    {
+      "epoch": 1.0622800306044375,
+      "grad_norm": 2.679684989109719,
+      "learning_rate": 9.478579030402849e-06,
+      "loss": 0.602,
+      "step": 6942
+    },
+    {
+      "epoch": 1.0624330527926549,
+      "grad_norm": 2.2237217293867833,
+      "learning_rate": 9.476104234521376e-06,
+      "loss": 0.4686,
+      "step": 6943
+    },
+    {
+      "epoch": 1.0625860749808722,
+      "grad_norm": 2.2552698510205804,
+      "learning_rate": 9.473629470814395e-06,
+      "loss": 0.482,
+      "step": 6944
+    },
+    {
+      "epoch": 1.0627390971690895,
+      "grad_norm": 2.3382411136262915,
+      "learning_rate": 9.471154739433889e-06,
+      "loss": 0.5241,
+      "step": 6945
+    },
+    {
+      "epoch": 1.0628921193573069,
+      "grad_norm": 2.277125644575901,
+      "learning_rate": 9.468680040531834e-06,
+      "loss": 0.4437,
+      "step": 6946
+    },
+    {
+      "epoch": 1.063045141545524,
+      "grad_norm": 2.078240326941889,
+      "learning_rate": 9.46620537426022e-06,
+      "loss": 0.4313,
+      "step": 6947
+    },
+    {
+      "epoch": 1.0631981637337413,
+      "grad_norm": 2.130682788759023,
+      "learning_rate": 9.463730740771025e-06,
+      "loss": 0.4409,
+      "step": 6948
+    },
+    {
+      "epoch": 1.0633511859219587,
+      "grad_norm": 2.30060169305822,
+      "learning_rate": 9.461256140216218e-06,
+      "loss": 0.4819,
+      "step": 6949
+    },
+    {
+      "epoch": 1.063504208110176,
+      "grad_norm": 2.519758537925622,
+      "learning_rate": 9.45878157274778e-06,
+      "loss": 0.5226,
+      "step": 6950
+    },
+    {
+      "epoch": 1.0636572302983933,
+      "grad_norm": 2.017059855190871,
+      "learning_rate": 9.456307038517682e-06,
+      "loss": 0.3927,
+      "step": 6951
+    },
+    {
+      "epoch": 1.0638102524866107,
+      "grad_norm": 2.4161156149987413,
+      "learning_rate": 9.453832537677899e-06,
+      "loss": 0.4671,
+      "step": 6952
+    },
+    {
+      "epoch": 1.0639632746748278,
+      "grad_norm": 2.4352087813015295,
+      "learning_rate": 9.451358070380395e-06,
+      "loss": 0.5307,
+      "step": 6953
+    },
+    {
+      "epoch": 1.064116296863045,
+      "grad_norm": 2.4821772090895924,
+      "learning_rate": 9.448883636777136e-06,
+      "loss": 0.6598,
+      "step": 6954
+    },
+    {
+      "epoch": 1.0642693190512624,
+      "grad_norm": 2.393312834043304,
+      "learning_rate": 9.446409237020093e-06,
+      "loss": 0.5168,
+      "step": 6955
+    },
+    {
+      "epoch": 1.0644223412394798,
+      "grad_norm": 2.321195584837522,
+      "learning_rate": 9.443934871261223e-06,
+      "loss": 0.4782,
+      "step": 6956
+    },
+    {
+      "epoch": 1.064575363427697,
+      "grad_norm": 2.083851712582457,
+      "learning_rate": 9.441460539652485e-06,
+      "loss": 0.3878,
+      "step": 6957
+    },
+    {
+      "epoch": 1.0647283856159142,
+      "grad_norm": 2.2876127565948106,
+      "learning_rate": 9.438986242345844e-06,
+      "loss": 0.4808,
+      "step": 6958
+    },
+    {
+      "epoch": 1.0648814078041315,
+      "grad_norm": 2.328857160236901,
+      "learning_rate": 9.436511979493249e-06,
+      "loss": 0.4874,
+      "step": 6959
+    },
+    {
+      "epoch": 1.0650344299923489,
+      "grad_norm": 2.483784227392007,
+      "learning_rate": 9.43403775124666e-06,
+      "loss": 0.4848,
+      "step": 6960
+    },
+    {
+      "epoch": 1.0651874521805662,
+      "grad_norm": 2.139738952325263,
+      "learning_rate": 9.43156355775803e-06,
+      "loss": 0.4908,
+      "step": 6961
+    },
+    {
+      "epoch": 1.0653404743687835,
+      "grad_norm": 2.215075116418688,
+      "learning_rate": 9.429089399179298e-06,
+      "loss": 0.4778,
+      "step": 6962
+    },
+    {
+      "epoch": 1.0654934965570009,
+      "grad_norm": 1.9814013494006328,
+      "learning_rate": 9.426615275662426e-06,
+      "loss": 0.441,
+      "step": 6963
+    },
+    {
+      "epoch": 1.065646518745218,
+      "grad_norm": 2.275252087564853,
+      "learning_rate": 9.424141187359347e-06,
+      "loss": 0.5158,
+      "step": 6964
+    },
+    {
+      "epoch": 1.0657995409334353,
+      "grad_norm": 2.2604361487894638,
+      "learning_rate": 9.421667134422018e-06,
+      "loss": 0.4746,
+      "step": 6965
+    },
+    {
+      "epoch": 1.0659525631216527,
+      "grad_norm": 2.539253336905643,
+      "learning_rate": 9.419193117002367e-06,
+      "loss": 0.4601,
+      "step": 6966
+    },
+    {
+      "epoch": 1.06610558530987,
+      "grad_norm": 2.3308452441992413,
+      "learning_rate": 9.416719135252338e-06,
+      "loss": 0.469,
+      "step": 6967
+    },
+    {
+      "epoch": 1.0662586074980873,
+      "grad_norm": 2.1799849355110417,
+      "learning_rate": 9.414245189323875e-06,
+      "loss": 0.4691,
+      "step": 6968
+    },
+    {
+      "epoch": 1.0664116296863044,
+      "grad_norm": 2.438373820976639,
+      "learning_rate": 9.411771279368902e-06,
+      "loss": 0.4941,
+      "step": 6969
+    },
+    {
+      "epoch": 1.0665646518745218,
+      "grad_norm": 2.1442590483045816,
+      "learning_rate": 9.409297405539355e-06,
+      "loss": 0.3921,
+      "step": 6970
+    },
+    {
+      "epoch": 1.066717674062739,
+      "grad_norm": 2.3576024377638167,
+      "learning_rate": 9.40682356798717e-06,
+      "loss": 0.4943,
+      "step": 6971
+    },
+    {
+      "epoch": 1.0668706962509564,
+      "grad_norm": 2.3218424806222435,
+      "learning_rate": 9.404349766864262e-06,
+      "loss": 0.5805,
+      "step": 6972
+    },
+    {
+      "epoch": 1.0670237184391738,
+      "grad_norm": 2.4645763128765745,
+      "learning_rate": 9.401876002322573e-06,
+      "loss": 0.4609,
+      "step": 6973
+    },
+    {
+      "epoch": 1.0671767406273909,
+      "grad_norm": 2.3130841744661073,
+      "learning_rate": 9.399402274514017e-06,
+      "loss": 0.5233,
+      "step": 6974
+    },
+    {
+      "epoch": 1.0673297628156082,
+      "grad_norm": 2.1876222922209703,
+      "learning_rate": 9.39692858359051e-06,
+      "loss": 0.4513,
+      "step": 6975
+    },
+    {
+      "epoch": 1.0674827850038255,
+      "grad_norm": 2.033847811009393,
+      "learning_rate": 9.394454929703987e-06,
+      "loss": 0.3781,
+      "step": 6976
+    },
+    {
+      "epoch": 1.0676358071920429,
+      "grad_norm": 2.482097524069422,
+      "learning_rate": 9.391981313006354e-06,
+      "loss": 0.4744,
+      "step": 6977
+    },
+    {
+      "epoch": 1.0677888293802602,
+      "grad_norm": 2.346499473043113,
+      "learning_rate": 9.38950773364952e-06,
+      "loss": 0.4851,
+      "step": 6978
+    },
+    {
+      "epoch": 1.0679418515684773,
+      "grad_norm": 2.157264052254601,
+      "learning_rate": 9.387034191785413e-06,
+      "loss": 0.4382,
+      "step": 6979
+    },
+    {
+      "epoch": 1.0680948737566947,
+      "grad_norm": 2.3721231776727056,
+      "learning_rate": 9.384560687565927e-06,
+      "loss": 0.485,
+      "step": 6980
+    },
+    {
+      "epoch": 1.068247895944912,
+      "grad_norm": 2.2298549333189306,
+      "learning_rate": 9.382087221142981e-06,
+      "loss": 0.4579,
+      "step": 6981
+    },
+    {
+      "epoch": 1.0684009181331293,
+      "grad_norm": 2.0889715056745377,
+      "learning_rate": 9.379613792668478e-06,
+      "loss": 0.4872,
+      "step": 6982
+    },
+    {
+      "epoch": 1.0685539403213467,
+      "grad_norm": 2.0541752026840365,
+      "learning_rate": 9.377140402294315e-06,
+      "loss": 0.4358,
+      "step": 6983
+    },
+    {
+      "epoch": 1.068706962509564,
+      "grad_norm": 2.327139410292757,
+      "learning_rate": 9.374667050172397e-06,
+      "loss": 0.4975,
+      "step": 6984
+    },
+    {
+      "epoch": 1.068859984697781,
+      "grad_norm": 2.367702726919621,
+      "learning_rate": 9.372193736454626e-06,
+      "loss": 0.5055,
+      "step": 6985
+    },
+    {
+      "epoch": 1.0690130068859984,
+      "grad_norm": 2.178558568246285,
+      "learning_rate": 9.369720461292889e-06,
+      "loss": 0.4234,
+      "step": 6986
+    },
+    {
+      "epoch": 1.0691660290742158,
+      "grad_norm": 2.2406729508801653,
+      "learning_rate": 9.367247224839084e-06,
+      "loss": 0.4451,
+      "step": 6987
+    },
+    {
+      "epoch": 1.069319051262433,
+      "grad_norm": 2.2256944403639953,
+      "learning_rate": 9.364774027245102e-06,
+      "loss": 0.443,
+      "step": 6988
+    },
+    {
+      "epoch": 1.0694720734506504,
+      "grad_norm": 2.0697411441110463,
+      "learning_rate": 9.362300868662837e-06,
+      "loss": 0.441,
+      "step": 6989
+    },
+    {
+      "epoch": 1.0696250956388675,
+      "grad_norm": 2.2535106828944182,
+      "learning_rate": 9.359827749244168e-06,
+      "loss": 0.3842,
+      "step": 6990
+    },
+    {
+      "epoch": 1.0697781178270849,
+      "grad_norm": 2.3355363703052214,
+      "learning_rate": 9.357354669140977e-06,
+      "loss": 0.4452,
+      "step": 6991
+    },
+    {
+      "epoch": 1.0699311400153022,
+      "grad_norm": 2.144156733012759,
+      "learning_rate": 9.354881628505156e-06,
+      "loss": 0.4491,
+      "step": 6992
+    },
+    {
+      "epoch": 1.0700841622035195,
+      "grad_norm": 2.304338238051539,
+      "learning_rate": 9.352408627488574e-06,
+      "loss": 0.4392,
+      "step": 6993
+    },
+    {
+      "epoch": 1.0702371843917369,
+      "grad_norm": 2.207433271716185,
+      "learning_rate": 9.349935666243112e-06,
+      "loss": 0.3962,
+      "step": 6994
+    },
+    {
+      "epoch": 1.0703902065799542,
+      "grad_norm": 2.339465394300809,
+      "learning_rate": 9.347462744920646e-06,
+      "loss": 0.5691,
+      "step": 6995
+    },
+    {
+      "epoch": 1.0705432287681713,
+      "grad_norm": 2.2116556612187286,
+      "learning_rate": 9.344989863673043e-06,
+      "loss": 0.3849,
+      "step": 6996
+    },
+    {
+      "epoch": 1.0706962509563887,
+      "grad_norm": 2.2286947116609244,
+      "learning_rate": 9.342517022652176e-06,
+      "loss": 0.4345,
+      "step": 6997
+    },
+    {
+      "epoch": 1.070849273144606,
+      "grad_norm": 2.523389393218545,
+      "learning_rate": 9.340044222009913e-06,
+      "loss": 0.5253,
+      "step": 6998
+    },
+    {
+      "epoch": 1.0710022953328233,
+      "grad_norm": 2.2312942926442703,
+      "learning_rate": 9.337571461898112e-06,
+      "loss": 0.504,
+      "step": 6999
+    },
+    {
+      "epoch": 1.0711553175210407,
+      "grad_norm": 2.2598765293416876,
+      "learning_rate": 9.33509874246864e-06,
+      "loss": 0.5225,
+      "step": 7000
+    },
+    {
+      "epoch": 1.0713083397092578,
+      "grad_norm": 2.503438620842996,
+      "learning_rate": 9.332626063873354e-06,
+      "loss": 0.5315,
+      "step": 7001
+    },
+    {
+      "epoch": 1.071461361897475,
+      "grad_norm": 2.5807369463416916,
+      "learning_rate": 9.330153426264117e-06,
+      "loss": 0.4877,
+      "step": 7002
+    },
+    {
+      "epoch": 1.0716143840856924,
+      "grad_norm": 2.568301950382052,
+      "learning_rate": 9.327680829792775e-06,
+      "loss": 0.5205,
+      "step": 7003
+    },
+    {
+      "epoch": 1.0717674062739098,
+      "grad_norm": 2.182053222736725,
+      "learning_rate": 9.325208274611184e-06,
+      "loss": 0.436,
+      "step": 7004
+    },
+    {
+      "epoch": 1.071920428462127,
+      "grad_norm": 2.265886222181865,
+      "learning_rate": 9.322735760871194e-06,
+      "loss": 0.4027,
+      "step": 7005
+    },
+    {
+      "epoch": 1.0720734506503442,
+      "grad_norm": 2.2566761587437125,
+      "learning_rate": 9.32026328872465e-06,
+      "loss": 0.4805,
+      "step": 7006
+    },
+    {
+      "epoch": 1.0722264728385615,
+      "grad_norm": 2.5470439085229133,
+      "learning_rate": 9.317790858323393e-06,
+      "loss": 0.5402,
+      "step": 7007
+    },
+    {
+      "epoch": 1.0723794950267789,
+      "grad_norm": 2.473910127790221,
+      "learning_rate": 9.315318469819275e-06,
+      "loss": 0.4526,
+      "step": 7008
+    },
+    {
+      "epoch": 1.0725325172149962,
+      "grad_norm": 2.1728995220377776,
+      "learning_rate": 9.31284612336412e-06,
+      "loss": 0.4184,
+      "step": 7009
+    },
+    {
+      "epoch": 1.0726855394032135,
+      "grad_norm": 2.166777588898116,
+      "learning_rate": 9.310373819109779e-06,
+      "loss": 0.4201,
+      "step": 7010
+    },
+    {
+      "epoch": 1.0728385615914307,
+      "grad_norm": 2.2110283208174177,
+      "learning_rate": 9.30790155720808e-06,
+      "loss": 0.4783,
+      "step": 7011
+    },
+    {
+      "epoch": 1.072991583779648,
+      "grad_norm": 2.2649284810005117,
+      "learning_rate": 9.30542933781085e-06,
+      "loss": 0.3991,
+      "step": 7012
+    },
+    {
+      "epoch": 1.0731446059678653,
+      "grad_norm": 2.1904501370858522,
+      "learning_rate": 9.302957161069921e-06,
+      "loss": 0.4015,
+      "step": 7013
+    },
+    {
+      "epoch": 1.0732976281560827,
+      "grad_norm": 2.1730718132967017,
+      "learning_rate": 9.300485027137125e-06,
+      "loss": 0.4699,
+      "step": 7014
+    },
+    {
+      "epoch": 1.0734506503443,
+      "grad_norm": 2.3653172817867314,
+      "learning_rate": 9.298012936164275e-06,
+      "loss": 0.5384,
+      "step": 7015
+    },
+    {
+      "epoch": 1.0736036725325173,
+      "grad_norm": 2.1901146142058394,
+      "learning_rate": 9.295540888303198e-06,
+      "loss": 0.5113,
+      "step": 7016
+    },
+    {
+      "epoch": 1.0737566947207344,
+      "grad_norm": 2.384766068642065,
+      "learning_rate": 9.29306888370571e-06,
+      "loss": 0.5185,
+      "step": 7017
+    },
+    {
+      "epoch": 1.0739097169089518,
+      "grad_norm": 2.6746763962083167,
+      "learning_rate": 9.29059692252363e-06,
+      "loss": 0.4647,
+      "step": 7018
+    },
+    {
+      "epoch": 1.074062739097169,
+      "grad_norm": 2.353980841278368,
+      "learning_rate": 9.288125004908768e-06,
+      "loss": 0.4805,
+      "step": 7019
+    },
+    {
+      "epoch": 1.0742157612853864,
+      "grad_norm": 2.311592396149525,
+      "learning_rate": 9.28565313101293e-06,
+      "loss": 0.5211,
+      "step": 7020
+    },
+    {
+      "epoch": 1.0743687834736038,
+      "grad_norm": 2.252228528893847,
+      "learning_rate": 9.283181300987934e-06,
+      "loss": 0.4451,
+      "step": 7021
+    },
+    {
+      "epoch": 1.0745218056618209,
+      "grad_norm": 2.2982993979133925,
+      "learning_rate": 9.280709514985575e-06,
+      "loss": 0.3907,
+      "step": 7022
+    },
+    {
+      "epoch": 1.0746748278500382,
+      "grad_norm": 2.3070546754874295,
+      "learning_rate": 9.278237773157657e-06,
+      "loss": 0.4716,
+      "step": 7023
+    },
+    {
+      "epoch": 1.0748278500382555,
+      "grad_norm": 2.6131421219502498,
+      "learning_rate": 9.275766075655986e-06,
+      "loss": 0.5161,
+      "step": 7024
+    },
+    {
+      "epoch": 1.0749808722264729,
+      "grad_norm": 2.4042427042020873,
+      "learning_rate": 9.27329442263235e-06,
+      "loss": 0.5227,
+      "step": 7025
+    },
+    {
+      "epoch": 1.0751338944146902,
+      "grad_norm": 2.3104549373393652,
+      "learning_rate": 9.270822814238547e-06,
+      "loss": 0.4751,
+      "step": 7026
+    },
+    {
+      "epoch": 1.0752869166029075,
+      "grad_norm": 2.1835962666687254,
+      "learning_rate": 9.268351250626368e-06,
+      "loss": 0.5067,
+      "step": 7027
+    },
+    {
+      "epoch": 1.0754399387911246,
+      "grad_norm": 2.2162958243208735,
+      "learning_rate": 9.265879731947599e-06,
+      "loss": 0.4971,
+      "step": 7028
+    },
+    {
+      "epoch": 1.075592960979342,
+      "grad_norm": 2.342103971102228,
+      "learning_rate": 9.263408258354028e-06,
+      "loss": 0.4492,
+      "step": 7029
+    },
+    {
+      "epoch": 1.0757459831675593,
+      "grad_norm": 2.5572314302614387,
+      "learning_rate": 9.260936829997437e-06,
+      "loss": 0.5074,
+      "step": 7030
+    },
+    {
+      "epoch": 1.0758990053557766,
+      "grad_norm": 2.2950024298977123,
+      "learning_rate": 9.258465447029609e-06,
+      "loss": 0.3915,
+      "step": 7031
+    },
+    {
+      "epoch": 1.076052027543994,
+      "grad_norm": 2.3647579223440163,
+      "learning_rate": 9.255994109602319e-06,
+      "loss": 0.5398,
+      "step": 7032
+    },
+    {
+      "epoch": 1.076205049732211,
+      "grad_norm": 2.336276675467057,
+      "learning_rate": 9.253522817867337e-06,
+      "loss": 0.4975,
+      "step": 7033
+    },
+    {
+      "epoch": 1.0763580719204284,
+      "grad_norm": 2.569841908595689,
+      "learning_rate": 9.251051571976444e-06,
+      "loss": 0.5007,
+      "step": 7034
+    },
+    {
+      "epoch": 1.0765110941086458,
+      "grad_norm": 1.9321188143755093,
+      "learning_rate": 9.248580372081403e-06,
+      "loss": 0.4187,
+      "step": 7035
+    },
+    {
+      "epoch": 1.076664116296863,
+      "grad_norm": 1.993423653114654,
+      "learning_rate": 9.246109218333978e-06,
+      "loss": 0.4471,
+      "step": 7036
+    },
+    {
+      "epoch": 1.0768171384850804,
+      "grad_norm": 2.8320947879207923,
+      "learning_rate": 9.243638110885938e-06,
+      "loss": 0.3765,
+      "step": 7037
+    },
+    {
+      "epoch": 1.0769701606732975,
+      "grad_norm": 2.121544366157612,
+      "learning_rate": 9.241167049889038e-06,
+      "loss": 0.4361,
+      "step": 7038
+    },
+    {
+      "epoch": 1.0771231828615149,
+      "grad_norm": 2.1231269443520153,
+      "learning_rate": 9.23869603549504e-06,
+      "loss": 0.4724,
+      "step": 7039
+    },
+    {
+      "epoch": 1.0772762050497322,
+      "grad_norm": 2.1396557730209924,
+      "learning_rate": 9.236225067855697e-06,
+      "loss": 0.4575,
+      "step": 7040
+    },
+    {
+      "epoch": 1.0774292272379495,
+      "grad_norm": 2.444995460337678,
+      "learning_rate": 9.233754147122759e-06,
+      "loss": 0.5192,
+      "step": 7041
+    },
+    {
+      "epoch": 1.0775822494261669,
+      "grad_norm": 2.09131681001774,
+      "learning_rate": 9.231283273447976e-06,
+      "loss": 0.417,
+      "step": 7042
+    },
+    {
+      "epoch": 1.077735271614384,
+      "grad_norm": 2.412492987813149,
+      "learning_rate": 9.228812446983098e-06,
+      "loss": 0.4716,
+      "step": 7043
+    },
+    {
+      "epoch": 1.0778882938026013,
+      "grad_norm": 2.376792204202723,
+      "learning_rate": 9.22634166787986e-06,
+      "loss": 0.5042,
+      "step": 7044
+    },
+    {
+      "epoch": 1.0780413159908186,
+      "grad_norm": 2.258267038660303,
+      "learning_rate": 9.22387093629001e-06,
+      "loss": 0.5356,
+      "step": 7045
+    },
+    {
+      "epoch": 1.078194338179036,
+      "grad_norm": 2.140885159688897,
+      "learning_rate": 9.221400252365279e-06,
+      "loss": 0.402,
+      "step": 7046
+    },
+    {
+      "epoch": 1.0783473603672533,
+      "grad_norm": 2.299639935119907,
+      "learning_rate": 9.218929616257406e-06,
+      "loss": 0.4874,
+      "step": 7047
+    },
+    {
+      "epoch": 1.0785003825554704,
+      "grad_norm": 2.0884411596909005,
+      "learning_rate": 9.216459028118123e-06,
+      "loss": 0.4117,
+      "step": 7048
+    },
+    {
+      "epoch": 1.0786534047436878,
+      "grad_norm": 2.3604023546791613,
+      "learning_rate": 9.213988488099152e-06,
+      "loss": 0.4785,
+      "step": 7049
+    },
+    {
+      "epoch": 1.078806426931905,
+      "grad_norm": 2.2206333646096197,
+      "learning_rate": 9.211517996352226e-06,
+      "loss": 0.5144,
+      "step": 7050
+    },
+    {
+      "epoch": 1.0789594491201224,
+      "grad_norm": 2.4418928939273794,
+      "learning_rate": 9.209047553029065e-06,
+      "loss": 0.5097,
+      "step": 7051
+    },
+    {
+      "epoch": 1.0791124713083398,
+      "grad_norm": 2.4200213816818756,
+      "learning_rate": 9.206577158281384e-06,
+      "loss": 0.4596,
+      "step": 7052
+    },
+    {
+      "epoch": 1.079265493496557,
+      "grad_norm": 2.1039066805422553,
+      "learning_rate": 9.204106812260908e-06,
+      "loss": 0.5227,
+      "step": 7053
+    },
+    {
+      "epoch": 1.0794185156847742,
+      "grad_norm": 2.3042584412347136,
+      "learning_rate": 9.20163651511934e-06,
+      "loss": 0.4305,
+      "step": 7054
+    },
+    {
+      "epoch": 1.0795715378729915,
+      "grad_norm": 2.4025169673069118,
+      "learning_rate": 9.199166267008404e-06,
+      "loss": 0.513,
+      "step": 7055
+    },
+    {
+      "epoch": 1.0797245600612089,
+      "grad_norm": 2.46461175249102,
+      "learning_rate": 9.196696068079799e-06,
+      "loss": 0.4345,
+      "step": 7056
+    },
+    {
+      "epoch": 1.0798775822494262,
+      "grad_norm": 2.0642559537388334,
+      "learning_rate": 9.194225918485226e-06,
+      "loss": 0.3485,
+      "step": 7057
+    },
+    {
+      "epoch": 1.0800306044376435,
+      "grad_norm": 2.305597405190117,
+      "learning_rate": 9.191755818376397e-06,
+      "loss": 0.4684,
+      "step": 7058
+    },
+    {
+      "epoch": 1.0801836266258606,
+      "grad_norm": 2.153729065281024,
+      "learning_rate": 9.189285767905006e-06,
+      "loss": 0.4609,
+      "step": 7059
+    },
+    {
+      "epoch": 1.080336648814078,
+      "grad_norm": 2.151763990460353,
+      "learning_rate": 9.18681576722274e-06,
+      "loss": 0.5002,
+      "step": 7060
+    },
+    {
+      "epoch": 1.0804896710022953,
+      "grad_norm": 2.121608247611442,
+      "learning_rate": 9.184345816481307e-06,
+      "loss": 0.6741,
+      "step": 7061
+    },
+    {
+      "epoch": 1.0806426931905126,
+      "grad_norm": 2.585835936462805,
+      "learning_rate": 9.181875915832381e-06,
+      "loss": 0.5757,
+      "step": 7062
+    },
+    {
+      "epoch": 1.08079571537873,
+      "grad_norm": 2.5048579827782746,
+      "learning_rate": 9.17940606542766e-06,
+      "loss": 0.4583,
+      "step": 7063
+    },
+    {
+      "epoch": 1.0809487375669473,
+      "grad_norm": 2.149409121009293,
+      "learning_rate": 9.176936265418825e-06,
+      "loss": 0.5809,
+      "step": 7064
+    },
+    {
+      "epoch": 1.0811017597551644,
+      "grad_norm": 2.1026339389941575,
+      "learning_rate": 9.174466515957548e-06,
+      "loss": 0.3908,
+      "step": 7065
+    },
+    {
+      "epoch": 1.0812547819433818,
+      "grad_norm": 2.281688039861431,
+      "learning_rate": 9.171996817195516e-06,
+      "loss": 0.4972,
+      "step": 7066
+    },
+    {
+      "epoch": 1.081407804131599,
+      "grad_norm": 2.359209844798118,
+      "learning_rate": 9.169527169284395e-06,
+      "loss": 0.5039,
+      "step": 7067
+    },
+    {
+      "epoch": 1.0815608263198164,
+      "grad_norm": 2.284727340548648,
+      "learning_rate": 9.167057572375863e-06,
+      "loss": 0.4852,
+      "step": 7068
+    },
+    {
+      "epoch": 1.0817138485080338,
+      "grad_norm": 2.4648971458789917,
+      "learning_rate": 9.164588026621582e-06,
+      "loss": 0.4096,
+      "step": 7069
+    },
+    {
+      "epoch": 1.0818668706962509,
+      "grad_norm": 2.050205220971181,
+      "learning_rate": 9.162118532173217e-06,
+      "loss": 0.4125,
+      "step": 7070
+    },
+    {
+      "epoch": 1.0820198928844682,
+      "grad_norm": 2.0811476169259704,
+      "learning_rate": 9.159649089182436e-06,
+      "loss": 0.4692,
+      "step": 7071
+    },
+    {
+      "epoch": 1.0821729150726855,
+      "grad_norm": 2.217536427398098,
+      "learning_rate": 9.157179697800889e-06,
+      "loss": 0.4477,
+      "step": 7072
+    },
+    {
+      "epoch": 1.0823259372609029,
+      "grad_norm": 2.542074757986943,
+      "learning_rate": 9.154710358180231e-06,
+      "loss": 0.5461,
+      "step": 7073
+    },
+    {
+      "epoch": 1.0824789594491202,
+      "grad_norm": 2.050089241280042,
+      "learning_rate": 9.152241070472122e-06,
+      "loss": 0.4656,
+      "step": 7074
+    },
+    {
+      "epoch": 1.0826319816373373,
+      "grad_norm": 2.2914035705247695,
+      "learning_rate": 9.149771834828201e-06,
+      "loss": 0.491,
+      "step": 7075
+    },
+    {
+      "epoch": 1.0827850038255546,
+      "grad_norm": 2.243598099910007,
+      "learning_rate": 9.147302651400119e-06,
+      "loss": 0.5017,
+      "step": 7076
+    },
+    {
+      "epoch": 1.082938026013772,
+      "grad_norm": 2.6116613995542837,
+      "learning_rate": 9.144833520339518e-06,
+      "loss": 0.5296,
+      "step": 7077
+    },
+    {
+      "epoch": 1.0830910482019893,
+      "grad_norm": 2.2604014415370592,
+      "learning_rate": 9.142364441798035e-06,
+      "loss": 0.4793,
+      "step": 7078
+    },
+    {
+      "epoch": 1.0832440703902066,
+      "grad_norm": 2.140404673851035,
+      "learning_rate": 9.139895415927307e-06,
+      "loss": 0.3525,
+      "step": 7079
+    },
+    {
+      "epoch": 1.0833970925784238,
+      "grad_norm": 2.2760551176385913,
+      "learning_rate": 9.137426442878969e-06,
+      "loss": 0.4887,
+      "step": 7080
+    },
+    {
+      "epoch": 1.083550114766641,
+      "grad_norm": 2.3819989085455298,
+      "learning_rate": 9.134957522804641e-06,
+      "loss": 0.4958,
+      "step": 7081
+    },
+    {
+      "epoch": 1.0837031369548584,
+      "grad_norm": 2.4328489903808523,
+      "learning_rate": 9.13248865585596e-06,
+      "loss": 0.4582,
+      "step": 7082
+    },
+    {
+      "epoch": 1.0838561591430758,
+      "grad_norm": 2.1001441662606575,
+      "learning_rate": 9.130019842184542e-06,
+      "loss": 0.4808,
+      "step": 7083
+    },
+    {
+      "epoch": 1.084009181331293,
+      "grad_norm": 2.126839624383373,
+      "learning_rate": 9.127551081942013e-06,
+      "loss": 0.505,
+      "step": 7084
+    },
+    {
+      "epoch": 1.0841622035195104,
+      "grad_norm": 2.0661438691181386,
+      "learning_rate": 9.125082375279983e-06,
+      "loss": 0.4676,
+      "step": 7085
+    },
+    {
+      "epoch": 1.0843152257077275,
+      "grad_norm": 2.104196112200205,
+      "learning_rate": 9.122613722350064e-06,
+      "loss": 0.383,
+      "step": 7086
+    },
+    {
+      "epoch": 1.0844682478959449,
+      "grad_norm": 2.2568665355471746,
+      "learning_rate": 9.120145123303874e-06,
+      "loss": 0.5048,
+      "step": 7087
+    },
+    {
+      "epoch": 1.0846212700841622,
+      "grad_norm": 2.2024165587118176,
+      "learning_rate": 9.11767657829301e-06,
+      "loss": 0.4727,
+      "step": 7088
+    },
+    {
+      "epoch": 1.0847742922723795,
+      "grad_norm": 2.612232217407369,
+      "learning_rate": 9.11520808746908e-06,
+      "loss": 0.5607,
+      "step": 7089
+    },
+    {
+      "epoch": 1.0849273144605969,
+      "grad_norm": 2.1450307502448847,
+      "learning_rate": 9.112739650983685e-06,
+      "loss": 0.5029,
+      "step": 7090
+    },
+    {
+      "epoch": 1.085080336648814,
+      "grad_norm": 2.578778549735402,
+      "learning_rate": 9.110271268988415e-06,
+      "loss": 0.5127,
+      "step": 7091
+    },
+    {
+      "epoch": 1.0852333588370313,
+      "grad_norm": 2.1489916563880165,
+      "learning_rate": 9.107802941634869e-06,
+      "loss": 0.4452,
+      "step": 7092
+    },
+    {
+      "epoch": 1.0853863810252486,
+      "grad_norm": 2.2916701107274147,
+      "learning_rate": 9.105334669074637e-06,
+      "loss": 0.4841,
+      "step": 7093
+    },
+    {
+      "epoch": 1.085539403213466,
+      "grad_norm": 2.152474576028727,
+      "learning_rate": 9.102866451459299e-06,
+      "loss": 0.416,
+      "step": 7094
+    },
+    {
+      "epoch": 1.0856924254016833,
+      "grad_norm": 1.9391836681067787,
+      "learning_rate": 9.100398288940443e-06,
+      "loss": 0.3911,
+      "step": 7095
+    },
+    {
+      "epoch": 1.0858454475899006,
+      "grad_norm": 2.284477748078566,
+      "learning_rate": 9.09793018166965e-06,
+      "loss": 0.4323,
+      "step": 7096
+    },
+    {
+      "epoch": 1.0859984697781178,
+      "grad_norm": 2.0810449204826558,
+      "learning_rate": 9.09546212979849e-06,
+      "loss": 0.4461,
+      "step": 7097
+    },
+    {
+      "epoch": 1.086151491966335,
+      "grad_norm": 2.102591065248399,
+      "learning_rate": 9.09299413347854e-06,
+      "loss": 0.4236,
+      "step": 7098
+    },
+    {
+      "epoch": 1.0863045141545524,
+      "grad_norm": 2.285514466521769,
+      "learning_rate": 9.090526192861367e-06,
+      "loss": 0.4432,
+      "step": 7099
+    },
+    {
+      "epoch": 1.0864575363427698,
+      "grad_norm": 2.2247005114609064,
+      "learning_rate": 9.088058308098542e-06,
+      "loss": 0.4757,
+      "step": 7100
+    },
+    {
+      "epoch": 1.086610558530987,
+      "grad_norm": 2.379153491410123,
+      "learning_rate": 9.085590479341623e-06,
+      "loss": 0.4765,
+      "step": 7101
+    },
+    {
+      "epoch": 1.0867635807192042,
+      "grad_norm": 2.321895207959366,
+      "learning_rate": 9.083122706742166e-06,
+      "loss": 0.5086,
+      "step": 7102
+    },
+    {
+      "epoch": 1.0869166029074215,
+      "grad_norm": 2.3449443560981345,
+      "learning_rate": 9.080654990451736e-06,
+      "loss": 0.4743,
+      "step": 7103
+    },
+    {
+      "epoch": 1.0870696250956389,
+      "grad_norm": 2.169152507590791,
+      "learning_rate": 9.078187330621876e-06,
+      "loss": 0.4113,
+      "step": 7104
+    },
+    {
+      "epoch": 1.0872226472838562,
+      "grad_norm": 2.2823786746054187,
+      "learning_rate": 9.075719727404138e-06,
+      "loss": 0.4769,
+      "step": 7105
+    },
+    {
+      "epoch": 1.0873756694720735,
+      "grad_norm": 2.2438777723684966,
+      "learning_rate": 9.073252180950072e-06,
+      "loss": 0.4935,
+      "step": 7106
+    },
+    {
+      "epoch": 1.0875286916602906,
+      "grad_norm": 2.191000328520221,
+      "learning_rate": 9.070784691411208e-06,
+      "loss": 0.4083,
+      "step": 7107
+    },
+    {
+      "epoch": 1.087681713848508,
+      "grad_norm": 2.2185239900801275,
+      "learning_rate": 9.068317258939096e-06,
+      "loss": 0.4524,
+      "step": 7108
+    },
+    {
+      "epoch": 1.0878347360367253,
+      "grad_norm": 2.469228148557464,
+      "learning_rate": 9.065849883685265e-06,
+      "loss": 0.4708,
+      "step": 7109
+    },
+    {
+      "epoch": 1.0879877582249426,
+      "grad_norm": 2.148317257106404,
+      "learning_rate": 9.063382565801247e-06,
+      "loss": 0.4117,
+      "step": 7110
+    },
+    {
+      "epoch": 1.08814078041316,
+      "grad_norm": 2.46960126434831,
+      "learning_rate": 9.060915305438567e-06,
+      "loss": 0.5797,
+      "step": 7111
+    },
+    {
+      "epoch": 1.088293802601377,
+      "grad_norm": 2.5654903458195295,
+      "learning_rate": 9.058448102748753e-06,
+      "loss": 0.5204,
+      "step": 7112
+    },
+    {
+      "epoch": 1.0884468247895944,
+      "grad_norm": 2.5226254011755085,
+      "learning_rate": 9.055980957883329e-06,
+      "loss": 0.5013,
+      "step": 7113
+    },
+    {
+      "epoch": 1.0885998469778118,
+      "grad_norm": 2.444152682210665,
+      "learning_rate": 9.053513870993803e-06,
+      "loss": 0.4974,
+      "step": 7114
+    },
+    {
+      "epoch": 1.088752869166029,
+      "grad_norm": 2.1433965549708223,
+      "learning_rate": 9.051046842231693e-06,
+      "loss": 0.3761,
+      "step": 7115
+    },
+    {
+      "epoch": 1.0889058913542464,
+      "grad_norm": 2.2244339091179848,
+      "learning_rate": 9.04857987174851e-06,
+      "loss": 0.4533,
+      "step": 7116
+    },
+    {
+      "epoch": 1.0890589135424638,
+      "grad_norm": 2.3900548171797342,
+      "learning_rate": 9.04611295969576e-06,
+      "loss": 0.4394,
+      "step": 7117
+    },
+    {
+      "epoch": 1.0892119357306809,
+      "grad_norm": 2.477441551752149,
+      "learning_rate": 9.043646106224942e-06,
+      "loss": 0.4756,
+      "step": 7118
+    },
+    {
+      "epoch": 1.0893649579188982,
+      "grad_norm": 2.2301190063141707,
+      "learning_rate": 9.04117931148756e-06,
+      "loss": 0.4872,
+      "step": 7119
+    },
+    {
+      "epoch": 1.0895179801071155,
+      "grad_norm": 2.2765796549488373,
+      "learning_rate": 9.038712575635105e-06,
+      "loss": 0.5097,
+      "step": 7120
+    },
+    {
+      "epoch": 1.0896710022953329,
+      "grad_norm": 2.516283088910062,
+      "learning_rate": 9.036245898819074e-06,
+      "loss": 0.4335,
+      "step": 7121
+    },
+    {
+      "epoch": 1.0898240244835502,
+      "grad_norm": 2.279848242904234,
+      "learning_rate": 9.033779281190953e-06,
+      "loss": 0.5019,
+      "step": 7122
+    },
+    {
+      "epoch": 1.0899770466717673,
+      "grad_norm": 2.1190201556445114,
+      "learning_rate": 9.031312722902223e-06,
+      "loss": 0.4462,
+      "step": 7123
+    },
+    {
+      "epoch": 1.0901300688599846,
+      "grad_norm": 2.3137880793300463,
+      "learning_rate": 9.028846224104369e-06,
+      "loss": 0.4546,
+      "step": 7124
+    },
+    {
+      "epoch": 1.090283091048202,
+      "grad_norm": 2.1680262820024723,
+      "learning_rate": 9.026379784948873e-06,
+      "loss": 0.4073,
+      "step": 7125
+    },
+    {
+      "epoch": 1.0904361132364193,
+      "grad_norm": 2.2731966671711166,
+      "learning_rate": 9.023913405587196e-06,
+      "loss": 0.4626,
+      "step": 7126
+    },
+    {
+      "epoch": 1.0905891354246366,
+      "grad_norm": 2.066266313161121,
+      "learning_rate": 9.021447086170818e-06,
+      "loss": 0.4734,
+      "step": 7127
+    },
+    {
+      "epoch": 1.090742157612854,
+      "grad_norm": 1.9706951129586714,
+      "learning_rate": 9.0189808268512e-06,
+      "loss": 0.4781,
+      "step": 7128
+    },
+    {
+      "epoch": 1.090895179801071,
+      "grad_norm": 2.141784621387068,
+      "learning_rate": 9.016514627779811e-06,
+      "loss": 0.484,
+      "step": 7129
+    },
+    {
+      "epoch": 1.0910482019892884,
+      "grad_norm": 2.63631118781373,
+      "learning_rate": 9.014048489108106e-06,
+      "loss": 0.4505,
+      "step": 7130
+    },
+    {
+      "epoch": 1.0912012241775058,
+      "grad_norm": 2.1987115618847795,
+      "learning_rate": 9.011582410987538e-06,
+      "loss": 0.4639,
+      "step": 7131
+    },
+    {
+      "epoch": 1.091354246365723,
+      "grad_norm": 2.134264485035292,
+      "learning_rate": 9.009116393569563e-06,
+      "loss": 0.4645,
+      "step": 7132
+    },
+    {
+      "epoch": 1.0915072685539404,
+      "grad_norm": 2.0750373860179785,
+      "learning_rate": 9.006650437005627e-06,
+      "loss": 0.4651,
+      "step": 7133
+    },
+    {
+      "epoch": 1.0916602907421575,
+      "grad_norm": 2.2855081732746036,
+      "learning_rate": 9.004184541447169e-06,
+      "loss": 0.4995,
+      "step": 7134
+    },
+    {
+      "epoch": 1.0918133129303749,
+      "grad_norm": 2.4098402902812914,
+      "learning_rate": 9.00171870704564e-06,
+      "loss": 0.4734,
+      "step": 7135
+    },
+    {
+      "epoch": 1.0919663351185922,
+      "grad_norm": 2.032376714257287,
+      "learning_rate": 8.999252933952465e-06,
+      "loss": 0.3866,
+      "step": 7136
+    },
+    {
+      "epoch": 1.0921193573068095,
+      "grad_norm": 2.15493218358809,
+      "learning_rate": 8.996787222319087e-06,
+      "loss": 0.5154,
+      "step": 7137
+    },
+    {
+      "epoch": 1.0922723794950269,
+      "grad_norm": 2.0858765860099173,
+      "learning_rate": 8.99432157229693e-06,
+      "loss": 0.3879,
+      "step": 7138
+    },
+    {
+      "epoch": 1.092425401683244,
+      "grad_norm": 2.5591535079462098,
+      "learning_rate": 8.991855984037414e-06,
+      "loss": 0.4634,
+      "step": 7139
+    },
+    {
+      "epoch": 1.0925784238714613,
+      "grad_norm": 2.4454327491285417,
+      "learning_rate": 8.989390457691973e-06,
+      "loss": 0.4858,
+      "step": 7140
+    },
+    {
+      "epoch": 1.0927314460596786,
+      "grad_norm": 1.8496428093473942,
+      "learning_rate": 8.986924993412012e-06,
+      "loss": 0.4182,
+      "step": 7141
+    },
+    {
+      "epoch": 1.092884468247896,
+      "grad_norm": 1.9629732376566862,
+      "learning_rate": 8.984459591348952e-06,
+      "loss": 0.414,
+      "step": 7142
+    },
+    {
+      "epoch": 1.0930374904361133,
+      "grad_norm": 2.382561085581707,
+      "learning_rate": 8.981994251654205e-06,
+      "loss": 0.4709,
+      "step": 7143
+    },
+    {
+      "epoch": 1.0931905126243304,
+      "grad_norm": 2.076307911023831,
+      "learning_rate": 8.979528974479169e-06,
+      "loss": 0.4202,
+      "step": 7144
+    },
+    {
+      "epoch": 1.0933435348125478,
+      "grad_norm": 2.1286263317947687,
+      "learning_rate": 8.977063759975251e-06,
+      "loss": 0.4613,
+      "step": 7145
+    },
+    {
+      "epoch": 1.093496557000765,
+      "grad_norm": 2.4719234461371986,
+      "learning_rate": 8.974598608293853e-06,
+      "loss": 0.4872,
+      "step": 7146
+    },
+    {
+      "epoch": 1.0936495791889824,
+      "grad_norm": 1.9484181799070321,
+      "learning_rate": 8.972133519586361e-06,
+      "loss": 0.349,
+      "step": 7147
+    },
+    {
+      "epoch": 1.0938026013771998,
+      "grad_norm": 2.149297644516257,
+      "learning_rate": 8.969668494004173e-06,
+      "loss": 0.4462,
+      "step": 7148
+    },
+    {
+      "epoch": 1.0939556235654169,
+      "grad_norm": 2.3154019028958643,
+      "learning_rate": 8.967203531698672e-06,
+      "loss": 0.4999,
+      "step": 7149
+    },
+    {
+      "epoch": 1.0941086457536342,
+      "grad_norm": 2.345475674509572,
+      "learning_rate": 8.964738632821245e-06,
+      "loss": 0.4747,
+      "step": 7150
+    },
+    {
+      "epoch": 1.0942616679418515,
+      "grad_norm": 2.028434119282612,
+      "learning_rate": 8.962273797523268e-06,
+      "loss": 0.3636,
+      "step": 7151
+    },
+    {
+      "epoch": 1.0944146901300689,
+      "grad_norm": 2.5352442491743523,
+      "learning_rate": 8.959809025956113e-06,
+      "loss": 0.4937,
+      "step": 7152
+    },
+    {
+      "epoch": 1.0945677123182862,
+      "grad_norm": 2.0639018071881408,
+      "learning_rate": 8.957344318271161e-06,
+      "loss": 0.3728,
+      "step": 7153
+    },
+    {
+      "epoch": 1.0947207345065035,
+      "grad_norm": 2.3340432211508357,
+      "learning_rate": 8.954879674619769e-06,
+      "loss": 0.4858,
+      "step": 7154
+    },
+    {
+      "epoch": 1.0948737566947206,
+      "grad_norm": 2.0559018209376685,
+      "learning_rate": 8.952415095153305e-06,
+      "loss": 0.4086,
+      "step": 7155
+    },
+    {
+      "epoch": 1.095026778882938,
+      "grad_norm": 2.316442921819214,
+      "learning_rate": 8.949950580023131e-06,
+      "loss": 0.4435,
+      "step": 7156
+    },
+    {
+      "epoch": 1.0951798010711553,
+      "grad_norm": 1.9374221563323648,
+      "learning_rate": 8.947486129380597e-06,
+      "loss": 0.4009,
+      "step": 7157
+    },
+    {
+      "epoch": 1.0953328232593726,
+      "grad_norm": 2.030063875991149,
+      "learning_rate": 8.945021743377057e-06,
+      "loss": 0.3603,
+      "step": 7158
+    },
+    {
+      "epoch": 1.09548584544759,
+      "grad_norm": 2.1826213973706667,
+      "learning_rate": 8.942557422163864e-06,
+      "loss": 0.4447,
+      "step": 7159
+    },
+    {
+      "epoch": 1.095638867635807,
+      "grad_norm": 2.252983268260625,
+      "learning_rate": 8.94009316589235e-06,
+      "loss": 0.4963,
+      "step": 7160
+    },
+    {
+      "epoch": 1.0957918898240244,
+      "grad_norm": 2.6669262080305516,
+      "learning_rate": 8.937628974713864e-06,
+      "loss": 0.5089,
+      "step": 7161
+    },
+    {
+      "epoch": 1.0959449120122418,
+      "grad_norm": 2.12773257212889,
+      "learning_rate": 8.93516484877974e-06,
+      "loss": 0.4969,
+      "step": 7162
+    },
+    {
+      "epoch": 1.096097934200459,
+      "grad_norm": 2.3072830017793313,
+      "learning_rate": 8.932700788241304e-06,
+      "loss": 0.479,
+      "step": 7163
+    },
+    {
+      "epoch": 1.0962509563886764,
+      "grad_norm": 2.2042128737534004,
+      "learning_rate": 8.93023679324989e-06,
+      "loss": 0.4839,
+      "step": 7164
+    },
+    {
+      "epoch": 1.0964039785768938,
+      "grad_norm": 2.793680246758663,
+      "learning_rate": 8.92777286395682e-06,
+      "loss": 0.4981,
+      "step": 7165
+    },
+    {
+      "epoch": 1.0965570007651109,
+      "grad_norm": 2.225104256306868,
+      "learning_rate": 8.925309000513414e-06,
+      "loss": 0.4377,
+      "step": 7166
+    },
+    {
+      "epoch": 1.0967100229533282,
+      "grad_norm": 1.9745454245625953,
+      "learning_rate": 8.922845203070984e-06,
+      "loss": 0.4017,
+      "step": 7167
+    },
+    {
+      "epoch": 1.0968630451415455,
+      "grad_norm": 2.368420047885716,
+      "learning_rate": 8.920381471780843e-06,
+      "loss": 0.5215,
+      "step": 7168
+    },
+    {
+      "epoch": 1.0970160673297629,
+      "grad_norm": 2.051795870663101,
+      "learning_rate": 8.917917806794305e-06,
+      "loss": 0.3828,
+      "step": 7169
+    },
+    {
+      "epoch": 1.0971690895179802,
+      "grad_norm": 2.123677254807295,
+      "learning_rate": 8.915454208262664e-06,
+      "loss": 0.4633,
+      "step": 7170
+    },
+    {
+      "epoch": 1.0973221117061973,
+      "grad_norm": 2.2627608124003604,
+      "learning_rate": 8.912990676337221e-06,
+      "loss": 0.4726,
+      "step": 7171
+    },
+    {
+      "epoch": 1.0974751338944146,
+      "grad_norm": 2.103757846488755,
+      "learning_rate": 8.910527211169277e-06,
+      "loss": 0.4227,
+      "step": 7172
+    },
+    {
+      "epoch": 1.097628156082632,
+      "grad_norm": 2.2143920295035544,
+      "learning_rate": 8.908063812910116e-06,
+      "loss": 0.4625,
+      "step": 7173
+    },
+    {
+      "epoch": 1.0977811782708493,
+      "grad_norm": 2.100586428663275,
+      "learning_rate": 8.905600481711027e-06,
+      "loss": 0.4364,
+      "step": 7174
+    },
+    {
+      "epoch": 1.0979342004590666,
+      "grad_norm": 2.058446058812767,
+      "learning_rate": 8.9031372177233e-06,
+      "loss": 0.4965,
+      "step": 7175
+    },
+    {
+      "epoch": 1.0980872226472838,
+      "grad_norm": 2.588950853621863,
+      "learning_rate": 8.9006740210982e-06,
+      "loss": 0.3974,
+      "step": 7176
+    },
+    {
+      "epoch": 1.098240244835501,
+      "grad_norm": 2.1324434817566065,
+      "learning_rate": 8.898210891987012e-06,
+      "loss": 0.5475,
+      "step": 7177
+    },
+    {
+      "epoch": 1.0983932670237184,
+      "grad_norm": 2.0296371639639674,
+      "learning_rate": 8.895747830541001e-06,
+      "loss": 0.4481,
+      "step": 7178
+    },
+    {
+      "epoch": 1.0985462892119358,
+      "grad_norm": 2.0897999254760973,
+      "learning_rate": 8.89328483691144e-06,
+      "loss": 0.4616,
+      "step": 7179
+    },
+    {
+      "epoch": 1.098699311400153,
+      "grad_norm": 2.14171265984956,
+      "learning_rate": 8.890821911249585e-06,
+      "loss": 0.4869,
+      "step": 7180
+    },
+    {
+      "epoch": 1.0988523335883702,
+      "grad_norm": 2.1623838358382805,
+      "learning_rate": 8.888359053706694e-06,
+      "loss": 0.4638,
+      "step": 7181
+    },
+    {
+      "epoch": 1.0990053557765875,
+      "grad_norm": 1.956343477779932,
+      "learning_rate": 8.885896264434025e-06,
+      "loss": 0.3838,
+      "step": 7182
+    },
+    {
+      "epoch": 1.0991583779648049,
+      "grad_norm": 2.3400499233335768,
+      "learning_rate": 8.883433543582824e-06,
+      "loss": 0.4655,
+      "step": 7183
+    },
+    {
+      "epoch": 1.0993114001530222,
+      "grad_norm": 2.399181740585056,
+      "learning_rate": 8.880970891304334e-06,
+      "loss": 0.4578,
+      "step": 7184
+    },
+    {
+      "epoch": 1.0994644223412395,
+      "grad_norm": 2.250008234169996,
+      "learning_rate": 8.878508307749804e-06,
+      "loss": 0.428,
+      "step": 7185
+    },
+    {
+      "epoch": 1.0996174445294569,
+      "grad_norm": 2.290696083811117,
+      "learning_rate": 8.87604579307046e-06,
+      "loss": 0.4269,
+      "step": 7186
+    },
+    {
+      "epoch": 1.099770466717674,
+      "grad_norm": 2.505122870008909,
+      "learning_rate": 8.873583347417546e-06,
+      "loss": 0.5127,
+      "step": 7187
+    },
+    {
+      "epoch": 1.0999234889058913,
+      "grad_norm": 2.353030872555066,
+      "learning_rate": 8.871120970942285e-06,
+      "loss": 0.4578,
+      "step": 7188
+    },
+    {
+      "epoch": 1.1000765110941086,
+      "grad_norm": 2.301167013530359,
+      "learning_rate": 8.868658663795898e-06,
+      "loss": 0.3812,
+      "step": 7189
+    },
+    {
+      "epoch": 1.100229533282326,
+      "grad_norm": 2.2533182412177513,
+      "learning_rate": 8.866196426129612e-06,
+      "loss": 0.427,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1003825554705433,
+      "grad_norm": 2.4637131571486406,
+      "learning_rate": 8.863734258094638e-06,
+      "loss": 0.4376,
+      "step": 7191
+    },
+    {
+      "epoch": 1.1005355776587604,
+      "grad_norm": 2.0197232101443223,
+      "learning_rate": 8.861272159842186e-06,
+      "loss": 0.4344,
+      "step": 7192
+    },
+    {
+      "epoch": 1.1006885998469778,
+      "grad_norm": 2.251483371734288,
+      "learning_rate": 8.858810131523467e-06,
+      "loss": 0.4298,
+      "step": 7193
+    },
+    {
+      "epoch": 1.100841622035195,
+      "grad_norm": 2.287008056541385,
+      "learning_rate": 8.85634817328968e-06,
+      "loss": 0.4718,
+      "step": 7194
+    },
+    {
+      "epoch": 1.1009946442234124,
+      "grad_norm": 2.209324158741729,
+      "learning_rate": 8.85388628529203e-06,
+      "loss": 0.4657,
+      "step": 7195
+    },
+    {
+      "epoch": 1.1011476664116298,
+      "grad_norm": 2.1505467139385304,
+      "learning_rate": 8.851424467681705e-06,
+      "loss": 0.4322,
+      "step": 7196
+    },
+    {
+      "epoch": 1.101300688599847,
+      "grad_norm": 2.1982904432190984,
+      "learning_rate": 8.848962720609894e-06,
+      "loss": 0.4675,
+      "step": 7197
+    },
+    {
+      "epoch": 1.1014537107880642,
+      "grad_norm": 2.1486253183804767,
+      "learning_rate": 8.84650104422779e-06,
+      "loss": 0.4145,
+      "step": 7198
+    },
+    {
+      "epoch": 1.1016067329762815,
+      "grad_norm": 2.0660350104739864,
+      "learning_rate": 8.844039438686566e-06,
+      "loss": 0.4289,
+      "step": 7199
+    },
+    {
+      "epoch": 1.1017597551644989,
+      "grad_norm": 2.2958785347357793,
+      "learning_rate": 8.841577904137402e-06,
+      "loss": 0.4849,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1019127773527162,
+      "grad_norm": 2.3249009204416216,
+      "learning_rate": 8.839116440731475e-06,
+      "loss": 0.4445,
+      "step": 7201
+    },
+    {
+      "epoch": 1.1020657995409335,
+      "grad_norm": 2.2802870963931188,
+      "learning_rate": 8.836655048619943e-06,
+      "loss": 0.4219,
+      "step": 7202
+    },
+    {
+      "epoch": 1.1022188217291506,
+      "grad_norm": 2.1524504956749304,
+      "learning_rate": 8.834193727953977e-06,
+      "loss": 0.4425,
+      "step": 7203
+    },
+    {
+      "epoch": 1.102371843917368,
+      "grad_norm": 2.3059346698587024,
+      "learning_rate": 8.831732478884739e-06,
+      "loss": 0.481,
+      "step": 7204
+    },
+    {
+      "epoch": 1.1025248661055853,
+      "grad_norm": 2.017182531907496,
+      "learning_rate": 8.829271301563375e-06,
+      "loss": 0.3548,
+      "step": 7205
+    },
+    {
+      "epoch": 1.1026778882938026,
+      "grad_norm": 2.1409796994991286,
+      "learning_rate": 8.826810196141042e-06,
+      "loss": 0.5375,
+      "step": 7206
+    },
+    {
+      "epoch": 1.10283091048202,
+      "grad_norm": 2.1409288991943924,
+      "learning_rate": 8.824349162768882e-06,
+      "loss": 0.4341,
+      "step": 7207
+    },
+    {
+      "epoch": 1.102983932670237,
+      "grad_norm": 2.198060299539323,
+      "learning_rate": 8.821888201598044e-06,
+      "loss": 0.4512,
+      "step": 7208
+    },
+    {
+      "epoch": 1.1031369548584544,
+      "grad_norm": 2.262625789778196,
+      "learning_rate": 8.819427312779658e-06,
+      "loss": 0.4792,
+      "step": 7209
+    },
+    {
+      "epoch": 1.1032899770466718,
+      "grad_norm": 2.2566093525445594,
+      "learning_rate": 8.816966496464858e-06,
+      "loss": 0.4995,
+      "step": 7210
+    },
+    {
+      "epoch": 1.103442999234889,
+      "grad_norm": 2.4283834426361994,
+      "learning_rate": 8.814505752804776e-06,
+      "loss": 0.4806,
+      "step": 7211
+    },
+    {
+      "epoch": 1.1035960214231064,
+      "grad_norm": 2.170904697292499,
+      "learning_rate": 8.812045081950532e-06,
+      "loss": 0.4919,
+      "step": 7212
+    },
+    {
+      "epoch": 1.1037490436113235,
+      "grad_norm": 1.9863702432643318,
+      "learning_rate": 8.809584484053245e-06,
+      "loss": 0.375,
+      "step": 7213
+    },
+    {
+      "epoch": 1.1039020657995409,
+      "grad_norm": 2.5916602921862286,
+      "learning_rate": 8.807123959264036e-06,
+      "loss": 0.4797,
+      "step": 7214
+    },
+    {
+      "epoch": 1.1040550879877582,
+      "grad_norm": 2.434544800191652,
+      "learning_rate": 8.804663507734004e-06,
+      "loss": 0.3995,
+      "step": 7215
+    },
+    {
+      "epoch": 1.1042081101759755,
+      "grad_norm": 2.499246869230129,
+      "learning_rate": 8.802203129614269e-06,
+      "loss": 0.5353,
+      "step": 7216
+    },
+    {
+      "epoch": 1.1043611323641929,
+      "grad_norm": 2.0781080712876117,
+      "learning_rate": 8.799742825055927e-06,
+      "loss": 0.38,
+      "step": 7217
+    },
+    {
+      "epoch": 1.1045141545524102,
+      "grad_norm": 2.394442587470399,
+      "learning_rate": 8.797282594210066e-06,
+      "loss": 0.4759,
+      "step": 7218
+    },
+    {
+      "epoch": 1.1046671767406273,
+      "grad_norm": 2.51859505988219,
+      "learning_rate": 8.79482243722779e-06,
+      "loss": 0.4989,
+      "step": 7219
+    },
+    {
+      "epoch": 1.1048201989288446,
+      "grad_norm": 2.075859944947023,
+      "learning_rate": 8.792362354260185e-06,
+      "loss": 0.4434,
+      "step": 7220
+    },
+    {
+      "epoch": 1.104973221117062,
+      "grad_norm": 2.4451130046615592,
+      "learning_rate": 8.789902345458326e-06,
+      "loss": 0.4663,
+      "step": 7221
+    },
+    {
+      "epoch": 1.1051262433052793,
+      "grad_norm": 2.3113134861788036,
+      "learning_rate": 8.787442410973305e-06,
+      "loss": 0.4555,
+      "step": 7222
+    },
+    {
+      "epoch": 1.1052792654934966,
+      "grad_norm": 2.1639552175738856,
+      "learning_rate": 8.784982550956184e-06,
+      "loss": 0.424,
+      "step": 7223
+    },
+    {
+      "epoch": 1.1054322876817138,
+      "grad_norm": 1.9803878884912123,
+      "learning_rate": 8.782522765558039e-06,
+      "loss": 0.4537,
+      "step": 7224
+    },
+    {
+      "epoch": 1.105585309869931,
+      "grad_norm": 2.2787793998286823,
+      "learning_rate": 8.780063054929937e-06,
+      "loss": 0.4371,
+      "step": 7225
+    },
+    {
+      "epoch": 1.1057383320581484,
+      "grad_norm": 2.3573504102686713,
+      "learning_rate": 8.77760341922293e-06,
+      "loss": 0.5008,
+      "step": 7226
+    },
+    {
+      "epoch": 1.1058913542463658,
+      "grad_norm": 2.1996380165093177,
+      "learning_rate": 8.775143858588082e-06,
+      "loss": 0.4169,
+      "step": 7227
+    },
+    {
+      "epoch": 1.106044376434583,
+      "grad_norm": 2.616789474953216,
+      "learning_rate": 8.772684373176443e-06,
+      "loss": 0.4648,
+      "step": 7228
+    },
+    {
+      "epoch": 1.1061973986228004,
+      "grad_norm": 2.128350016291785,
+      "learning_rate": 8.770224963139054e-06,
+      "loss": 0.4217,
+      "step": 7229
+    },
+    {
+      "epoch": 1.1063504208110175,
+      "grad_norm": 2.579626546907676,
+      "learning_rate": 8.767765628626962e-06,
+      "loss": 0.5056,
+      "step": 7230
+    },
+    {
+      "epoch": 1.1065034429992349,
+      "grad_norm": 2.2513999128668636,
+      "learning_rate": 8.765306369791204e-06,
+      "loss": 0.4425,
+      "step": 7231
+    },
+    {
+      "epoch": 1.1066564651874522,
+      "grad_norm": 2.442135717799566,
+      "learning_rate": 8.762847186782813e-06,
+      "loss": 0.506,
+      "step": 7232
+    },
+    {
+      "epoch": 1.1068094873756695,
+      "grad_norm": 2.090965212912735,
+      "learning_rate": 8.760388079752815e-06,
+      "loss": 0.4475,
+      "step": 7233
+    },
+    {
+      "epoch": 1.1069625095638869,
+      "grad_norm": 2.2371641896853802,
+      "learning_rate": 8.757929048852233e-06,
+      "loss": 0.484,
+      "step": 7234
+    },
+    {
+      "epoch": 1.107115531752104,
+      "grad_norm": 2.3009833508605007,
+      "learning_rate": 8.75547009423209e-06,
+      "loss": 0.3488,
+      "step": 7235
+    },
+    {
+      "epoch": 1.1072685539403213,
+      "grad_norm": 2.1163473200298957,
+      "learning_rate": 8.753011216043396e-06,
+      "loss": 0.3902,
+      "step": 7236
+    },
+    {
+      "epoch": 1.1074215761285386,
+      "grad_norm": 2.3626068037558685,
+      "learning_rate": 8.75055241443716e-06,
+      "loss": 0.4614,
+      "step": 7237
+    },
+    {
+      "epoch": 1.107574598316756,
+      "grad_norm": 1.8373619367737042,
+      "learning_rate": 8.748093689564392e-06,
+      "loss": 0.3726,
+      "step": 7238
+    },
+    {
+      "epoch": 1.1077276205049733,
+      "grad_norm": 1.9346840039209652,
+      "learning_rate": 8.745635041576082e-06,
+      "loss": 0.3809,
+      "step": 7239
+    },
+    {
+      "epoch": 1.1078806426931904,
+      "grad_norm": 2.1799202758607312,
+      "learning_rate": 8.743176470623236e-06,
+      "loss": 0.4697,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1080336648814078,
+      "grad_norm": 2.351668788150053,
+      "learning_rate": 8.74071797685684e-06,
+      "loss": 0.3959,
+      "step": 7241
+    },
+    {
+      "epoch": 1.108186687069625,
+      "grad_norm": 2.0925064474546686,
+      "learning_rate": 8.738259560427876e-06,
+      "loss": 0.41,
+      "step": 7242
+    },
+    {
+      "epoch": 1.1083397092578424,
+      "grad_norm": 2.260684681391703,
+      "learning_rate": 8.73580122148733e-06,
+      "loss": 0.4818,
+      "step": 7243
+    },
+    {
+      "epoch": 1.1084927314460598,
+      "grad_norm": 2.1853032894152995,
+      "learning_rate": 8.733342960186174e-06,
+      "loss": 0.4506,
+      "step": 7244
+    },
+    {
+      "epoch": 1.1086457536342769,
+      "grad_norm": 1.9906423609607566,
+      "learning_rate": 8.730884776675386e-06,
+      "loss": 0.3836,
+      "step": 7245
+    },
+    {
+      "epoch": 1.1087987758224942,
+      "grad_norm": 1.9829231974861674,
+      "learning_rate": 8.728426671105929e-06,
+      "loss": 0.497,
+      "step": 7246
+    },
+    {
+      "epoch": 1.1089517980107115,
+      "grad_norm": 2.0781493892344316,
+      "learning_rate": 8.72596864362876e-06,
+      "loss": 0.4439,
+      "step": 7247
+    },
+    {
+      "epoch": 1.1091048201989289,
+      "grad_norm": 2.380914255234149,
+      "learning_rate": 8.723510694394845e-06,
+      "loss": 0.4626,
+      "step": 7248
+    },
+    {
+      "epoch": 1.1092578423871462,
+      "grad_norm": 2.430407170746528,
+      "learning_rate": 8.721052823555131e-06,
+      "loss": 0.4854,
+      "step": 7249
+    },
+    {
+      "epoch": 1.1094108645753635,
+      "grad_norm": 2.4319507114869667,
+      "learning_rate": 8.718595031260564e-06,
+      "loss": 0.3742,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1095638867635806,
+      "grad_norm": 2.093265509251606,
+      "learning_rate": 8.716137317662093e-06,
+      "loss": 0.388,
+      "step": 7251
+    },
+    {
+      "epoch": 1.109716908951798,
+      "grad_norm": 2.4559992923329106,
+      "learning_rate": 8.713679682910648e-06,
+      "loss": 0.4244,
+      "step": 7252
+    },
+    {
+      "epoch": 1.1098699311400153,
+      "grad_norm": 2.1829029164513982,
+      "learning_rate": 8.711222127157166e-06,
+      "loss": 0.3357,
+      "step": 7253
+    },
+    {
+      "epoch": 1.1100229533282326,
+      "grad_norm": 2.3403552031528903,
+      "learning_rate": 8.70876465055258e-06,
+      "loss": 0.4064,
+      "step": 7254
+    },
+    {
+      "epoch": 1.11017597551645,
+      "grad_norm": 1.989255895084661,
+      "learning_rate": 8.7063072532478e-06,
+      "loss": 0.3819,
+      "step": 7255
+    },
+    {
+      "epoch": 1.110328997704667,
+      "grad_norm": 2.184660067497534,
+      "learning_rate": 8.703849935393758e-06,
+      "loss": 0.4368,
+      "step": 7256
+    },
+    {
+      "epoch": 1.1104820198928844,
+      "grad_norm": 2.349449352310267,
+      "learning_rate": 8.701392697141363e-06,
+      "loss": 0.4845,
+      "step": 7257
+    },
+    {
+      "epoch": 1.1106350420811018,
+      "grad_norm": 2.198702813821092,
+      "learning_rate": 8.698935538641519e-06,
+      "loss": 0.4256,
+      "step": 7258
+    },
+    {
+      "epoch": 1.110788064269319,
+      "grad_norm": 2.342120943064589,
+      "learning_rate": 8.696478460045134e-06,
+      "loss": 0.451,
+      "step": 7259
+    },
+    {
+      "epoch": 1.1109410864575364,
+      "grad_norm": 2.187839864774295,
+      "learning_rate": 8.694021461503105e-06,
+      "loss": 0.4097,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1110941086457535,
+      "grad_norm": 2.165677195044475,
+      "learning_rate": 8.691564543166332e-06,
+      "loss": 0.4467,
+      "step": 7261
+    },
+    {
+      "epoch": 1.1112471308339709,
+      "grad_norm": 2.278960948831545,
+      "learning_rate": 8.689107705185697e-06,
+      "loss": 0.4216,
+      "step": 7262
+    },
+    {
+      "epoch": 1.1114001530221882,
+      "grad_norm": 2.1925602427988085,
+      "learning_rate": 8.686650947712084e-06,
+      "loss": 0.4818,
+      "step": 7263
+    },
+    {
+      "epoch": 1.1115531752104055,
+      "grad_norm": 2.4280978672138365,
+      "learning_rate": 8.684194270896376e-06,
+      "loss": 0.4793,
+      "step": 7264
+    },
+    {
+      "epoch": 1.1117061973986229,
+      "grad_norm": 2.413291630851117,
+      "learning_rate": 8.681737674889445e-06,
+      "loss": 0.436,
+      "step": 7265
+    },
+    {
+      "epoch": 1.1118592195868402,
+      "grad_norm": 2.182619190908532,
+      "learning_rate": 8.679281159842157e-06,
+      "loss": 0.4091,
+      "step": 7266
+    },
+    {
+      "epoch": 1.1120122417750573,
+      "grad_norm": 2.352484482668082,
+      "learning_rate": 8.676824725905386e-06,
+      "loss": 0.4431,
+      "step": 7267
+    },
+    {
+      "epoch": 1.1121652639632746,
+      "grad_norm": 2.104077049629743,
+      "learning_rate": 8.67436837322998e-06,
+      "loss": 0.4664,
+      "step": 7268
+    },
+    {
+      "epoch": 1.112318286151492,
+      "grad_norm": 1.9325614114153729,
+      "learning_rate": 8.671912101966799e-06,
+      "loss": 0.4152,
+      "step": 7269
+    },
+    {
+      "epoch": 1.1124713083397093,
+      "grad_norm": 2.2218232070237933,
+      "learning_rate": 8.669455912266696e-06,
+      "loss": 0.4681,
+      "step": 7270
+    },
+    {
+      "epoch": 1.1126243305279266,
+      "grad_norm": 2.3511511595238925,
+      "learning_rate": 8.666999804280503e-06,
+      "loss": 0.4319,
+      "step": 7271
+    },
+    {
+      "epoch": 1.1127773527161438,
+      "grad_norm": 2.5154012800065515,
+      "learning_rate": 8.664543778159071e-06,
+      "loss": 0.5433,
+      "step": 7272
+    },
+    {
+      "epoch": 1.112930374904361,
+      "grad_norm": 2.094015164774007,
+      "learning_rate": 8.662087834053232e-06,
+      "loss": 0.4057,
+      "step": 7273
+    },
+    {
+      "epoch": 1.1130833970925784,
+      "grad_norm": 2.407236738502629,
+      "learning_rate": 8.659631972113809e-06,
+      "loss": 0.4519,
+      "step": 7274
+    },
+    {
+      "epoch": 1.1132364192807958,
+      "grad_norm": 2.3070387985152503,
+      "learning_rate": 8.65717619249163e-06,
+      "loss": 0.4029,
+      "step": 7275
+    },
+    {
+      "epoch": 1.113389441469013,
+      "grad_norm": 2.2050815747283057,
+      "learning_rate": 8.654720495337514e-06,
+      "loss": 0.4293,
+      "step": 7276
+    },
+    {
+      "epoch": 1.1135424636572302,
+      "grad_norm": 1.9874516443915964,
+      "learning_rate": 8.652264880802279e-06,
+      "loss": 0.3515,
+      "step": 7277
+    },
+    {
+      "epoch": 1.1136954858454475,
+      "grad_norm": 2.283468114473653,
+      "learning_rate": 8.649809349036729e-06,
+      "loss": 0.5043,
+      "step": 7278
+    },
+    {
+      "epoch": 1.1138485080336649,
+      "grad_norm": 2.228147809122041,
+      "learning_rate": 8.647353900191665e-06,
+      "loss": 0.4794,
+      "step": 7279
+    },
+    {
+      "epoch": 1.1140015302218822,
+      "grad_norm": 2.181698558558112,
+      "learning_rate": 8.644898534417893e-06,
+      "loss": 0.4362,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1141545524100995,
+      "grad_norm": 2.7327555296956514,
+      "learning_rate": 8.642443251866202e-06,
+      "loss": 0.5515,
+      "step": 7281
+    },
+    {
+      "epoch": 1.1143075745983166,
+      "grad_norm": 2.172183101436314,
+      "learning_rate": 8.639988052687383e-06,
+      "loss": 0.5035,
+      "step": 7282
+    },
+    {
+      "epoch": 1.114460596786534,
+      "grad_norm": 2.3012502821953564,
+      "learning_rate": 8.637532937032218e-06,
+      "loss": 0.42,
+      "step": 7283
+    },
+    {
+      "epoch": 1.1146136189747513,
+      "grad_norm": 2.3750682620962333,
+      "learning_rate": 8.635077905051484e-06,
+      "loss": 0.4683,
+      "step": 7284
+    },
+    {
+      "epoch": 1.1147666411629686,
+      "grad_norm": 2.295739800025103,
+      "learning_rate": 8.632622956895956e-06,
+      "loss": 0.4134,
+      "step": 7285
+    },
+    {
+      "epoch": 1.114919663351186,
+      "grad_norm": 2.5572348746935862,
+      "learning_rate": 8.630168092716403e-06,
+      "loss": 0.4558,
+      "step": 7286
+    },
+    {
+      "epoch": 1.1150726855394033,
+      "grad_norm": 2.2569408220861042,
+      "learning_rate": 8.627713312663582e-06,
+      "loss": 0.4033,
+      "step": 7287
+    },
+    {
+      "epoch": 1.1152257077276204,
+      "grad_norm": 2.270829840814031,
+      "learning_rate": 8.625258616888258e-06,
+      "loss": 0.5059,
+      "step": 7288
+    },
+    {
+      "epoch": 1.1153787299158378,
+      "grad_norm": 2.402699754110219,
+      "learning_rate": 8.622804005541179e-06,
+      "loss": 0.462,
+      "step": 7289
+    },
+    {
+      "epoch": 1.115531752104055,
+      "grad_norm": 2.4722485610391782,
+      "learning_rate": 8.620349478773095e-06,
+      "loss": 0.4529,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1156847742922724,
+      "grad_norm": 2.1893417797352344,
+      "learning_rate": 8.617895036734747e-06,
+      "loss": 0.4421,
+      "step": 7291
+    },
+    {
+      "epoch": 1.1158377964804898,
+      "grad_norm": 2.1580635465283633,
+      "learning_rate": 8.61544067957687e-06,
+      "loss": 0.3836,
+      "step": 7292
+    },
+    {
+      "epoch": 1.1159908186687069,
+      "grad_norm": 2.4058476302646374,
+      "learning_rate": 8.6129864074502e-06,
+      "loss": 0.4373,
+      "step": 7293
+    },
+    {
+      "epoch": 1.1161438408569242,
+      "grad_norm": 2.3222777156880037,
+      "learning_rate": 8.610532220505459e-06,
+      "loss": 0.5089,
+      "step": 7294
+    },
+    {
+      "epoch": 1.1162968630451415,
+      "grad_norm": 2.1307415290698852,
+      "learning_rate": 8.608078118893368e-06,
+      "loss": 0.5199,
+      "step": 7295
+    },
+    {
+      "epoch": 1.1164498852333589,
+      "grad_norm": 2.383794896617956,
+      "learning_rate": 8.60562410276465e-06,
+      "loss": 0.4593,
+      "step": 7296
+    },
+    {
+      "epoch": 1.1166029074215762,
+      "grad_norm": 2.5837278205838663,
+      "learning_rate": 8.603170172270002e-06,
+      "loss": 0.4638,
+      "step": 7297
+    },
+    {
+      "epoch": 1.1167559296097935,
+      "grad_norm": 2.2525760948688944,
+      "learning_rate": 8.600716327560146e-06,
+      "loss": 0.4474,
+      "step": 7298
+    },
+    {
+      "epoch": 1.1169089517980106,
+      "grad_norm": 2.615571933223899,
+      "learning_rate": 8.598262568785775e-06,
+      "loss": 0.4843,
+      "step": 7299
+    },
+    {
+      "epoch": 1.117061973986228,
+      "grad_norm": 2.439123207803759,
+      "learning_rate": 8.595808896097575e-06,
+      "loss": 0.4902,
+      "step": 7300
+    },
+    {
+      "epoch": 1.1172149961744453,
+      "grad_norm": 2.2731590292985677,
+      "learning_rate": 8.593355309646252e-06,
+      "loss": 0.514,
+      "step": 7301
+    },
+    {
+      "epoch": 1.1173680183626626,
+      "grad_norm": 2.2254431206595773,
+      "learning_rate": 8.590901809582482e-06,
+      "loss": 0.4406,
+      "step": 7302
+    },
+    {
+      "epoch": 1.11752104055088,
+      "grad_norm": 2.1556714339537226,
+      "learning_rate": 8.588448396056938e-06,
+      "loss": 0.4179,
+      "step": 7303
+    },
+    {
+      "epoch": 1.117674062739097,
+      "grad_norm": 2.2555805850430817,
+      "learning_rate": 8.585995069220305e-06,
+      "loss": 0.4607,
+      "step": 7304
+    },
+    {
+      "epoch": 1.1178270849273144,
+      "grad_norm": 2.095833604865627,
+      "learning_rate": 8.583541829223243e-06,
+      "loss": 0.5086,
+      "step": 7305
+    },
+    {
+      "epoch": 1.1179801071155318,
+      "grad_norm": 2.3270010835496397,
+      "learning_rate": 8.581088676216421e-06,
+      "loss": 0.4744,
+      "step": 7306
+    },
+    {
+      "epoch": 1.118133129303749,
+      "grad_norm": 2.491796490650458,
+      "learning_rate": 8.578635610350497e-06,
+      "loss": 0.4065,
+      "step": 7307
+    },
+    {
+      "epoch": 1.1182861514919664,
+      "grad_norm": 2.182575815083912,
+      "learning_rate": 8.576182631776115e-06,
+      "loss": 0.4834,
+      "step": 7308
+    },
+    {
+      "epoch": 1.1184391736801835,
+      "grad_norm": 2.173635352835729,
+      "learning_rate": 8.57372974064393e-06,
+      "loss": 0.4341,
+      "step": 7309
+    },
+    {
+      "epoch": 1.1185921958684009,
+      "grad_norm": 2.3043141043629274,
+      "learning_rate": 8.571276937104584e-06,
+      "loss": 0.464,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1187452180566182,
+      "grad_norm": 2.316805699809123,
+      "learning_rate": 8.568824221308705e-06,
+      "loss": 0.4992,
+      "step": 7311
+    },
+    {
+      "epoch": 1.1188982402448355,
+      "grad_norm": 2.100553392834741,
+      "learning_rate": 8.566371593406933e-06,
+      "loss": 0.447,
+      "step": 7312
+    },
+    {
+      "epoch": 1.1190512624330529,
+      "grad_norm": 2.3347884127487055,
+      "learning_rate": 8.563919053549887e-06,
+      "loss": 0.4517,
+      "step": 7313
+    },
+    {
+      "epoch": 1.11920428462127,
+      "grad_norm": 2.011899264289727,
+      "learning_rate": 8.561466601888195e-06,
+      "loss": 0.381,
+      "step": 7314
+    },
+    {
+      "epoch": 1.1193573068094873,
+      "grad_norm": 2.264732519715413,
+      "learning_rate": 8.559014238572463e-06,
+      "loss": 0.4625,
+      "step": 7315
+    },
+    {
+      "epoch": 1.1195103289977046,
+      "grad_norm": 2.1982141461822944,
+      "learning_rate": 8.556561963753303e-06,
+      "loss": 0.433,
+      "step": 7316
+    },
+    {
+      "epoch": 1.119663351185922,
+      "grad_norm": 2.2287969140588038,
+      "learning_rate": 8.554109777581322e-06,
+      "loss": 0.4322,
+      "step": 7317
+    },
+    {
+      "epoch": 1.1198163733741393,
+      "grad_norm": 2.0805687752430218,
+      "learning_rate": 8.551657680207114e-06,
+      "loss": 0.4368,
+      "step": 7318
+    },
+    {
+      "epoch": 1.1199693955623566,
+      "grad_norm": 2.167889830590893,
+      "learning_rate": 8.549205671781273e-06,
+      "loss": 0.6545,
+      "step": 7319
+    },
+    {
+      "epoch": 1.1201224177505738,
+      "grad_norm": 2.528070985327336,
+      "learning_rate": 8.54675375245439e-06,
+      "loss": 0.5258,
+      "step": 7320
+    },
+    {
+      "epoch": 1.120275439938791,
+      "grad_norm": 1.787445525421544,
+      "learning_rate": 8.544301922377043e-06,
+      "loss": 0.3962,
+      "step": 7321
+    },
+    {
+      "epoch": 1.1204284621270084,
+      "grad_norm": 2.244017891063301,
+      "learning_rate": 8.54185018169981e-06,
+      "loss": 0.4742,
+      "step": 7322
+    },
+    {
+      "epoch": 1.1205814843152258,
+      "grad_norm": 2.249034278168932,
+      "learning_rate": 8.539398530573264e-06,
+      "loss": 0.4758,
+      "step": 7323
+    },
+    {
+      "epoch": 1.120734506503443,
+      "grad_norm": 2.0505223499372796,
+      "learning_rate": 8.536946969147963e-06,
+      "loss": 0.4034,
+      "step": 7324
+    },
+    {
+      "epoch": 1.1208875286916602,
+      "grad_norm": 2.433144003092916,
+      "learning_rate": 8.534495497574472e-06,
+      "loss": 0.5097,
+      "step": 7325
+    },
+    {
+      "epoch": 1.1210405508798775,
+      "grad_norm": 2.136664385459225,
+      "learning_rate": 8.532044116003347e-06,
+      "loss": 0.4433,
+      "step": 7326
+    },
+    {
+      "epoch": 1.1211935730680949,
+      "grad_norm": 2.360936102413662,
+      "learning_rate": 8.52959282458514e-06,
+      "loss": 0.4784,
+      "step": 7327
+    },
+    {
+      "epoch": 1.1213465952563122,
+      "grad_norm": 2.1996065493633106,
+      "learning_rate": 8.527141623470387e-06,
+      "loss": 0.4174,
+      "step": 7328
+    },
+    {
+      "epoch": 1.1214996174445295,
+      "grad_norm": 2.2465195864524463,
+      "learning_rate": 8.524690512809626e-06,
+      "loss": 0.3862,
+      "step": 7329
+    },
+    {
+      "epoch": 1.1216526396327469,
+      "grad_norm": 2.1053183068145045,
+      "learning_rate": 8.522239492753395e-06,
+      "loss": 0.475,
+      "step": 7330
+    },
+    {
+      "epoch": 1.121805661820964,
+      "grad_norm": 2.418720533759001,
+      "learning_rate": 8.519788563452218e-06,
+      "loss": 0.4415,
+      "step": 7331
+    },
+    {
+      "epoch": 1.1219586840091813,
+      "grad_norm": 1.9996187128183676,
+      "learning_rate": 8.517337725056613e-06,
+      "loss": 0.4324,
+      "step": 7332
+    },
+    {
+      "epoch": 1.1221117061973986,
+      "grad_norm": 2.1481582833504693,
+      "learning_rate": 8.514886977717102e-06,
+      "loss": 0.4025,
+      "step": 7333
+    },
+    {
+      "epoch": 1.122264728385616,
+      "grad_norm": 2.2852880745037676,
+      "learning_rate": 8.51243632158419e-06,
+      "loss": 0.5204,
+      "step": 7334
+    },
+    {
+      "epoch": 1.1224177505738333,
+      "grad_norm": 1.9810063469638823,
+      "learning_rate": 8.509985756808382e-06,
+      "loss": 0.417,
+      "step": 7335
+    },
+    {
+      "epoch": 1.1225707727620504,
+      "grad_norm": 2.354682978168584,
+      "learning_rate": 8.507535283540181e-06,
+      "loss": 0.4445,
+      "step": 7336
+    },
+    {
+      "epoch": 1.1227237949502678,
+      "grad_norm": 2.049284352078451,
+      "learning_rate": 8.505084901930073e-06,
+      "loss": 0.4622,
+      "step": 7337
+    },
+    {
+      "epoch": 1.122876817138485,
+      "grad_norm": 2.2985255632558106,
+      "learning_rate": 8.502634612128551e-06,
+      "loss": 0.4512,
+      "step": 7338
+    },
+    {
+      "epoch": 1.1230298393267024,
+      "grad_norm": 2.289298123164532,
+      "learning_rate": 8.500184414286096e-06,
+      "loss": 0.4481,
+      "step": 7339
+    },
+    {
+      "epoch": 1.1231828615149198,
+      "grad_norm": 2.066046084729003,
+      "learning_rate": 8.49773430855318e-06,
+      "loss": 0.5145,
+      "step": 7340
+    },
+    {
+      "epoch": 1.1233358837031369,
+      "grad_norm": 2.4768352381868155,
+      "learning_rate": 8.495284295080281e-06,
+      "loss": 0.507,
+      "step": 7341
+    },
+    {
+      "epoch": 1.1234889058913542,
+      "grad_norm": 2.379390037498325,
+      "learning_rate": 8.492834374017856e-06,
+      "loss": 0.4829,
+      "step": 7342
+    },
+    {
+      "epoch": 1.1236419280795715,
+      "grad_norm": 2.107590759872093,
+      "learning_rate": 8.490384545516373e-06,
+      "loss": 0.411,
+      "step": 7343
+    },
+    {
+      "epoch": 1.1237949502677889,
+      "grad_norm": 1.9761377140115508,
+      "learning_rate": 8.487934809726277e-06,
+      "loss": 0.4598,
+      "step": 7344
+    },
+    {
+      "epoch": 1.1239479724560062,
+      "grad_norm": 1.9653246388829446,
+      "learning_rate": 8.485485166798019e-06,
+      "loss": 0.4399,
+      "step": 7345
+    },
+    {
+      "epoch": 1.1241009946442233,
+      "grad_norm": 2.3005849600149917,
+      "learning_rate": 8.483035616882046e-06,
+      "loss": 0.4133,
+      "step": 7346
+    },
+    {
+      "epoch": 1.1242540168324406,
+      "grad_norm": 2.108212619768816,
+      "learning_rate": 8.480586160128785e-06,
+      "loss": 0.4129,
+      "step": 7347
+    },
+    {
+      "epoch": 1.124407039020658,
+      "grad_norm": 2.189179438990085,
+      "learning_rate": 8.478136796688674e-06,
+      "loss": 0.4236,
+      "step": 7348
+    },
+    {
+      "epoch": 1.1245600612088753,
+      "grad_norm": 2.0614034092253117,
+      "learning_rate": 8.47568752671214e-06,
+      "loss": 0.4482,
+      "step": 7349
+    },
+    {
+      "epoch": 1.1247130833970926,
+      "grad_norm": 2.3106721350735357,
+      "learning_rate": 8.473238350349593e-06,
+      "loss": 0.4582,
+      "step": 7350
+    },
+    {
+      "epoch": 1.12486610558531,
+      "grad_norm": 2.069049495107061,
+      "learning_rate": 8.470789267751452e-06,
+      "loss": 0.4362,
+      "step": 7351
+    },
+    {
+      "epoch": 1.125019127773527,
+      "grad_norm": 2.232653664335247,
+      "learning_rate": 8.468340279068128e-06,
+      "loss": 0.4467,
+      "step": 7352
+    },
+    {
+      "epoch": 1.1251721499617444,
+      "grad_norm": 2.2760784233432263,
+      "learning_rate": 8.465891384450015e-06,
+      "loss": 0.4328,
+      "step": 7353
+    },
+    {
+      "epoch": 1.1253251721499617,
+      "grad_norm": 1.9416601069747,
+      "learning_rate": 8.463442584047516e-06,
+      "loss": 0.3905,
+      "step": 7354
+    },
+    {
+      "epoch": 1.125478194338179,
+      "grad_norm": 2.023979021978067,
+      "learning_rate": 8.460993878011014e-06,
+      "loss": 0.4174,
+      "step": 7355
+    },
+    {
+      "epoch": 1.1256312165263964,
+      "grad_norm": 2.585206276544525,
+      "learning_rate": 8.458545266490906e-06,
+      "loss": 0.454,
+      "step": 7356
+    },
+    {
+      "epoch": 1.1257842387146135,
+      "grad_norm": 2.0345591716158298,
+      "learning_rate": 8.456096749637561e-06,
+      "loss": 0.3921,
+      "step": 7357
+    },
+    {
+      "epoch": 1.1259372609028309,
+      "grad_norm": 2.470455072554232,
+      "learning_rate": 8.453648327601352e-06,
+      "loss": 0.4929,
+      "step": 7358
+    },
+    {
+      "epoch": 1.1260902830910482,
+      "grad_norm": 2.3686162736588385,
+      "learning_rate": 8.451200000532653e-06,
+      "loss": 0.4353,
+      "step": 7359
+    },
+    {
+      "epoch": 1.1262433052792655,
+      "grad_norm": 2.135323906416676,
+      "learning_rate": 8.448751768581818e-06,
+      "loss": 0.3795,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1263963274674829,
+      "grad_norm": 2.33267928641562,
+      "learning_rate": 8.446303631899203e-06,
+      "loss": 0.4306,
+      "step": 7361
+    },
+    {
+      "epoch": 1.1265493496557002,
+      "grad_norm": 2.4218289282958114,
+      "learning_rate": 8.443855590635165e-06,
+      "loss": 0.451,
+      "step": 7362
+    },
+    {
+      "epoch": 1.1267023718439173,
+      "grad_norm": 2.132079052539921,
+      "learning_rate": 8.441407644940038e-06,
+      "loss": 0.4516,
+      "step": 7363
+    },
+    {
+      "epoch": 1.1268553940321346,
+      "grad_norm": 2.270796945887167,
+      "learning_rate": 8.438959794964168e-06,
+      "loss": 0.4633,
+      "step": 7364
+    },
+    {
+      "epoch": 1.127008416220352,
+      "grad_norm": 2.23165428476424,
+      "learning_rate": 8.436512040857885e-06,
+      "loss": 0.4663,
+      "step": 7365
+    },
+    {
+      "epoch": 1.1271614384085693,
+      "grad_norm": 2.474571748588033,
+      "learning_rate": 8.43406438277151e-06,
+      "loss": 0.4026,
+      "step": 7366
+    },
+    {
+      "epoch": 1.1273144605967866,
+      "grad_norm": 1.9023084237155796,
+      "learning_rate": 8.43161682085537e-06,
+      "loss": 0.4036,
+      "step": 7367
+    },
+    {
+      "epoch": 1.1274674827850037,
+      "grad_norm": 2.2417421961042034,
+      "learning_rate": 8.429169355259778e-06,
+      "loss": 0.401,
+      "step": 7368
+    },
+    {
+      "epoch": 1.127620504973221,
+      "grad_norm": 2.4185358282453864,
+      "learning_rate": 8.426721986135037e-06,
+      "loss": 0.5283,
+      "step": 7369
+    },
+    {
+      "epoch": 1.1277735271614384,
+      "grad_norm": 2.27465022088665,
+      "learning_rate": 8.424274713631455e-06,
+      "loss": 0.4105,
+      "step": 7370
+    },
+    {
+      "epoch": 1.1279265493496557,
+      "grad_norm": 2.2414361092641815,
+      "learning_rate": 8.421827537899325e-06,
+      "loss": 0.4595,
+      "step": 7371
+    },
+    {
+      "epoch": 1.128079571537873,
+      "grad_norm": 2.2130236838711457,
+      "learning_rate": 8.419380459088943e-06,
+      "loss": 0.3782,
+      "step": 7372
+    },
+    {
+      "epoch": 1.1282325937260902,
+      "grad_norm": 2.289444586623291,
+      "learning_rate": 8.41693347735059e-06,
+      "loss": 0.4152,
+      "step": 7373
+    },
+    {
+      "epoch": 1.1283856159143075,
+      "grad_norm": 2.1547623456356573,
+      "learning_rate": 8.414486592834544e-06,
+      "loss": 0.3986,
+      "step": 7374
+    },
+    {
+      "epoch": 1.1285386381025249,
+      "grad_norm": 2.3965413396527033,
+      "learning_rate": 8.412039805691082e-06,
+      "loss": 0.4543,
+      "step": 7375
+    },
+    {
+      "epoch": 1.1286916602907422,
+      "grad_norm": 2.0682256832583192,
+      "learning_rate": 8.409593116070465e-06,
+      "loss": 0.3893,
+      "step": 7376
+    },
+    {
+      "epoch": 1.1288446824789595,
+      "grad_norm": 2.1555593129350967,
+      "learning_rate": 8.407146524122956e-06,
+      "loss": 0.4843,
+      "step": 7377
+    },
+    {
+      "epoch": 1.1289977046671766,
+      "grad_norm": 2.235134611714402,
+      "learning_rate": 8.404700029998813e-06,
+      "loss": 0.4193,
+      "step": 7378
+    },
+    {
+      "epoch": 1.129150726855394,
+      "grad_norm": 2.0948079300685025,
+      "learning_rate": 8.402253633848275e-06,
+      "loss": 0.4189,
+      "step": 7379
+    },
+    {
+      "epoch": 1.1293037490436113,
+      "grad_norm": 2.5618929392411385,
+      "learning_rate": 8.399807335821599e-06,
+      "loss": 0.4643,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1294567712318286,
+      "grad_norm": 2.1511509812008645,
+      "learning_rate": 8.397361136069016e-06,
+      "loss": 0.4467,
+      "step": 7381
+    },
+    {
+      "epoch": 1.129609793420046,
+      "grad_norm": 2.386196178306402,
+      "learning_rate": 8.394915034740746e-06,
+      "loss": 0.4485,
+      "step": 7382
+    },
+    {
+      "epoch": 1.129762815608263,
+      "grad_norm": 2.102940746700613,
+      "learning_rate": 8.39246903198703e-06,
+      "loss": 0.4044,
+      "step": 7383
+    },
+    {
+      "epoch": 1.1299158377964804,
+      "grad_norm": 2.122835102459708,
+      "learning_rate": 8.390023127958076e-06,
+      "loss": 0.3913,
+      "step": 7384
+    },
+    {
+      "epoch": 1.1300688599846977,
+      "grad_norm": 2.1996709406789043,
+      "learning_rate": 8.387577322804102e-06,
+      "loss": 0.4211,
+      "step": 7385
+    },
+    {
+      "epoch": 1.130221882172915,
+      "grad_norm": 2.23042537125408,
+      "learning_rate": 8.385131616675316e-06,
+      "loss": 0.3713,
+      "step": 7386
+    },
+    {
+      "epoch": 1.1303749043611324,
+      "grad_norm": 2.1043667069097745,
+      "learning_rate": 8.382686009721907e-06,
+      "loss": 0.4529,
+      "step": 7387
+    },
+    {
+      "epoch": 1.1305279265493497,
+      "grad_norm": 2.227960780238535,
+      "learning_rate": 8.380240502094083e-06,
+      "loss": 0.4243,
+      "step": 7388
+    },
+    {
+      "epoch": 1.1306809487375669,
+      "grad_norm": 2.4418394916459008,
+      "learning_rate": 8.377795093942026e-06,
+      "loss": 0.48,
+      "step": 7389
+    },
+    {
+      "epoch": 1.1308339709257842,
+      "grad_norm": 2.484896572042387,
+      "learning_rate": 8.375349785415915e-06,
+      "loss": 0.376,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1309869931140015,
+      "grad_norm": 2.1153869841882553,
+      "learning_rate": 8.372904576665931e-06,
+      "loss": 0.4547,
+      "step": 7391
+    },
+    {
+      "epoch": 1.1311400153022189,
+      "grad_norm": 2.135980963386847,
+      "learning_rate": 8.37045946784224e-06,
+      "loss": 0.3852,
+      "step": 7392
+    },
+    {
+      "epoch": 1.1312930374904362,
+      "grad_norm": 2.3502007142725856,
+      "learning_rate": 8.368014459095014e-06,
+      "loss": 0.4678,
+      "step": 7393
+    },
+    {
+      "epoch": 1.1314460596786535,
+      "grad_norm": 2.2129245462799525,
+      "learning_rate": 8.3655695505744e-06,
+      "loss": 0.3617,
+      "step": 7394
+    },
+    {
+      "epoch": 1.1315990818668706,
+      "grad_norm": 2.4771896435377085,
+      "learning_rate": 8.363124742430552e-06,
+      "loss": 0.4377,
+      "step": 7395
+    },
+    {
+      "epoch": 1.131752104055088,
+      "grad_norm": 2.1686976601970414,
+      "learning_rate": 8.360680034813621e-06,
+      "loss": 0.4344,
+      "step": 7396
+    },
+    {
+      "epoch": 1.1319051262433053,
+      "grad_norm": 2.219669874900676,
+      "learning_rate": 8.358235427873741e-06,
+      "loss": 0.3737,
+      "step": 7397
+    },
+    {
+      "epoch": 1.1320581484315226,
+      "grad_norm": 2.7668635869641043,
+      "learning_rate": 8.355790921761045e-06,
+      "loss": 0.4628,
+      "step": 7398
+    },
+    {
+      "epoch": 1.13221117061974,
+      "grad_norm": 2.257255357947881,
+      "learning_rate": 8.353346516625662e-06,
+      "loss": 0.4475,
+      "step": 7399
+    },
+    {
+      "epoch": 1.132364192807957,
+      "grad_norm": 1.7893405512069935,
+      "learning_rate": 8.350902212617708e-06,
+      "loss": 0.3202,
+      "step": 7400
+    },
+    {
+      "epoch": 1.1325172149961744,
+      "grad_norm": 2.188151848584422,
+      "learning_rate": 8.348458009887305e-06,
+      "loss": 0.4079,
+      "step": 7401
+    },
+    {
+      "epoch": 1.1326702371843917,
+      "grad_norm": 2.1347935928648623,
+      "learning_rate": 8.346013908584556e-06,
+      "loss": 0.4305,
+      "step": 7402
+    },
+    {
+      "epoch": 1.132823259372609,
+      "grad_norm": 2.3676766269499026,
+      "learning_rate": 8.343569908859559e-06,
+      "loss": 0.5252,
+      "step": 7403
+    },
+    {
+      "epoch": 1.1329762815608264,
+      "grad_norm": 2.284607042130858,
+      "learning_rate": 8.341126010862417e-06,
+      "loss": 0.4101,
+      "step": 7404
+    },
+    {
+      "epoch": 1.1331293037490435,
+      "grad_norm": 2.3344145212373126,
+      "learning_rate": 8.33868221474322e-06,
+      "loss": 0.4269,
+      "step": 7405
+    },
+    {
+      "epoch": 1.1332823259372609,
+      "grad_norm": 2.371166608832448,
+      "learning_rate": 8.33623852065204e-06,
+      "loss": 0.4966,
+      "step": 7406
+    },
+    {
+      "epoch": 1.1334353481254782,
+      "grad_norm": 2.0768928444278565,
+      "learning_rate": 8.333794928738963e-06,
+      "loss": 0.4451,
+      "step": 7407
+    },
+    {
+      "epoch": 1.1335883703136955,
+      "grad_norm": 2.2826255992028726,
+      "learning_rate": 8.331351439154058e-06,
+      "loss": 0.4068,
+      "step": 7408
+    },
+    {
+      "epoch": 1.1337413925019129,
+      "grad_norm": 2.1691141270829264,
+      "learning_rate": 8.328908052047392e-06,
+      "loss": 0.449,
+      "step": 7409
+    },
+    {
+      "epoch": 1.13389441469013,
+      "grad_norm": 2.0987622141051334,
+      "learning_rate": 8.326464767569018e-06,
+      "loss": 0.4117,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1340474368783473,
+      "grad_norm": 1.9073261756337854,
+      "learning_rate": 8.324021585868987e-06,
+      "loss": 0.3589,
+      "step": 7411
+    },
+    {
+      "epoch": 1.1342004590665646,
+      "grad_norm": 2.145654462179916,
+      "learning_rate": 8.321578507097351e-06,
+      "loss": 0.4972,
+      "step": 7412
+    },
+    {
+      "epoch": 1.134353481254782,
+      "grad_norm": 2.285381702770092,
+      "learning_rate": 8.319135531404143e-06,
+      "loss": 0.4493,
+      "step": 7413
+    },
+    {
+      "epoch": 1.1345065034429993,
+      "grad_norm": 2.0970514541562153,
+      "learning_rate": 8.316692658939396e-06,
+      "loss": 0.4087,
+      "step": 7414
+    },
+    {
+      "epoch": 1.1346595256312164,
+      "grad_norm": 2.2048734136664607,
+      "learning_rate": 8.314249889853141e-06,
+      "loss": 0.4177,
+      "step": 7415
+    },
+    {
+      "epoch": 1.1348125478194337,
+      "grad_norm": 2.1074902936965225,
+      "learning_rate": 8.311807224295391e-06,
+      "loss": 0.4048,
+      "step": 7416
+    },
+    {
+      "epoch": 1.134965570007651,
+      "grad_norm": 2.4923095199826726,
+      "learning_rate": 8.309364662416167e-06,
+      "loss": 0.4147,
+      "step": 7417
+    },
+    {
+      "epoch": 1.1351185921958684,
+      "grad_norm": 2.108320231994888,
+      "learning_rate": 8.306922204365476e-06,
+      "loss": 0.3964,
+      "step": 7418
+    },
+    {
+      "epoch": 1.1352716143840857,
+      "grad_norm": 2.3299560327530404,
+      "learning_rate": 8.304479850293308e-06,
+      "loss": 0.4453,
+      "step": 7419
+    },
+    {
+      "epoch": 1.135424636572303,
+      "grad_norm": 1.9868538731801448,
+      "learning_rate": 8.302037600349669e-06,
+      "loss": 0.3799,
+      "step": 7420
+    },
+    {
+      "epoch": 1.1355776587605202,
+      "grad_norm": 1.9462273485980057,
+      "learning_rate": 8.29959545468454e-06,
+      "loss": 0.4148,
+      "step": 7421
+    },
+    {
+      "epoch": 1.1357306809487375,
+      "grad_norm": 2.234376101383108,
+      "learning_rate": 8.297153413447914e-06,
+      "loss": 0.4836,
+      "step": 7422
+    },
+    {
+      "epoch": 1.1358837031369549,
+      "grad_norm": 2.0793008207140047,
+      "learning_rate": 8.294711476789754e-06,
+      "loss": 0.4197,
+      "step": 7423
+    },
+    {
+      "epoch": 1.1360367253251722,
+      "grad_norm": 2.257538544402379,
+      "learning_rate": 8.292269644860032e-06,
+      "loss": 0.4509,
+      "step": 7424
+    },
+    {
+      "epoch": 1.1361897475133895,
+      "grad_norm": 2.200890531426487,
+      "learning_rate": 8.289827917808715e-06,
+      "loss": 0.472,
+      "step": 7425
+    },
+    {
+      "epoch": 1.1363427697016066,
+      "grad_norm": 2.278989470946727,
+      "learning_rate": 8.287386295785755e-06,
+      "loss": 0.3801,
+      "step": 7426
+    },
+    {
+      "epoch": 1.136495791889824,
+      "grad_norm": 2.39384075061946,
+      "learning_rate": 8.284944778941098e-06,
+      "loss": 0.5339,
+      "step": 7427
+    },
+    {
+      "epoch": 1.1366488140780413,
+      "grad_norm": 2.1283550645597367,
+      "learning_rate": 8.282503367424697e-06,
+      "loss": 0.4964,
+      "step": 7428
+    },
+    {
+      "epoch": 1.1368018362662586,
+      "grad_norm": 2.0117879432634798,
+      "learning_rate": 8.280062061386478e-06,
+      "loss": 0.4534,
+      "step": 7429
+    },
+    {
+      "epoch": 1.136954858454476,
+      "grad_norm": 2.2323821418668883,
+      "learning_rate": 8.277620860976382e-06,
+      "loss": 0.468,
+      "step": 7430
+    },
+    {
+      "epoch": 1.1371078806426933,
+      "grad_norm": 2.1809204167250678,
+      "learning_rate": 8.275179766344325e-06,
+      "loss": 0.4446,
+      "step": 7431
+    },
+    {
+      "epoch": 1.1372609028309104,
+      "grad_norm": 2.193452628067021,
+      "learning_rate": 8.272738777640224e-06,
+      "loss": 0.4116,
+      "step": 7432
+    },
+    {
+      "epoch": 1.1374139250191277,
+      "grad_norm": 2.288415436946979,
+      "learning_rate": 8.270297895013992e-06,
+      "loss": 0.496,
+      "step": 7433
+    },
+    {
+      "epoch": 1.137566947207345,
+      "grad_norm": 2.18335735896481,
+      "learning_rate": 8.267857118615536e-06,
+      "loss": 0.4222,
+      "step": 7434
+    },
+    {
+      "epoch": 1.1377199693955624,
+      "grad_norm": 2.181193321791857,
+      "learning_rate": 8.265416448594747e-06,
+      "loss": 0.372,
+      "step": 7435
+    },
+    {
+      "epoch": 1.1378729915837797,
+      "grad_norm": 2.268232897097735,
+      "learning_rate": 8.26297588510152e-06,
+      "loss": 0.494,
+      "step": 7436
+    },
+    {
+      "epoch": 1.1380260137719969,
+      "grad_norm": 2.122086095027104,
+      "learning_rate": 8.260535428285739e-06,
+      "loss": 0.3753,
+      "step": 7437
+    },
+    {
+      "epoch": 1.1381790359602142,
+      "grad_norm": 2.526263316541223,
+      "learning_rate": 8.258095078297286e-06,
+      "loss": 0.442,
+      "step": 7438
+    },
+    {
+      "epoch": 1.1383320581484315,
+      "grad_norm": 2.1160673691906653,
+      "learning_rate": 8.255654835286025e-06,
+      "loss": 0.3435,
+      "step": 7439
+    },
+    {
+      "epoch": 1.1384850803366489,
+      "grad_norm": 2.114820062384801,
+      "learning_rate": 8.253214699401824e-06,
+      "loss": 0.4397,
+      "step": 7440
+    },
+    {
+      "epoch": 1.1386381025248662,
+      "grad_norm": 2.3148968139526325,
+      "learning_rate": 8.250774670794545e-06,
+      "loss": 0.4095,
+      "step": 7441
+    },
+    {
+      "epoch": 1.1387911247130833,
+      "grad_norm": 2.3769438544271257,
+      "learning_rate": 8.248334749614035e-06,
+      "loss": 0.4592,
+      "step": 7442
+    },
+    {
+      "epoch": 1.1389441469013006,
+      "grad_norm": 1.9651717159386775,
+      "learning_rate": 8.245894936010139e-06,
+      "loss": 0.4159,
+      "step": 7443
+    },
+    {
+      "epoch": 1.139097169089518,
+      "grad_norm": 1.93039744862964,
+      "learning_rate": 8.243455230132699e-06,
+      "loss": 0.389,
+      "step": 7444
+    },
+    {
+      "epoch": 1.1392501912777353,
+      "grad_norm": 2.2701040105758095,
+      "learning_rate": 8.241015632131543e-06,
+      "loss": 0.4002,
+      "step": 7445
+    },
+    {
+      "epoch": 1.1394032134659526,
+      "grad_norm": 2.456302414590909,
+      "learning_rate": 8.2385761421565e-06,
+      "loss": 0.5128,
+      "step": 7446
+    },
+    {
+      "epoch": 1.1395562356541697,
+      "grad_norm": 2.441590053887338,
+      "learning_rate": 8.236136760357388e-06,
+      "loss": 0.4405,
+      "step": 7447
+    },
+    {
+      "epoch": 1.139709257842387,
+      "grad_norm": 2.142496840381271,
+      "learning_rate": 8.233697486884014e-06,
+      "loss": 0.3596,
+      "step": 7448
+    },
+    {
+      "epoch": 1.1398622800306044,
+      "grad_norm": 2.2436694830487847,
+      "learning_rate": 8.23125832188619e-06,
+      "loss": 0.3891,
+      "step": 7449
+    },
+    {
+      "epoch": 1.1400153022188217,
+      "grad_norm": 2.3067821364697085,
+      "learning_rate": 8.228819265513713e-06,
+      "loss": 0.356,
+      "step": 7450
+    },
+    {
+      "epoch": 1.140168324407039,
+      "grad_norm": 2.2973291155628903,
+      "learning_rate": 8.22638031791637e-06,
+      "loss": 0.4147,
+      "step": 7451
+    },
+    {
+      "epoch": 1.1403213465952562,
+      "grad_norm": 2.3150537810209344,
+      "learning_rate": 8.223941479243952e-06,
+      "loss": 0.4021,
+      "step": 7452
+    },
+    {
+      "epoch": 1.1404743687834735,
+      "grad_norm": 2.1101437473072373,
+      "learning_rate": 8.221502749646233e-06,
+      "loss": 0.3884,
+      "step": 7453
+    },
+    {
+      "epoch": 1.1406273909716909,
+      "grad_norm": 2.2940668477947015,
+      "learning_rate": 8.219064129272993e-06,
+      "loss": 0.4888,
+      "step": 7454
+    },
+    {
+      "epoch": 1.1407804131599082,
+      "grad_norm": 2.5443419586116107,
+      "learning_rate": 8.21662561827399e-06,
+      "loss": 0.4524,
+      "step": 7455
+    },
+    {
+      "epoch": 1.1409334353481255,
+      "grad_norm": 2.2535777983806433,
+      "learning_rate": 8.214187216798982e-06,
+      "loss": 0.4283,
+      "step": 7456
+    },
+    {
+      "epoch": 1.1410864575363429,
+      "grad_norm": 2.2630948114110048,
+      "learning_rate": 8.211748924997727e-06,
+      "loss": 0.3295,
+      "step": 7457
+    },
+    {
+      "epoch": 1.14123947972456,
+      "grad_norm": 2.3979824581281535,
+      "learning_rate": 8.20931074301996e-06,
+      "loss": 0.4793,
+      "step": 7458
+    },
+    {
+      "epoch": 1.1413925019127773,
+      "grad_norm": 2.849113718168752,
+      "learning_rate": 8.206872671015431e-06,
+      "loss": 0.4199,
+      "step": 7459
+    },
+    {
+      "epoch": 1.1415455241009946,
+      "grad_norm": 2.4386612095527807,
+      "learning_rate": 8.204434709133868e-06,
+      "loss": 0.3872,
+      "step": 7460
+    },
+    {
+      "epoch": 1.141698546289212,
+      "grad_norm": 2.6483087030550845,
+      "learning_rate": 8.201996857524985e-06,
+      "loss": 0.4643,
+      "step": 7461
+    },
+    {
+      "epoch": 1.1418515684774293,
+      "grad_norm": 1.8880437723686634,
+      "learning_rate": 8.199559116338516e-06,
+      "loss": 0.3426,
+      "step": 7462
+    },
+    {
+      "epoch": 1.1420045906656466,
+      "grad_norm": 2.252181667335893,
+      "learning_rate": 8.197121485724165e-06,
+      "loss": 0.3879,
+      "step": 7463
+    },
+    {
+      "epoch": 1.1421576128538637,
+      "grad_norm": 2.2676544247977684,
+      "learning_rate": 8.194683965831632e-06,
+      "loss": 0.4494,
+      "step": 7464
+    },
+    {
+      "epoch": 1.142310635042081,
+      "grad_norm": 2.5427993887949705,
+      "learning_rate": 8.192246556810623e-06,
+      "loss": 0.4561,
+      "step": 7465
+    },
+    {
+      "epoch": 1.1424636572302984,
+      "grad_norm": 2.4818084453252025,
+      "learning_rate": 8.189809258810821e-06,
+      "loss": 0.4889,
+      "step": 7466
+    },
+    {
+      "epoch": 1.1426166794185157,
+      "grad_norm": 2.1968397362218144,
+      "learning_rate": 8.187372071981918e-06,
+      "loss": 0.4769,
+      "step": 7467
+    },
+    {
+      "epoch": 1.142769701606733,
+      "grad_norm": 2.382228924959736,
+      "learning_rate": 8.184934996473588e-06,
+      "loss": 0.4139,
+      "step": 7468
+    },
+    {
+      "epoch": 1.1429227237949502,
+      "grad_norm": 2.2384393655517107,
+      "learning_rate": 8.182498032435495e-06,
+      "loss": 0.4245,
+      "step": 7469
+    },
+    {
+      "epoch": 1.1430757459831675,
+      "grad_norm": 2.1015126974026064,
+      "learning_rate": 8.180061180017312e-06,
+      "loss": 0.3678,
+      "step": 7470
+    },
+    {
+      "epoch": 1.1432287681713849,
+      "grad_norm": 2.428458221772654,
+      "learning_rate": 8.177624439368692e-06,
+      "loss": 0.378,
+      "step": 7471
+    },
+    {
+      "epoch": 1.1433817903596022,
+      "grad_norm": 2.2063195875131902,
+      "learning_rate": 8.175187810639281e-06,
+      "loss": 0.4007,
+      "step": 7472
+    },
+    {
+      "epoch": 1.1435348125478195,
+      "grad_norm": 2.1957528996819184,
+      "learning_rate": 8.172751293978727e-06,
+      "loss": 0.4849,
+      "step": 7473
+    },
+    {
+      "epoch": 1.1436878347360366,
+      "grad_norm": 2.3310595610922826,
+      "learning_rate": 8.170314889536663e-06,
+      "loss": 0.4719,
+      "step": 7474
+    },
+    {
+      "epoch": 1.143840856924254,
+      "grad_norm": 2.1956159717965815,
+      "learning_rate": 8.167878597462724e-06,
+      "loss": 0.4452,
+      "step": 7475
+    },
+    {
+      "epoch": 1.1439938791124713,
+      "grad_norm": 2.6239595075786855,
+      "learning_rate": 8.165442417906523e-06,
+      "loss": 0.401,
+      "step": 7476
+    },
+    {
+      "epoch": 1.1441469013006886,
+      "grad_norm": 2.4291514293723875,
+      "learning_rate": 8.163006351017681e-06,
+      "loss": 0.4379,
+      "step": 7477
+    },
+    {
+      "epoch": 1.144299923488906,
+      "grad_norm": 2.1139992285994014,
+      "learning_rate": 8.160570396945808e-06,
+      "loss": 0.4388,
+      "step": 7478
+    },
+    {
+      "epoch": 1.144452945677123,
+      "grad_norm": 2.1633213676632015,
+      "learning_rate": 8.1581345558405e-06,
+      "loss": 0.4151,
+      "step": 7479
+    },
+    {
+      "epoch": 1.1446059678653404,
+      "grad_norm": 2.2390982385776397,
+      "learning_rate": 8.155698827851354e-06,
+      "loss": 0.412,
+      "step": 7480
+    },
+    {
+      "epoch": 1.1447589900535577,
+      "grad_norm": 2.2089163980451088,
+      "learning_rate": 8.153263213127961e-06,
+      "loss": 0.4565,
+      "step": 7481
+    },
+    {
+      "epoch": 1.144912012241775,
+      "grad_norm": 2.176734428087088,
+      "learning_rate": 8.150827711819894e-06,
+      "loss": 0.4647,
+      "step": 7482
+    },
+    {
+      "epoch": 1.1450650344299924,
+      "grad_norm": 2.253278709863533,
+      "learning_rate": 8.148392324076733e-06,
+      "loss": 0.4017,
+      "step": 7483
+    },
+    {
+      "epoch": 1.1452180566182095,
+      "grad_norm": 2.152174433588315,
+      "learning_rate": 8.145957050048047e-06,
+      "loss": 0.3346,
+      "step": 7484
+    },
+    {
+      "epoch": 1.1453710788064269,
+      "grad_norm": 2.047079847330625,
+      "learning_rate": 8.143521889883385e-06,
+      "loss": 0.3476,
+      "step": 7485
+    },
+    {
+      "epoch": 1.1455241009946442,
+      "grad_norm": 2.4767797743598443,
+      "learning_rate": 8.141086843732311e-06,
+      "loss": 0.4391,
+      "step": 7486
+    },
+    {
+      "epoch": 1.1456771231828615,
+      "grad_norm": 2.0852254844893476,
+      "learning_rate": 8.138651911744362e-06,
+      "loss": 0.3421,
+      "step": 7487
+    },
+    {
+      "epoch": 1.1458301453710789,
+      "grad_norm": 2.3333992177155394,
+      "learning_rate": 8.136217094069085e-06,
+      "loss": 0.4186,
+      "step": 7488
+    },
+    {
+      "epoch": 1.1459831675592962,
+      "grad_norm": 2.242432354816371,
+      "learning_rate": 8.133782390856007e-06,
+      "loss": 0.4564,
+      "step": 7489
+    },
+    {
+      "epoch": 1.1461361897475133,
+      "grad_norm": 2.1801434144194447,
+      "learning_rate": 8.131347802254649e-06,
+      "loss": 0.4897,
+      "step": 7490
+    },
+    {
+      "epoch": 1.1462892119357306,
+      "grad_norm": 2.7231872288313315,
+      "learning_rate": 8.128913328414538e-06,
+      "loss": 0.417,
+      "step": 7491
+    },
+    {
+      "epoch": 1.146442234123948,
+      "grad_norm": 2.113303644811708,
+      "learning_rate": 8.126478969485176e-06,
+      "loss": 0.5045,
+      "step": 7492
+    },
+    {
+      "epoch": 1.1465952563121653,
+      "grad_norm": 2.095819049578236,
+      "learning_rate": 8.124044725616069e-06,
+      "loss": 0.426,
+      "step": 7493
+    },
+    {
+      "epoch": 1.1467482785003826,
+      "grad_norm": 2.040810152227828,
+      "learning_rate": 8.121610596956718e-06,
+      "loss": 0.4056,
+      "step": 7494
+    },
+    {
+      "epoch": 1.1469013006886,
+      "grad_norm": 2.1692836388967125,
+      "learning_rate": 8.119176583656603e-06,
+      "loss": 0.44,
+      "step": 7495
+    },
+    {
+      "epoch": 1.147054322876817,
+      "grad_norm": 2.2409161920316434,
+      "learning_rate": 8.116742685865217e-06,
+      "loss": 0.3894,
+      "step": 7496
+    },
+    {
+      "epoch": 1.1472073450650344,
+      "grad_norm": 2.1774102605411776,
+      "learning_rate": 8.114308903732028e-06,
+      "loss": 0.4041,
+      "step": 7497
+    },
+    {
+      "epoch": 1.1473603672532517,
+      "grad_norm": 1.728871470396512,
+      "learning_rate": 8.111875237406506e-06,
+      "loss": 0.3032,
+      "step": 7498
+    },
+    {
+      "epoch": 1.147513389441469,
+      "grad_norm": 2.3391781721148743,
+      "learning_rate": 8.109441687038111e-06,
+      "loss": 0.4852,
+      "step": 7499
+    },
+    {
+      "epoch": 1.1476664116296864,
+      "grad_norm": 2.104355287358558,
+      "learning_rate": 8.107008252776301e-06,
+      "loss": 0.4481,
+      "step": 7500
+    },
+    {
+      "epoch": 1.1478194338179035,
+      "grad_norm": 2.2728732595750474,
+      "learning_rate": 8.104574934770516e-06,
+      "loss": 0.3932,
+      "step": 7501
+    },
+    {
+      "epoch": 1.1479724560061209,
+      "grad_norm": 2.7508028520265686,
+      "learning_rate": 8.102141733170202e-06,
+      "loss": 0.422,
+      "step": 7502
+    },
+    {
+      "epoch": 1.1481254781943382,
+      "grad_norm": 2.3486796594495885,
+      "learning_rate": 8.099708648124785e-06,
+      "loss": 0.542,
+      "step": 7503
+    },
+    {
+      "epoch": 1.1482785003825555,
+      "grad_norm": 2.2815892384437806,
+      "learning_rate": 8.097275679783698e-06,
+      "loss": 0.4602,
+      "step": 7504
+    },
+    {
+      "epoch": 1.1484315225707729,
+      "grad_norm": 2.371707040162622,
+      "learning_rate": 8.094842828296354e-06,
+      "loss": 0.4856,
+      "step": 7505
+    },
+    {
+      "epoch": 1.14858454475899,
+      "grad_norm": 1.8999703078986259,
+      "learning_rate": 8.092410093812161e-06,
+      "loss": 0.3464,
+      "step": 7506
+    },
+    {
+      "epoch": 1.1487375669472073,
+      "grad_norm": 1.9999448169209695,
+      "learning_rate": 8.089977476480533e-06,
+      "loss": 0.3793,
+      "step": 7507
+    },
+    {
+      "epoch": 1.1488905891354246,
+      "grad_norm": 2.101338405608802,
+      "learning_rate": 8.08754497645086e-06,
+      "loss": 0.367,
+      "step": 7508
+    },
+    {
+      "epoch": 1.149043611323642,
+      "grad_norm": 2.254916559111806,
+      "learning_rate": 8.085112593872524e-06,
+      "loss": 0.4823,
+      "step": 7509
+    },
+    {
+      "epoch": 1.1491966335118593,
+      "grad_norm": 2.2298089722186103,
+      "learning_rate": 8.082680328894923e-06,
+      "loss": 0.4959,
+      "step": 7510
+    },
+    {
+      "epoch": 1.1493496557000764,
+      "grad_norm": 2.017233130477621,
+      "learning_rate": 8.080248181667417e-06,
+      "loss": 0.4274,
+      "step": 7511
+    },
+    {
+      "epoch": 1.1495026778882937,
+      "grad_norm": 2.358574713364482,
+      "learning_rate": 8.077816152339383e-06,
+      "loss": 0.5025,
+      "step": 7512
+    },
+    {
+      "epoch": 1.149655700076511,
+      "grad_norm": 2.0398205910476657,
+      "learning_rate": 8.075384241060183e-06,
+      "loss": 0.4287,
+      "step": 7513
+    },
+    {
+      "epoch": 1.1498087222647284,
+      "grad_norm": 2.0429358861791305,
+      "learning_rate": 8.072952447979159e-06,
+      "loss": 0.3928,
+      "step": 7514
+    },
+    {
+      "epoch": 1.1499617444529457,
+      "grad_norm": 2.141370554482417,
+      "learning_rate": 8.070520773245667e-06,
+      "loss": 0.3852,
+      "step": 7515
+    },
+    {
+      "epoch": 1.1501147666411629,
+      "grad_norm": 2.169035469722936,
+      "learning_rate": 8.068089217009045e-06,
+      "loss": 0.3634,
+      "step": 7516
+    },
+    {
+      "epoch": 1.1502677888293802,
+      "grad_norm": 2.4022699362669058,
+      "learning_rate": 8.065657779418617e-06,
+      "loss": 0.4754,
+      "step": 7517
+    },
+    {
+      "epoch": 1.1504208110175975,
+      "grad_norm": 2.0431104935895847,
+      "learning_rate": 8.063226460623714e-06,
+      "loss": 0.371,
+      "step": 7518
+    },
+    {
+      "epoch": 1.1505738332058149,
+      "grad_norm": 2.0616275176271235,
+      "learning_rate": 8.060795260773649e-06,
+      "loss": 0.4364,
+      "step": 7519
+    },
+    {
+      "epoch": 1.1507268553940322,
+      "grad_norm": 2.389435632037756,
+      "learning_rate": 8.058364180017738e-06,
+      "loss": 0.4303,
+      "step": 7520
+    },
+    {
+      "epoch": 1.1508798775822495,
+      "grad_norm": 2.390938065826764,
+      "learning_rate": 8.055933218505275e-06,
+      "loss": 0.4162,
+      "step": 7521
+    },
+    {
+      "epoch": 1.1510328997704666,
+      "grad_norm": 2.0817572989263735,
+      "learning_rate": 8.053502376385555e-06,
+      "loss": 0.4079,
+      "step": 7522
+    },
+    {
+      "epoch": 1.151185921958684,
+      "grad_norm": 1.9572766628941676,
+      "learning_rate": 8.051071653807877e-06,
+      "loss": 0.3664,
+      "step": 7523
+    },
+    {
+      "epoch": 1.1513389441469013,
+      "grad_norm": 1.97525311017114,
+      "learning_rate": 8.048641050921505e-06,
+      "loss": 0.3817,
+      "step": 7524
+    },
+    {
+      "epoch": 1.1514919663351186,
+      "grad_norm": 2.189527479601311,
+      "learning_rate": 8.046210567875725e-06,
+      "loss": 0.4619,
+      "step": 7525
+    },
+    {
+      "epoch": 1.151644988523336,
+      "grad_norm": 2.0684633513104136,
+      "learning_rate": 8.043780204819796e-06,
+      "loss": 0.3981,
+      "step": 7526
+    },
+    {
+      "epoch": 1.1517980107115533,
+      "grad_norm": 2.199448040768017,
+      "learning_rate": 8.041349961902976e-06,
+      "loss": 0.3961,
+      "step": 7527
+    },
+    {
+      "epoch": 1.1519510328997704,
+      "grad_norm": 2.2150480648254236,
+      "learning_rate": 8.038919839274519e-06,
+      "loss": 0.4216,
+      "step": 7528
+    },
+    {
+      "epoch": 1.1521040550879877,
+      "grad_norm": 2.292813109070026,
+      "learning_rate": 8.036489837083668e-06,
+      "loss": 0.3829,
+      "step": 7529
+    },
+    {
+      "epoch": 1.152257077276205,
+      "grad_norm": 2.065860761981022,
+      "learning_rate": 8.034059955479652e-06,
+      "loss": 0.398,
+      "step": 7530
+    },
+    {
+      "epoch": 1.1524100994644224,
+      "grad_norm": 2.1957137938311764,
+      "learning_rate": 8.031630194611708e-06,
+      "loss": 0.3777,
+      "step": 7531
+    },
+    {
+      "epoch": 1.1525631216526397,
+      "grad_norm": 2.395586780136042,
+      "learning_rate": 8.029200554629052e-06,
+      "loss": 0.4595,
+      "step": 7532
+    },
+    {
+      "epoch": 1.1527161438408569,
+      "grad_norm": 2.354007174674176,
+      "learning_rate": 8.026771035680905e-06,
+      "loss": 0.4501,
+      "step": 7533
+    },
+    {
+      "epoch": 1.1528691660290742,
+      "grad_norm": 2.1309239235307276,
+      "learning_rate": 8.024341637916465e-06,
+      "loss": 0.4385,
+      "step": 7534
+    },
+    {
+      "epoch": 1.1530221882172915,
+      "grad_norm": 2.0989633908232554,
+      "learning_rate": 8.02191236148493e-06,
+      "loss": 0.4596,
+      "step": 7535
+    },
+    {
+      "epoch": 1.1531752104055089,
+      "grad_norm": 2.0051365553355307,
+      "learning_rate": 8.019483206535501e-06,
+      "loss": 0.3603,
+      "step": 7536
+    },
+    {
+      "epoch": 1.1533282325937262,
+      "grad_norm": 2.1379711591576602,
+      "learning_rate": 8.017054173217354e-06,
+      "loss": 0.4969,
+      "step": 7537
+    },
+    {
+      "epoch": 1.1534812547819433,
+      "grad_norm": 2.0414756729412007,
+      "learning_rate": 8.014625261679666e-06,
+      "loss": 0.3722,
+      "step": 7538
+    },
+    {
+      "epoch": 1.1536342769701606,
+      "grad_norm": 1.960059689751741,
+      "learning_rate": 8.012196472071612e-06,
+      "loss": 0.5155,
+      "step": 7539
+    },
+    {
+      "epoch": 1.153787299158378,
+      "grad_norm": 1.9148456988264915,
+      "learning_rate": 8.009767804542341e-06,
+      "loss": 0.3806,
+      "step": 7540
+    },
+    {
+      "epoch": 1.1539403213465953,
+      "grad_norm": 2.1346428516068507,
+      "learning_rate": 8.007339259241022e-06,
+      "loss": 0.3781,
+      "step": 7541
+    },
+    {
+      "epoch": 1.1540933435348126,
+      "grad_norm": 2.341555927456598,
+      "learning_rate": 8.004910836316796e-06,
+      "loss": 0.4134,
+      "step": 7542
+    },
+    {
+      "epoch": 1.1542463657230297,
+      "grad_norm": 2.1533856307933448,
+      "learning_rate": 8.002482535918792e-06,
+      "loss": 0.3918,
+      "step": 7543
+    },
+    {
+      "epoch": 1.154399387911247,
+      "grad_norm": 2.203898129272653,
+      "learning_rate": 8.000054358196156e-06,
+      "loss": 0.3875,
+      "step": 7544
+    },
+    {
+      "epoch": 1.1545524100994644,
+      "grad_norm": 2.0530530827506106,
+      "learning_rate": 7.997626303298008e-06,
+      "loss": 0.4047,
+      "step": 7545
+    },
+    {
+      "epoch": 1.1547054322876817,
+      "grad_norm": 2.1846781568623026,
+      "learning_rate": 7.995198371373455e-06,
+      "loss": 0.4748,
+      "step": 7546
+    },
+    {
+      "epoch": 1.154858454475899,
+      "grad_norm": 2.369642955439486,
+      "learning_rate": 7.992770562571616e-06,
+      "loss": 0.4418,
+      "step": 7547
+    },
+    {
+      "epoch": 1.1550114766641162,
+      "grad_norm": 2.876693910486257,
+      "learning_rate": 7.990342877041588e-06,
+      "loss": 0.3995,
+      "step": 7548
+    },
+    {
+      "epoch": 1.1551644988523335,
+      "grad_norm": 2.2881944470526503,
+      "learning_rate": 7.987915314932467e-06,
+      "loss": 0.4277,
+      "step": 7549
+    },
+    {
+      "epoch": 1.1553175210405509,
+      "grad_norm": 2.082026776655421,
+      "learning_rate": 7.98548787639334e-06,
+      "loss": 0.474,
+      "step": 7550
+    },
+    {
+      "epoch": 1.1554705432287682,
+      "grad_norm": 2.4770772431505623,
+      "learning_rate": 7.98306056157328e-06,
+      "loss": 0.4932,
+      "step": 7551
+    },
+    {
+      "epoch": 1.1556235654169855,
+      "grad_norm": 2.400411126727818,
+      "learning_rate": 7.980633370621361e-06,
+      "loss": 0.417,
+      "step": 7552
+    },
+    {
+      "epoch": 1.1557765876052026,
+      "grad_norm": 2.249144501782921,
+      "learning_rate": 7.97820630368665e-06,
+      "loss": 0.4198,
+      "step": 7553
+    },
+    {
+      "epoch": 1.15592960979342,
+      "grad_norm": 2.4416647224932526,
+      "learning_rate": 7.975779360918196e-06,
+      "loss": 0.4977,
+      "step": 7554
+    },
+    {
+      "epoch": 1.1560826319816373,
+      "grad_norm": 2.315267686111871,
+      "learning_rate": 7.973352542465052e-06,
+      "loss": 0.3786,
+      "step": 7555
+    },
+    {
+      "epoch": 1.1562356541698546,
+      "grad_norm": 2.12994967417012,
+      "learning_rate": 7.970925848476253e-06,
+      "loss": 0.4149,
+      "step": 7556
+    },
+    {
+      "epoch": 1.156388676358072,
+      "grad_norm": 2.4649859720017604,
+      "learning_rate": 7.968499279100841e-06,
+      "loss": 0.5204,
+      "step": 7557
+    },
+    {
+      "epoch": 1.1565416985462893,
+      "grad_norm": 2.546695379639944,
+      "learning_rate": 7.966072834487832e-06,
+      "loss": 0.4823,
+      "step": 7558
+    },
+    {
+      "epoch": 1.1566947207345064,
+      "grad_norm": 2.111908023754244,
+      "learning_rate": 7.963646514786246e-06,
+      "loss": 0.4027,
+      "step": 7559
+    },
+    {
+      "epoch": 1.1568477429227237,
+      "grad_norm": 2.079272062182409,
+      "learning_rate": 7.961220320145099e-06,
+      "loss": 0.3618,
+      "step": 7560
+    },
+    {
+      "epoch": 1.157000765110941,
+      "grad_norm": 2.0099331247968517,
+      "learning_rate": 7.958794250713381e-06,
+      "loss": 0.3389,
+      "step": 7561
+    },
+    {
+      "epoch": 1.1571537872991584,
+      "grad_norm": 2.3728785319967693,
+      "learning_rate": 7.956368306640097e-06,
+      "loss": 0.4472,
+      "step": 7562
+    },
+    {
+      "epoch": 1.1573068094873757,
+      "grad_norm": 2.124001664400862,
+      "learning_rate": 7.953942488074233e-06,
+      "loss": 0.4035,
+      "step": 7563
+    },
+    {
+      "epoch": 1.157459831675593,
+      "grad_norm": 2.2562599575262685,
+      "learning_rate": 7.95151679516476e-06,
+      "loss": 0.4425,
+      "step": 7564
+    },
+    {
+      "epoch": 1.1576128538638102,
+      "grad_norm": 2.1706628764944798,
+      "learning_rate": 7.949091228060657e-06,
+      "loss": 0.4145,
+      "step": 7565
+    },
+    {
+      "epoch": 1.1577658760520275,
+      "grad_norm": 1.9386706013096928,
+      "learning_rate": 7.946665786910885e-06,
+      "loss": 0.3171,
+      "step": 7566
+    },
+    {
+      "epoch": 1.1579188982402449,
+      "grad_norm": 2.2644379442856297,
+      "learning_rate": 7.944240471864398e-06,
+      "loss": 0.4577,
+      "step": 7567
+    },
+    {
+      "epoch": 1.1580719204284622,
+      "grad_norm": 2.2324278833799864,
+      "learning_rate": 7.941815283070147e-06,
+      "loss": 0.4197,
+      "step": 7568
+    },
+    {
+      "epoch": 1.1582249426166795,
+      "grad_norm": 2.4366820568521512,
+      "learning_rate": 7.939390220677068e-06,
+      "loss": 0.4956,
+      "step": 7569
+    },
+    {
+      "epoch": 1.1583779648048966,
+      "grad_norm": 2.274130967656845,
+      "learning_rate": 7.936965284834102e-06,
+      "loss": 0.4817,
+      "step": 7570
+    },
+    {
+      "epoch": 1.158530986993114,
+      "grad_norm": 2.214947057393825,
+      "learning_rate": 7.934540475690167e-06,
+      "loss": 0.4042,
+      "step": 7571
+    },
+    {
+      "epoch": 1.1586840091813313,
+      "grad_norm": 2.2361223952851823,
+      "learning_rate": 7.932115793394177e-06,
+      "loss": 0.3838,
+      "step": 7572
+    },
+    {
+      "epoch": 1.1588370313695486,
+      "grad_norm": 2.4647857683566823,
+      "learning_rate": 7.929691238095053e-06,
+      "loss": 0.4323,
+      "step": 7573
+    },
+    {
+      "epoch": 1.158990053557766,
+      "grad_norm": 2.2605032776577287,
+      "learning_rate": 7.927266809941684e-06,
+      "loss": 0.4403,
+      "step": 7574
+    },
+    {
+      "epoch": 1.159143075745983,
+      "grad_norm": 2.307062509681339,
+      "learning_rate": 7.924842509082968e-06,
+      "loss": 0.4219,
+      "step": 7575
+    },
+    {
+      "epoch": 1.1592960979342004,
+      "grad_norm": 2.2580402550302336,
+      "learning_rate": 7.922418335667796e-06,
+      "loss": 0.4076,
+      "step": 7576
+    },
+    {
+      "epoch": 1.1594491201224177,
+      "grad_norm": 1.926770123976774,
+      "learning_rate": 7.919994289845038e-06,
+      "loss": 0.3209,
+      "step": 7577
+    },
+    {
+      "epoch": 1.159602142310635,
+      "grad_norm": 1.88033873862793,
+      "learning_rate": 7.917570371763568e-06,
+      "loss": 0.3106,
+      "step": 7578
+    },
+    {
+      "epoch": 1.1597551644988524,
+      "grad_norm": 2.3143555284094726,
+      "learning_rate": 7.915146581572253e-06,
+      "loss": 0.4199,
+      "step": 7579
+    },
+    {
+      "epoch": 1.1599081866870695,
+      "grad_norm": 2.5111769532452524,
+      "learning_rate": 7.912722919419936e-06,
+      "loss": 0.4582,
+      "step": 7580
+    },
+    {
+      "epoch": 1.1600612088752869,
+      "grad_norm": 2.159383285210872,
+      "learning_rate": 7.910299385455472e-06,
+      "loss": 0.3882,
+      "step": 7581
+    },
+    {
+      "epoch": 1.1602142310635042,
+      "grad_norm": 2.0463292222580227,
+      "learning_rate": 7.9078759798277e-06,
+      "loss": 0.3195,
+      "step": 7582
+    },
+    {
+      "epoch": 1.1603672532517215,
+      "grad_norm": 2.315690123330286,
+      "learning_rate": 7.905452702685446e-06,
+      "loss": 0.4594,
+      "step": 7583
+    },
+    {
+      "epoch": 1.1605202754399389,
+      "grad_norm": 2.4911195808536464,
+      "learning_rate": 7.903029554177535e-06,
+      "loss": 0.4797,
+      "step": 7584
+    },
+    {
+      "epoch": 1.160673297628156,
+      "grad_norm": 2.0446690195381847,
+      "learning_rate": 7.900606534452782e-06,
+      "loss": 0.4061,
+      "step": 7585
+    },
+    {
+      "epoch": 1.1608263198163733,
+      "grad_norm": 2.1561715447958574,
+      "learning_rate": 7.898183643659998e-06,
+      "loss": 0.4537,
+      "step": 7586
+    },
+    {
+      "epoch": 1.1609793420045906,
+      "grad_norm": 2.3006046638561366,
+      "learning_rate": 7.895760881947976e-06,
+      "loss": 0.5039,
+      "step": 7587
+    },
+    {
+      "epoch": 1.161132364192808,
+      "grad_norm": 2.495860237065185,
+      "learning_rate": 7.89333824946551e-06,
+      "loss": 0.4654,
+      "step": 7588
+    },
+    {
+      "epoch": 1.1612853863810253,
+      "grad_norm": 2.1959113314709873,
+      "learning_rate": 7.890915746361388e-06,
+      "loss": 0.3754,
+      "step": 7589
+    },
+    {
+      "epoch": 1.1614384085692426,
+      "grad_norm": 2.2157690704589954,
+      "learning_rate": 7.888493372784375e-06,
+      "loss": 0.4643,
+      "step": 7590
+    },
+    {
+      "epoch": 1.1615914307574597,
+      "grad_norm": 2.1585183274845083,
+      "learning_rate": 7.886071128883245e-06,
+      "loss": 0.454,
+      "step": 7591
+    },
+    {
+      "epoch": 1.161744452945677,
+      "grad_norm": 2.3488517000482165,
+      "learning_rate": 7.883649014806762e-06,
+      "loss": 0.4765,
+      "step": 7592
+    },
+    {
+      "epoch": 1.1618974751338944,
+      "grad_norm": 2.3071998807798417,
+      "learning_rate": 7.881227030703666e-06,
+      "loss": 0.4174,
+      "step": 7593
+    },
+    {
+      "epoch": 1.1620504973221117,
+      "grad_norm": 2.0712309131234066,
+      "learning_rate": 7.878805176722708e-06,
+      "loss": 0.4144,
+      "step": 7594
+    },
+    {
+      "epoch": 1.162203519510329,
+      "grad_norm": 2.09316779209673,
+      "learning_rate": 7.876383453012626e-06,
+      "loss": 0.4189,
+      "step": 7595
+    },
+    {
+      "epoch": 1.1623565416985464,
+      "grad_norm": 1.9814309260192482,
+      "learning_rate": 7.873961859722139e-06,
+      "loss": 0.3391,
+      "step": 7596
+    },
+    {
+      "epoch": 1.1625095638867635,
+      "grad_norm": 2.1146901984425783,
+      "learning_rate": 7.871540396999974e-06,
+      "loss": 0.3784,
+      "step": 7597
+    },
+    {
+      "epoch": 1.1626625860749809,
+      "grad_norm": 2.0404107507386264,
+      "learning_rate": 7.869119064994836e-06,
+      "loss": 0.4149,
+      "step": 7598
+    },
+    {
+      "epoch": 1.1628156082631982,
+      "grad_norm": 2.083732346909204,
+      "learning_rate": 7.866697863855439e-06,
+      "loss": 0.4429,
+      "step": 7599
+    },
+    {
+      "epoch": 1.1629686304514155,
+      "grad_norm": 2.3775803032947636,
+      "learning_rate": 7.864276793730468e-06,
+      "loss": 0.3781,
+      "step": 7600
+    },
+    {
+      "epoch": 1.1631216526396329,
+      "grad_norm": 2.115787174371097,
+      "learning_rate": 7.861855854768611e-06,
+      "loss": 0.3634,
+      "step": 7601
+    },
+    {
+      "epoch": 1.16327467482785,
+      "grad_norm": 2.203471890719275,
+      "learning_rate": 7.859435047118558e-06,
+      "loss": 0.407,
+      "step": 7602
+    },
+    {
+      "epoch": 1.1634276970160673,
+      "grad_norm": 2.121708542637144,
+      "learning_rate": 7.857014370928968e-06,
+      "loss": 0.4098,
+      "step": 7603
+    },
+    {
+      "epoch": 1.1635807192042846,
+      "grad_norm": 2.324513780877982,
+      "learning_rate": 7.854593826348506e-06,
+      "loss": 0.4553,
+      "step": 7604
+    },
+    {
+      "epoch": 1.163733741392502,
+      "grad_norm": 2.1593388977675976,
+      "learning_rate": 7.852173413525837e-06,
+      "loss": 0.3437,
+      "step": 7605
+    },
+    {
+      "epoch": 1.1638867635807193,
+      "grad_norm": 2.2737818262158362,
+      "learning_rate": 7.849753132609595e-06,
+      "loss": 0.4773,
+      "step": 7606
+    },
+    {
+      "epoch": 1.1640397857689364,
+      "grad_norm": 2.4196407421903534,
+      "learning_rate": 7.847332983748427e-06,
+      "loss": 0.4916,
+      "step": 7607
+    },
+    {
+      "epoch": 1.1641928079571537,
+      "grad_norm": 2.2265966790800853,
+      "learning_rate": 7.844912967090965e-06,
+      "loss": 0.3524,
+      "step": 7608
+    },
+    {
+      "epoch": 1.164345830145371,
+      "grad_norm": 2.300774220942024,
+      "learning_rate": 7.842493082785823e-06,
+      "loss": 0.4525,
+      "step": 7609
+    },
+    {
+      "epoch": 1.1644988523335884,
+      "grad_norm": 2.1765958482202246,
+      "learning_rate": 7.840073330981623e-06,
+      "loss": 0.4291,
+      "step": 7610
+    },
+    {
+      "epoch": 1.1646518745218057,
+      "grad_norm": 2.545845268184133,
+      "learning_rate": 7.837653711826973e-06,
+      "loss": 0.4652,
+      "step": 7611
+    },
+    {
+      "epoch": 1.1648048967100229,
+      "grad_norm": 2.2236738456132437,
+      "learning_rate": 7.835234225470462e-06,
+      "loss": 0.4389,
+      "step": 7612
+    },
+    {
+      "epoch": 1.1649579188982402,
+      "grad_norm": 2.161295283860229,
+      "learning_rate": 7.832814872060688e-06,
+      "loss": 0.429,
+      "step": 7613
+    },
+    {
+      "epoch": 1.1651109410864575,
+      "grad_norm": 2.267543706470478,
+      "learning_rate": 7.83039565174623e-06,
+      "loss": 0.4238,
+      "step": 7614
+    },
+    {
+      "epoch": 1.1652639632746749,
+      "grad_norm": 2.096609865562028,
+      "learning_rate": 7.827976564675666e-06,
+      "loss": 0.4182,
+      "step": 7615
+    },
+    {
+      "epoch": 1.1654169854628922,
+      "grad_norm": 1.9474587819076785,
+      "learning_rate": 7.825557610997556e-06,
+      "loss": 0.4151,
+      "step": 7616
+    },
+    {
+      "epoch": 1.1655700076511093,
+      "grad_norm": 2.231859405584811,
+      "learning_rate": 7.823138790860457e-06,
+      "loss": 0.4394,
+      "step": 7617
+    },
+    {
+      "epoch": 1.1657230298393266,
+      "grad_norm": 2.1358801598651036,
+      "learning_rate": 7.820720104412926e-06,
+      "loss": 0.4202,
+      "step": 7618
+    },
+    {
+      "epoch": 1.165876052027544,
+      "grad_norm": 2.273267880308986,
+      "learning_rate": 7.818301551803495e-06,
+      "loss": 0.4307,
+      "step": 7619
+    },
+    {
+      "epoch": 1.1660290742157613,
+      "grad_norm": 2.2176807501043183,
+      "learning_rate": 7.8158831331807e-06,
+      "loss": 0.3873,
+      "step": 7620
+    },
+    {
+      "epoch": 1.1661820964039786,
+      "grad_norm": 2.164388307223667,
+      "learning_rate": 7.813464848693071e-06,
+      "loss": 0.418,
+      "step": 7621
+    },
+    {
+      "epoch": 1.166335118592196,
+      "grad_norm": 2.3247978450960405,
+      "learning_rate": 7.81104669848911e-06,
+      "loss": 0.4521,
+      "step": 7622
+    },
+    {
+      "epoch": 1.166488140780413,
+      "grad_norm": 2.1664042575123195,
+      "learning_rate": 7.808628682717344e-06,
+      "loss": 0.4207,
+      "step": 7623
+    },
+    {
+      "epoch": 1.1666411629686304,
+      "grad_norm": 2.0898246135348004,
+      "learning_rate": 7.806210801526262e-06,
+      "loss": 0.3832,
+      "step": 7624
+    },
+    {
+      "epoch": 1.1667941851568477,
+      "grad_norm": 2.0408662892538523,
+      "learning_rate": 7.80379305506435e-06,
+      "loss": 0.3524,
+      "step": 7625
+    },
+    {
+      "epoch": 1.166947207345065,
+      "grad_norm": 2.2001590021513397,
+      "learning_rate": 7.801375443480106e-06,
+      "loss": 0.4377,
+      "step": 7626
+    },
+    {
+      "epoch": 1.1671002295332824,
+      "grad_norm": 2.35247171593286,
+      "learning_rate": 7.798957966921992e-06,
+      "loss": 0.4511,
+      "step": 7627
+    },
+    {
+      "epoch": 1.1672532517214997,
+      "grad_norm": 2.093124942157242,
+      "learning_rate": 7.796540625538482e-06,
+      "loss": 0.3984,
+      "step": 7628
+    },
+    {
+      "epoch": 1.1674062739097169,
+      "grad_norm": 2.3270121913506743,
+      "learning_rate": 7.794123419478034e-06,
+      "loss": 0.4626,
+      "step": 7629
+    },
+    {
+      "epoch": 1.1675592960979342,
+      "grad_norm": 2.235583289520111,
+      "learning_rate": 7.791706348889092e-06,
+      "loss": 0.421,
+      "step": 7630
+    },
+    {
+      "epoch": 1.1677123182861515,
+      "grad_norm": 2.2800471006152336,
+      "learning_rate": 7.789289413920104e-06,
+      "loss": 0.4149,
+      "step": 7631
+    },
+    {
+      "epoch": 1.1678653404743689,
+      "grad_norm": 2.11975063545846,
+      "learning_rate": 7.786872614719504e-06,
+      "loss": 0.4212,
+      "step": 7632
+    },
+    {
+      "epoch": 1.1680183626625862,
+      "grad_norm": 1.976637872464587,
+      "learning_rate": 7.78445595143571e-06,
+      "loss": 0.3641,
+      "step": 7633
+    },
+    {
+      "epoch": 1.1681713848508033,
+      "grad_norm": 2.5678946494338426,
+      "learning_rate": 7.782039424217145e-06,
+      "loss": 0.3952,
+      "step": 7634
+    },
+    {
+      "epoch": 1.1683244070390206,
+      "grad_norm": 2.4311254543526712,
+      "learning_rate": 7.779623033212213e-06,
+      "loss": 0.5068,
+      "step": 7635
+    },
+    {
+      "epoch": 1.168477429227238,
+      "grad_norm": 2.125309875951746,
+      "learning_rate": 7.777206778569323e-06,
+      "loss": 0.39,
+      "step": 7636
+    },
+    {
+      "epoch": 1.1686304514154553,
+      "grad_norm": 1.9029565175583325,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.3747,
+      "step": 7637
+    },
+    {
+      "epoch": 1.1687834736036726,
+      "grad_norm": 1.8974621969583034,
+      "learning_rate": 7.772374678963199e-06,
+      "loss": 0.3954,
+      "step": 7638
+    },
+    {
+      "epoch": 1.1689364957918897,
+      "grad_norm": 2.1731954934867397,
+      "learning_rate": 7.769958834296733e-06,
+      "loss": 0.4806,
+      "step": 7639
+    },
+    {
+      "epoch": 1.169089517980107,
+      "grad_norm": 2.164959683654725,
+      "learning_rate": 7.767543126585815e-06,
+      "loss": 0.3921,
+      "step": 7640
+    },
+    {
+      "epoch": 1.1692425401683244,
+      "grad_norm": 2.0519145883295917,
+      "learning_rate": 7.765127555978805e-06,
+      "loss": 0.3916,
+      "step": 7641
+    },
+    {
+      "epoch": 1.1693955623565417,
+      "grad_norm": 2.022149191067335,
+      "learning_rate": 7.762712122624059e-06,
+      "loss": 0.5355,
+      "step": 7642
+    },
+    {
+      "epoch": 1.169548584544759,
+      "grad_norm": 2.2553693306957934,
+      "learning_rate": 7.76029682666991e-06,
+      "loss": 0.4656,
+      "step": 7643
+    },
+    {
+      "epoch": 1.1697016067329762,
+      "grad_norm": 2.1677346659488395,
+      "learning_rate": 7.757881668264696e-06,
+      "loss": 0.4061,
+      "step": 7644
+    },
+    {
+      "epoch": 1.1698546289211935,
+      "grad_norm": 2.2303487915393183,
+      "learning_rate": 7.755466647556742e-06,
+      "loss": 0.4142,
+      "step": 7645
+    },
+    {
+      "epoch": 1.1700076511094109,
+      "grad_norm": 2.2087393847961745,
+      "learning_rate": 7.753051764694358e-06,
+      "loss": 0.4007,
+      "step": 7646
+    },
+    {
+      "epoch": 1.1701606732976282,
+      "grad_norm": 2.1866091937779326,
+      "learning_rate": 7.750637019825858e-06,
+      "loss": 0.3785,
+      "step": 7647
+    },
+    {
+      "epoch": 1.1703136954858455,
+      "grad_norm": 2.333612890088411,
+      "learning_rate": 7.74822241309954e-06,
+      "loss": 0.488,
+      "step": 7648
+    },
+    {
+      "epoch": 1.1704667176740626,
+      "grad_norm": 2.152719088963061,
+      "learning_rate": 7.745807944663688e-06,
+      "loss": 0.421,
+      "step": 7649
+    },
+    {
+      "epoch": 1.17061973986228,
+      "grad_norm": 2.322270994926128,
+      "learning_rate": 7.74339361466659e-06,
+      "loss": 0.4393,
+      "step": 7650
+    },
+    {
+      "epoch": 1.1707727620504973,
+      "grad_norm": 2.195968124463279,
+      "learning_rate": 7.740979423256518e-06,
+      "loss": 0.4031,
+      "step": 7651
+    },
+    {
+      "epoch": 1.1709257842387146,
+      "grad_norm": 2.136519735206509,
+      "learning_rate": 7.738565370581739e-06,
+      "loss": 0.3473,
+      "step": 7652
+    },
+    {
+      "epoch": 1.171078806426932,
+      "grad_norm": 2.2023901399428305,
+      "learning_rate": 7.736151456790506e-06,
+      "loss": 0.4934,
+      "step": 7653
+    },
+    {
+      "epoch": 1.171231828615149,
+      "grad_norm": 2.2416044379867994,
+      "learning_rate": 7.733737682031066e-06,
+      "loss": 0.4688,
+      "step": 7654
+    },
+    {
+      "epoch": 1.1713848508033664,
+      "grad_norm": 1.9033203376997059,
+      "learning_rate": 7.731324046451665e-06,
+      "loss": 0.4178,
+      "step": 7655
+    },
+    {
+      "epoch": 1.1715378729915837,
+      "grad_norm": 2.007408800033132,
+      "learning_rate": 7.728910550200528e-06,
+      "loss": 0.4644,
+      "step": 7656
+    },
+    {
+      "epoch": 1.171690895179801,
+      "grad_norm": 2.187477421883523,
+      "learning_rate": 7.726497193425875e-06,
+      "loss": 0.4285,
+      "step": 7657
+    },
+    {
+      "epoch": 1.1718439173680184,
+      "grad_norm": 2.3383291826778363,
+      "learning_rate": 7.72408397627593e-06,
+      "loss": 0.4946,
+      "step": 7658
+    },
+    {
+      "epoch": 1.1719969395562357,
+      "grad_norm": 2.097809814469768,
+      "learning_rate": 7.721670898898886e-06,
+      "loss": 0.4198,
+      "step": 7659
+    },
+    {
+      "epoch": 1.1721499617444529,
+      "grad_norm": 2.134611181161027,
+      "learning_rate": 7.719257961442946e-06,
+      "loss": 0.3677,
+      "step": 7660
+    },
+    {
+      "epoch": 1.1723029839326702,
+      "grad_norm": 2.0635524558988165,
+      "learning_rate": 7.716845164056301e-06,
+      "loss": 0.3861,
+      "step": 7661
+    },
+    {
+      "epoch": 1.1724560061208875,
+      "grad_norm": 2.1668260879700285,
+      "learning_rate": 7.714432506887119e-06,
+      "loss": 0.3923,
+      "step": 7662
+    },
+    {
+      "epoch": 1.1726090283091049,
+      "grad_norm": 2.165575599965999,
+      "learning_rate": 7.712019990083583e-06,
+      "loss": 0.3705,
+      "step": 7663
+    },
+    {
+      "epoch": 1.1727620504973222,
+      "grad_norm": 2.3510883434857717,
+      "learning_rate": 7.709607613793847e-06,
+      "loss": 0.4414,
+      "step": 7664
+    },
+    {
+      "epoch": 1.1729150726855395,
+      "grad_norm": 2.4418747152055733,
+      "learning_rate": 7.707195378166071e-06,
+      "loss": 0.4232,
+      "step": 7665
+    },
+    {
+      "epoch": 1.1730680948737566,
+      "grad_norm": 1.8774606626370103,
+      "learning_rate": 7.704783283348396e-06,
+      "loss": 0.3496,
+      "step": 7666
+    },
+    {
+      "epoch": 1.173221117061974,
+      "grad_norm": 2.4097174796880125,
+      "learning_rate": 7.702371329488954e-06,
+      "loss": 0.402,
+      "step": 7667
+    },
+    {
+      "epoch": 1.1733741392501913,
+      "grad_norm": 2.141900723164932,
+      "learning_rate": 7.699959516735884e-06,
+      "loss": 0.4609,
+      "step": 7668
+    },
+    {
+      "epoch": 1.1735271614384086,
+      "grad_norm": 1.9648011311206592,
+      "learning_rate": 7.697547845237294e-06,
+      "loss": 0.307,
+      "step": 7669
+    },
+    {
+      "epoch": 1.173680183626626,
+      "grad_norm": 2.014416906522851,
+      "learning_rate": 7.695136315141295e-06,
+      "loss": 0.4412,
+      "step": 7670
+    },
+    {
+      "epoch": 1.173833205814843,
+      "grad_norm": 1.9892079843533914,
+      "learning_rate": 7.692724926595998e-06,
+      "loss": 0.386,
+      "step": 7671
+    },
+    {
+      "epoch": 1.1739862280030604,
+      "grad_norm": 2.123848925579515,
+      "learning_rate": 7.690313679749484e-06,
+      "loss": 0.4088,
+      "step": 7672
+    },
+    {
+      "epoch": 1.1741392501912777,
+      "grad_norm": 2.4469979298476146,
+      "learning_rate": 7.687902574749844e-06,
+      "loss": 0.538,
+      "step": 7673
+    },
+    {
+      "epoch": 1.174292272379495,
+      "grad_norm": 2.1653847556064414,
+      "learning_rate": 7.685491611745155e-06,
+      "loss": 0.4599,
+      "step": 7674
+    },
+    {
+      "epoch": 1.1744452945677124,
+      "grad_norm": 2.1231595403483468,
+      "learning_rate": 7.683080790883477e-06,
+      "loss": 0.4341,
+      "step": 7675
+    },
+    {
+      "epoch": 1.1745983167559295,
+      "grad_norm": 2.180437643743477,
+      "learning_rate": 7.680670112312871e-06,
+      "loss": 0.387,
+      "step": 7676
+    },
+    {
+      "epoch": 1.1747513389441469,
+      "grad_norm": 2.1797587090293455,
+      "learning_rate": 7.678259576181391e-06,
+      "loss": 0.4162,
+      "step": 7677
+    },
+    {
+      "epoch": 1.1749043611323642,
+      "grad_norm": 1.9300874158917527,
+      "learning_rate": 7.675849182637069e-06,
+      "loss": 0.3977,
+      "step": 7678
+    },
+    {
+      "epoch": 1.1750573833205815,
+      "grad_norm": 2.176277192004729,
+      "learning_rate": 7.67343893182794e-06,
+      "loss": 0.3371,
+      "step": 7679
+    },
+    {
+      "epoch": 1.1752104055087988,
+      "grad_norm": 2.1934070557560537,
+      "learning_rate": 7.671028823902029e-06,
+      "loss": 0.4144,
+      "step": 7680
+    },
+    {
+      "epoch": 1.175363427697016,
+      "grad_norm": 2.2349384227823914,
+      "learning_rate": 7.668618859007351e-06,
+      "loss": 0.3794,
+      "step": 7681
+    },
+    {
+      "epoch": 1.1755164498852333,
+      "grad_norm": 2.2353031611686416,
+      "learning_rate": 7.666209037291909e-06,
+      "loss": 0.439,
+      "step": 7682
+    },
+    {
+      "epoch": 1.1756694720734506,
+      "grad_norm": 2.511919213185073,
+      "learning_rate": 7.663799358903698e-06,
+      "loss": 0.455,
+      "step": 7683
+    },
+    {
+      "epoch": 1.175822494261668,
+      "grad_norm": 2.0920595984190626,
+      "learning_rate": 7.661389823990711e-06,
+      "loss": 0.3794,
+      "step": 7684
+    },
+    {
+      "epoch": 1.1759755164498853,
+      "grad_norm": 2.247879750828007,
+      "learning_rate": 7.65898043270092e-06,
+      "loss": 0.4846,
+      "step": 7685
+    },
+    {
+      "epoch": 1.1761285386381024,
+      "grad_norm": 2.1945571365625653,
+      "learning_rate": 7.656571185182298e-06,
+      "loss": 0.4249,
+      "step": 7686
+    },
+    {
+      "epoch": 1.1762815608263197,
+      "grad_norm": 2.043387306270332,
+      "learning_rate": 7.654162081582812e-06,
+      "loss": 0.3923,
+      "step": 7687
+    },
+    {
+      "epoch": 1.176434583014537,
+      "grad_norm": 2.1797498439512193,
+      "learning_rate": 7.651753122050404e-06,
+      "loss": 0.4203,
+      "step": 7688
+    },
+    {
+      "epoch": 1.1765876052027544,
+      "grad_norm": 1.9302525156159196,
+      "learning_rate": 7.649344306733026e-06,
+      "loss": 0.3811,
+      "step": 7689
+    },
+    {
+      "epoch": 1.1767406273909717,
+      "grad_norm": 2.3062243258549735,
+      "learning_rate": 7.646935635778612e-06,
+      "loss": 0.4832,
+      "step": 7690
+    },
+    {
+      "epoch": 1.176893649579189,
+      "grad_norm": 2.2193174384523315,
+      "learning_rate": 7.64452710933508e-06,
+      "loss": 0.4152,
+      "step": 7691
+    },
+    {
+      "epoch": 1.1770466717674062,
+      "grad_norm": 2.1635717890959754,
+      "learning_rate": 7.642118727550358e-06,
+      "loss": 0.3953,
+      "step": 7692
+    },
+    {
+      "epoch": 1.1771996939556235,
+      "grad_norm": 1.9823843125750913,
+      "learning_rate": 7.639710490572348e-06,
+      "loss": 0.3899,
+      "step": 7693
+    },
+    {
+      "epoch": 1.1773527161438408,
+      "grad_norm": 2.0966731774522684,
+      "learning_rate": 7.637302398548949e-06,
+      "loss": 0.3648,
+      "step": 7694
+    },
+    {
+      "epoch": 1.1775057383320582,
+      "grad_norm": 2.159828350633099,
+      "learning_rate": 7.634894451628053e-06,
+      "loss": 0.4285,
+      "step": 7695
+    },
+    {
+      "epoch": 1.1776587605202755,
+      "grad_norm": 2.314588251676099,
+      "learning_rate": 7.632486649957539e-06,
+      "loss": 0.4209,
+      "step": 7696
+    },
+    {
+      "epoch": 1.1778117827084928,
+      "grad_norm": 2.234225308705491,
+      "learning_rate": 7.630078993685286e-06,
+      "loss": 0.3981,
+      "step": 7697
+    },
+    {
+      "epoch": 1.17796480489671,
+      "grad_norm": 2.360790111872005,
+      "learning_rate": 7.627671482959152e-06,
+      "loss": 0.4557,
+      "step": 7698
+    },
+    {
+      "epoch": 1.1781178270849273,
+      "grad_norm": 2.206115497805982,
+      "learning_rate": 7.625264117926989e-06,
+      "loss": 0.4567,
+      "step": 7699
+    },
+    {
+      "epoch": 1.1782708492731446,
+      "grad_norm": 2.7009488220205395,
+      "learning_rate": 7.622856898736652e-06,
+      "loss": 0.3969,
+      "step": 7700
+    },
+    {
+      "epoch": 1.178423871461362,
+      "grad_norm": 1.9240114837643656,
+      "learning_rate": 7.6204498255359674e-06,
+      "loss": 0.4017,
+      "step": 7701
+    },
+    {
+      "epoch": 1.1785768936495793,
+      "grad_norm": 2.0628563892650558,
+      "learning_rate": 7.618042898472771e-06,
+      "loss": 0.4437,
+      "step": 7702
+    },
+    {
+      "epoch": 1.1787299158377964,
+      "grad_norm": 1.8012689942102933,
+      "learning_rate": 7.6156361176948804e-06,
+      "loss": 0.3697,
+      "step": 7703
+    },
+    {
+      "epoch": 1.1788829380260137,
+      "grad_norm": 2.22070307433343,
+      "learning_rate": 7.613229483350095e-06,
+      "loss": 0.4171,
+      "step": 7704
+    },
+    {
+      "epoch": 1.179035960214231,
+      "grad_norm": 2.312836250788219,
+      "learning_rate": 7.610822995586234e-06,
+      "loss": 0.4379,
+      "step": 7705
+    },
+    {
+      "epoch": 1.1791889824024484,
+      "grad_norm": 2.2321765546942665,
+      "learning_rate": 7.608416654551077e-06,
+      "loss": 0.3946,
+      "step": 7706
+    },
+    {
+      "epoch": 1.1793420045906657,
+      "grad_norm": 2.1449678132243677,
+      "learning_rate": 7.606010460392402e-06,
+      "loss": 0.3959,
+      "step": 7707
+    },
+    {
+      "epoch": 1.1794950267788828,
+      "grad_norm": 2.1178482207877836,
+      "learning_rate": 7.603604413257998e-06,
+      "loss": 0.4641,
+      "step": 7708
+    },
+    {
+      "epoch": 1.1796480489671002,
+      "grad_norm": 2.277472552097201,
+      "learning_rate": 7.601198513295618e-06,
+      "loss": 0.4133,
+      "step": 7709
+    },
+    {
+      "epoch": 1.1798010711553175,
+      "grad_norm": 2.3060438068546674,
+      "learning_rate": 7.598792760653022e-06,
+      "loss": 0.3717,
+      "step": 7710
+    },
+    {
+      "epoch": 1.1799540933435348,
+      "grad_norm": 2.212645777195339,
+      "learning_rate": 7.5963871554779586e-06,
+      "loss": 0.4233,
+      "step": 7711
+    },
+    {
+      "epoch": 1.1801071155317522,
+      "grad_norm": 2.288782015626369,
+      "learning_rate": 7.593981697918159e-06,
+      "loss": 0.3845,
+      "step": 7712
+    },
+    {
+      "epoch": 1.1802601377199693,
+      "grad_norm": 1.9850372928797795,
+      "learning_rate": 7.5915763881213576e-06,
+      "loss": 0.3853,
+      "step": 7713
+    },
+    {
+      "epoch": 1.1804131599081866,
+      "grad_norm": 2.1714931935713127,
+      "learning_rate": 7.589171226235276e-06,
+      "loss": 0.3907,
+      "step": 7714
+    },
+    {
+      "epoch": 1.180566182096404,
+      "grad_norm": 1.9973204967252072,
+      "learning_rate": 7.586766212407615e-06,
+      "loss": 0.3507,
+      "step": 7715
+    },
+    {
+      "epoch": 1.1807192042846213,
+      "grad_norm": 2.041853555787875,
+      "learning_rate": 7.584361346786082e-06,
+      "loss": 0.4383,
+      "step": 7716
+    },
+    {
+      "epoch": 1.1808722264728386,
+      "grad_norm": 2.0279848709164843,
+      "learning_rate": 7.581956629518369e-06,
+      "loss": 0.4229,
+      "step": 7717
+    },
+    {
+      "epoch": 1.1810252486610557,
+      "grad_norm": 2.06147392563384,
+      "learning_rate": 7.579552060752162e-06,
+      "loss": 0.4688,
+      "step": 7718
+    },
+    {
+      "epoch": 1.181178270849273,
+      "grad_norm": 2.257617092359455,
+      "learning_rate": 7.57714764063513e-06,
+      "loss": 0.4485,
+      "step": 7719
+    },
+    {
+      "epoch": 1.1813312930374904,
+      "grad_norm": 2.089931676977981,
+      "learning_rate": 7.574743369314937e-06,
+      "loss": 0.4228,
+      "step": 7720
+    },
+    {
+      "epoch": 1.1814843152257077,
+      "grad_norm": 2.018731297347733,
+      "learning_rate": 7.572339246939247e-06,
+      "loss": 0.3592,
+      "step": 7721
+    },
+    {
+      "epoch": 1.181637337413925,
+      "grad_norm": 1.9133562808684277,
+      "learning_rate": 7.569935273655696e-06,
+      "loss": 0.349,
+      "step": 7722
+    },
+    {
+      "epoch": 1.1817903596021424,
+      "grad_norm": 2.3225293651885828,
+      "learning_rate": 7.567531449611927e-06,
+      "loss": 0.3854,
+      "step": 7723
+    },
+    {
+      "epoch": 1.1819433817903595,
+      "grad_norm": 2.080329898307183,
+      "learning_rate": 7.565127774955571e-06,
+      "loss": 0.4338,
+      "step": 7724
+    },
+    {
+      "epoch": 1.1820964039785768,
+      "grad_norm": 2.127662122852471,
+      "learning_rate": 7.562724249834239e-06,
+      "loss": 0.4683,
+      "step": 7725
+    },
+    {
+      "epoch": 1.1822494261667942,
+      "grad_norm": 2.1114355637464133,
+      "learning_rate": 7.560320874395547e-06,
+      "loss": 0.3616,
+      "step": 7726
+    },
+    {
+      "epoch": 1.1824024483550115,
+      "grad_norm": 2.3223541842836237,
+      "learning_rate": 7.557917648787097e-06,
+      "loss": 0.4541,
+      "step": 7727
+    },
+    {
+      "epoch": 1.1825554705432288,
+      "grad_norm": 2.153187629446042,
+      "learning_rate": 7.555514573156474e-06,
+      "loss": 0.4665,
+      "step": 7728
+    },
+    {
+      "epoch": 1.1827084927314462,
+      "grad_norm": 2.115658998412094,
+      "learning_rate": 7.553111647651266e-06,
+      "loss": 0.3996,
+      "step": 7729
+    },
+    {
+      "epoch": 1.1828615149196633,
+      "grad_norm": 2.293884325074409,
+      "learning_rate": 7.5507088724190445e-06,
+      "loss": 0.4079,
+      "step": 7730
+    },
+    {
+      "epoch": 1.1830145371078806,
+      "grad_norm": 2.3765090288857658,
+      "learning_rate": 7.54830624760737e-06,
+      "loss": 0.4425,
+      "step": 7731
+    },
+    {
+      "epoch": 1.183167559296098,
+      "grad_norm": 1.9207125051904854,
+      "learning_rate": 7.5459037733638016e-06,
+      "loss": 0.3427,
+      "step": 7732
+    },
+    {
+      "epoch": 1.1833205814843153,
+      "grad_norm": 2.115216638237201,
+      "learning_rate": 7.54350144983588e-06,
+      "loss": 0.4216,
+      "step": 7733
+    },
+    {
+      "epoch": 1.1834736036725326,
+      "grad_norm": 2.072322814839418,
+      "learning_rate": 7.54109927717115e-06,
+      "loss": 0.4091,
+      "step": 7734
+    },
+    {
+      "epoch": 1.1836266258607497,
+      "grad_norm": 1.972905904780838,
+      "learning_rate": 7.53869725551713e-06,
+      "loss": 0.3766,
+      "step": 7735
+    },
+    {
+      "epoch": 1.183779648048967,
+      "grad_norm": 2.169516411743189,
+      "learning_rate": 7.536295385021338e-06,
+      "loss": 0.3662,
+      "step": 7736
+    },
+    {
+      "epoch": 1.1839326702371844,
+      "grad_norm": 2.0957775433273538,
+      "learning_rate": 7.5338936658312885e-06,
+      "loss": 0.398,
+      "step": 7737
+    },
+    {
+      "epoch": 1.1840856924254017,
+      "grad_norm": 2.1343320274405793,
+      "learning_rate": 7.531492098094473e-06,
+      "loss": 0.4014,
+      "step": 7738
+    },
+    {
+      "epoch": 1.184238714613619,
+      "grad_norm": 2.378363664272812,
+      "learning_rate": 7.5290906819583865e-06,
+      "loss": 0.4425,
+      "step": 7739
+    },
+    {
+      "epoch": 1.1843917368018362,
+      "grad_norm": 2.2596488259073535,
+      "learning_rate": 7.52668941757051e-06,
+      "loss": 0.4086,
+      "step": 7740
+    },
+    {
+      "epoch": 1.1845447589900535,
+      "grad_norm": 2.2518267482599135,
+      "learning_rate": 7.524288305078307e-06,
+      "loss": 0.3888,
+      "step": 7741
+    },
+    {
+      "epoch": 1.1846977811782708,
+      "grad_norm": 2.2319015517268386,
+      "learning_rate": 7.521887344629246e-06,
+      "loss": 0.4249,
+      "step": 7742
+    },
+    {
+      "epoch": 1.1848508033664882,
+      "grad_norm": 2.099102832105148,
+      "learning_rate": 7.519486536370779e-06,
+      "loss": 0.3981,
+      "step": 7743
+    },
+    {
+      "epoch": 1.1850038255547055,
+      "grad_norm": 2.326704059378749,
+      "learning_rate": 7.517085880450345e-06,
+      "loss": 0.4581,
+      "step": 7744
+    },
+    {
+      "epoch": 1.1851568477429226,
+      "grad_norm": 2.3914827027493297,
+      "learning_rate": 7.514685377015383e-06,
+      "loss": 0.394,
+      "step": 7745
+    },
+    {
+      "epoch": 1.18530986993114,
+      "grad_norm": 1.8983373970935058,
+      "learning_rate": 7.512285026213311e-06,
+      "loss": 0.3382,
+      "step": 7746
+    },
+    {
+      "epoch": 1.1854628921193573,
+      "grad_norm": 2.3908709484472523,
+      "learning_rate": 7.509884828191552e-06,
+      "loss": 0.4461,
+      "step": 7747
+    },
+    {
+      "epoch": 1.1856159143075746,
+      "grad_norm": 2.483737125826371,
+      "learning_rate": 7.5074847830975054e-06,
+      "loss": 0.4697,
+      "step": 7748
+    },
+    {
+      "epoch": 1.185768936495792,
+      "grad_norm": 2.1147090426635224,
+      "learning_rate": 7.505084891078566e-06,
+      "loss": 0.3533,
+      "step": 7749
+    },
+    {
+      "epoch": 1.185921958684009,
+      "grad_norm": 2.1135289311255407,
+      "learning_rate": 7.5026851522821286e-06,
+      "loss": 0.4444,
+      "step": 7750
+    },
+    {
+      "epoch": 1.1860749808722264,
+      "grad_norm": 2.1675730484291864,
+      "learning_rate": 7.500285566855564e-06,
+      "loss": 0.3763,
+      "step": 7751
+    },
+    {
+      "epoch": 1.1862280030604437,
+      "grad_norm": 2.1575555766890706,
+      "learning_rate": 7.497886134946238e-06,
+      "loss": 0.4758,
+      "step": 7752
+    },
+    {
+      "epoch": 1.186381025248661,
+      "grad_norm": 1.7799980041205554,
+      "learning_rate": 7.49548685670152e-06,
+      "loss": 0.3235,
+      "step": 7753
+    },
+    {
+      "epoch": 1.1865340474368784,
+      "grad_norm": 2.344961863500604,
+      "learning_rate": 7.493087732268744e-06,
+      "loss": 0.4517,
+      "step": 7754
+    },
+    {
+      "epoch": 1.1866870696250955,
+      "grad_norm": 2.3828288288807684,
+      "learning_rate": 7.490688761795262e-06,
+      "loss": 0.4251,
+      "step": 7755
+    },
+    {
+      "epoch": 1.1868400918133128,
+      "grad_norm": 2.224048831830438,
+      "learning_rate": 7.4882899454284e-06,
+      "loss": 0.4147,
+      "step": 7756
+    },
+    {
+      "epoch": 1.1869931140015302,
+      "grad_norm": 2.3063506729017655,
+      "learning_rate": 7.485891283315475e-06,
+      "loss": 0.3875,
+      "step": 7757
+    },
+    {
+      "epoch": 1.1871461361897475,
+      "grad_norm": 2.369613366127151,
+      "learning_rate": 7.483492775603803e-06,
+      "loss": 0.4198,
+      "step": 7758
+    },
+    {
+      "epoch": 1.1872991583779648,
+      "grad_norm": 2.2857022695312095,
+      "learning_rate": 7.481094422440688e-06,
+      "loss": 0.4651,
+      "step": 7759
+    },
+    {
+      "epoch": 1.1874521805661822,
+      "grad_norm": 1.9585673421612289,
+      "learning_rate": 7.478696223973413e-06,
+      "loss": 0.3196,
+      "step": 7760
+    },
+    {
+      "epoch": 1.1876052027543993,
+      "grad_norm": 2.103682155279868,
+      "learning_rate": 7.476298180349269e-06,
+      "loss": 0.403,
+      "step": 7761
+    },
+    {
+      "epoch": 1.1877582249426166,
+      "grad_norm": 1.8589192664213134,
+      "learning_rate": 7.4739002917155235e-06,
+      "loss": 0.3739,
+      "step": 7762
+    },
+    {
+      "epoch": 1.187911247130834,
+      "grad_norm": 2.2039683638286203,
+      "learning_rate": 7.471502558219448e-06,
+      "loss": 0.4164,
+      "step": 7763
+    },
+    {
+      "epoch": 1.1880642693190513,
+      "grad_norm": 2.285658684641684,
+      "learning_rate": 7.469104980008291e-06,
+      "loss": 0.4457,
+      "step": 7764
+    },
+    {
+      "epoch": 1.1882172915072686,
+      "grad_norm": 2.0870031789407077,
+      "learning_rate": 7.466707557229294e-06,
+      "loss": 0.4712,
+      "step": 7765
+    },
+    {
+      "epoch": 1.188370313695486,
+      "grad_norm": 1.981110909098681,
+      "learning_rate": 7.464310290029702e-06,
+      "loss": 0.442,
+      "step": 7766
+    },
+    {
+      "epoch": 1.188523335883703,
+      "grad_norm": 2.146376730190204,
+      "learning_rate": 7.461913178556731e-06,
+      "loss": 0.4094,
+      "step": 7767
+    },
+    {
+      "epoch": 1.1886763580719204,
+      "grad_norm": 2.1228310670875827,
+      "learning_rate": 7.4595162229576005e-06,
+      "loss": 0.425,
+      "step": 7768
+    },
+    {
+      "epoch": 1.1888293802601377,
+      "grad_norm": 2.0450434435134115,
+      "learning_rate": 7.45711942337952e-06,
+      "loss": 0.3992,
+      "step": 7769
+    },
+    {
+      "epoch": 1.188982402448355,
+      "grad_norm": 2.072140769338632,
+      "learning_rate": 7.45472277996968e-06,
+      "loss": 0.3567,
+      "step": 7770
+    },
+    {
+      "epoch": 1.1891354246365724,
+      "grad_norm": 2.1452549233035163,
+      "learning_rate": 7.452326292875273e-06,
+      "loss": 0.3996,
+      "step": 7771
+    },
+    {
+      "epoch": 1.1892884468247895,
+      "grad_norm": 2.0564488471141025,
+      "learning_rate": 7.449929962243478e-06,
+      "loss": 0.4003,
+      "step": 7772
+    },
+    {
+      "epoch": 1.1894414690130068,
+      "grad_norm": 2.2765367450207177,
+      "learning_rate": 7.447533788221454e-06,
+      "loss": 0.4523,
+      "step": 7773
+    },
+    {
+      "epoch": 1.1895944912012242,
+      "grad_norm": 2.474886870340549,
+      "learning_rate": 7.445137770956368e-06,
+      "loss": 0.4561,
+      "step": 7774
+    },
+    {
+      "epoch": 1.1897475133894415,
+      "grad_norm": 2.296110569783041,
+      "learning_rate": 7.442741910595365e-06,
+      "loss": 0.4152,
+      "step": 7775
+    },
+    {
+      "epoch": 1.1899005355776588,
+      "grad_norm": 2.3301560185251726,
+      "learning_rate": 7.440346207285589e-06,
+      "loss": 0.4871,
+      "step": 7776
+    },
+    {
+      "epoch": 1.190053557765876,
+      "grad_norm": 1.8039552048846907,
+      "learning_rate": 7.437950661174164e-06,
+      "loss": 0.3184,
+      "step": 7777
+    },
+    {
+      "epoch": 1.1902065799540933,
+      "grad_norm": 2.1926490325759,
+      "learning_rate": 7.435555272408208e-06,
+      "loss": 0.4312,
+      "step": 7778
+    },
+    {
+      "epoch": 1.1903596021423106,
+      "grad_norm": 2.104249637935987,
+      "learning_rate": 7.433160041134841e-06,
+      "loss": 0.3409,
+      "step": 7779
+    },
+    {
+      "epoch": 1.190512624330528,
+      "grad_norm": 2.20756794295377,
+      "learning_rate": 7.430764967501155e-06,
+      "loss": 0.3749,
+      "step": 7780
+    },
+    {
+      "epoch": 1.1906656465187453,
+      "grad_norm": 2.069229329976173,
+      "learning_rate": 7.428370051654241e-06,
+      "loss": 0.3903,
+      "step": 7781
+    },
+    {
+      "epoch": 1.1908186687069624,
+      "grad_norm": 2.2207924833034087,
+      "learning_rate": 7.425975293741186e-06,
+      "loss": 0.4389,
+      "step": 7782
+    },
+    {
+      "epoch": 1.1909716908951797,
+      "grad_norm": 2.2331530214251623,
+      "learning_rate": 7.4235806939090515e-06,
+      "loss": 0.4131,
+      "step": 7783
+    },
+    {
+      "epoch": 1.191124713083397,
+      "grad_norm": 2.2504109035986026,
+      "learning_rate": 7.421186252304913e-06,
+      "loss": 0.4731,
+      "step": 7784
+    },
+    {
+      "epoch": 1.1912777352716144,
+      "grad_norm": 2.371044724991984,
+      "learning_rate": 7.4187919690758145e-06,
+      "loss": 0.4377,
+      "step": 7785
+    },
+    {
+      "epoch": 1.1914307574598317,
+      "grad_norm": 2.242523834065211,
+      "learning_rate": 7.416397844368792e-06,
+      "loss": 0.3621,
+      "step": 7786
+    },
+    {
+      "epoch": 1.1915837796480488,
+      "grad_norm": 2.2945870940039668,
+      "learning_rate": 7.414003878330893e-06,
+      "loss": 0.3849,
+      "step": 7787
+    },
+    {
+      "epoch": 1.1917368018362662,
+      "grad_norm": 2.266870340168214,
+      "learning_rate": 7.411610071109131e-06,
+      "loss": 0.4038,
+      "step": 7788
+    },
+    {
+      "epoch": 1.1918898240244835,
+      "grad_norm": 2.1237321269660834,
+      "learning_rate": 7.409216422850514e-06,
+      "loss": 0.3927,
+      "step": 7789
+    },
+    {
+      "epoch": 1.1920428462127008,
+      "grad_norm": 2.3142228870616126,
+      "learning_rate": 7.406822933702058e-06,
+      "loss": 0.4434,
+      "step": 7790
+    },
+    {
+      "epoch": 1.1921958684009182,
+      "grad_norm": 2.0955604959527068,
+      "learning_rate": 7.404429603810747e-06,
+      "loss": 0.3222,
+      "step": 7791
+    },
+    {
+      "epoch": 1.1923488905891355,
+      "grad_norm": 1.9617282423466649,
+      "learning_rate": 7.402036433323569e-06,
+      "loss": 0.3517,
+      "step": 7792
+    },
+    {
+      "epoch": 1.1925019127773526,
+      "grad_norm": 1.9869255508557389,
+      "learning_rate": 7.399643422387499e-06,
+      "loss": 0.3596,
+      "step": 7793
+    },
+    {
+      "epoch": 1.19265493496557,
+      "grad_norm": 2.2481717873813367,
+      "learning_rate": 7.397250571149496e-06,
+      "loss": 0.4018,
+      "step": 7794
+    },
+    {
+      "epoch": 1.1928079571537873,
+      "grad_norm": 2.267844809299893,
+      "learning_rate": 7.3948578797565185e-06,
+      "loss": 0.4399,
+      "step": 7795
+    },
+    {
+      "epoch": 1.1929609793420046,
+      "grad_norm": 2.296245229262881,
+      "learning_rate": 7.392465348355512e-06,
+      "loss": 0.4375,
+      "step": 7796
+    },
+    {
+      "epoch": 1.193114001530222,
+      "grad_norm": 2.4275607536962585,
+      "learning_rate": 7.390072977093405e-06,
+      "loss": 0.4886,
+      "step": 7797
+    },
+    {
+      "epoch": 1.1932670237184393,
+      "grad_norm": 2.1878475452491633,
+      "learning_rate": 7.387680766117129e-06,
+      "loss": 0.4665,
+      "step": 7798
+    },
+    {
+      "epoch": 1.1934200459066564,
+      "grad_norm": 2.2456587752248445,
+      "learning_rate": 7.3852887155735955e-06,
+      "loss": 0.4437,
+      "step": 7799
+    },
+    {
+      "epoch": 1.1935730680948737,
+      "grad_norm": 2.395071968132589,
+      "learning_rate": 7.382896825609714e-06,
+      "loss": 0.456,
+      "step": 7800
+    },
+    {
+      "epoch": 1.193726090283091,
+      "grad_norm": 2.1758010578869738,
+      "learning_rate": 7.380505096372375e-06,
+      "loss": 0.3582,
+      "step": 7801
+    },
+    {
+      "epoch": 1.1938791124713084,
+      "grad_norm": 2.457008422878569,
+      "learning_rate": 7.3781135280084615e-06,
+      "loss": 0.4464,
+      "step": 7802
+    },
+    {
+      "epoch": 1.1940321346595257,
+      "grad_norm": 2.0253508595174394,
+      "learning_rate": 7.375722120664859e-06,
+      "loss": 0.3641,
+      "step": 7803
+    },
+    {
+      "epoch": 1.1941851568477428,
+      "grad_norm": 2.207183732033762,
+      "learning_rate": 7.373330874488422e-06,
+      "loss": 0.4679,
+      "step": 7804
+    },
+    {
+      "epoch": 1.1943381790359602,
+      "grad_norm": 1.9874369076730363,
+      "learning_rate": 7.370939789626016e-06,
+      "loss": 0.3426,
+      "step": 7805
+    },
+    {
+      "epoch": 1.1944912012241775,
+      "grad_norm": 2.2001806026700153,
+      "learning_rate": 7.368548866224483e-06,
+      "loss": 0.3841,
+      "step": 7806
+    },
+    {
+      "epoch": 1.1946442234123948,
+      "grad_norm": 1.9436191856311789,
+      "learning_rate": 7.366158104430654e-06,
+      "loss": 0.3105,
+      "step": 7807
+    },
+    {
+      "epoch": 1.1947972456006122,
+      "grad_norm": 2.0716591900676145,
+      "learning_rate": 7.363767504391362e-06,
+      "loss": 0.3567,
+      "step": 7808
+    },
+    {
+      "epoch": 1.1949502677888293,
+      "grad_norm": 2.3756768448765504,
+      "learning_rate": 7.361377066253424e-06,
+      "loss": 0.4912,
+      "step": 7809
+    },
+    {
+      "epoch": 1.1951032899770466,
+      "grad_norm": 2.044315279391537,
+      "learning_rate": 7.358986790163636e-06,
+      "loss": 0.3538,
+      "step": 7810
+    },
+    {
+      "epoch": 1.195256312165264,
+      "grad_norm": 1.9855034404666652,
+      "learning_rate": 7.356596676268804e-06,
+      "loss": 0.3895,
+      "step": 7811
+    },
+    {
+      "epoch": 1.1954093343534813,
+      "grad_norm": 2.315701745879983,
+      "learning_rate": 7.354206724715709e-06,
+      "loss": 0.4436,
+      "step": 7812
+    },
+    {
+      "epoch": 1.1955623565416986,
+      "grad_norm": 2.134487707383376,
+      "learning_rate": 7.3518169356511335e-06,
+      "loss": 0.3905,
+      "step": 7813
+    },
+    {
+      "epoch": 1.1957153787299157,
+      "grad_norm": 2.0187548298884543,
+      "learning_rate": 7.349427309221838e-06,
+      "loss": 0.3984,
+      "step": 7814
+    },
+    {
+      "epoch": 1.195868400918133,
+      "grad_norm": 2.271455573960681,
+      "learning_rate": 7.347037845574578e-06,
+      "loss": 0.3748,
+      "step": 7815
+    },
+    {
+      "epoch": 1.1960214231063504,
+      "grad_norm": 2.5486034536590987,
+      "learning_rate": 7.344648544856108e-06,
+      "loss": 0.5062,
+      "step": 7816
+    },
+    {
+      "epoch": 1.1961744452945677,
+      "grad_norm": 2.222378931229507,
+      "learning_rate": 7.342259407213155e-06,
+      "loss": 0.3106,
+      "step": 7817
+    },
+    {
+      "epoch": 1.196327467482785,
+      "grad_norm": 2.384115619572824,
+      "learning_rate": 7.339870432792448e-06,
+      "loss": 0.4188,
+      "step": 7818
+    },
+    {
+      "epoch": 1.1964804896710022,
+      "grad_norm": 2.4136145453264652,
+      "learning_rate": 7.337481621740707e-06,
+      "loss": 0.4704,
+      "step": 7819
+    },
+    {
+      "epoch": 1.1966335118592195,
+      "grad_norm": 2.081016144614866,
+      "learning_rate": 7.335092974204632e-06,
+      "loss": 0.409,
+      "step": 7820
+    },
+    {
+      "epoch": 1.1967865340474368,
+      "grad_norm": 2.0694517272109563,
+      "learning_rate": 7.332704490330924e-06,
+      "loss": 0.3886,
+      "step": 7821
+    },
+    {
+      "epoch": 1.1969395562356542,
+      "grad_norm": 2.015225672566741,
+      "learning_rate": 7.3303161702662715e-06,
+      "loss": 0.3692,
+      "step": 7822
+    },
+    {
+      "epoch": 1.1970925784238715,
+      "grad_norm": 2.503758938970734,
+      "learning_rate": 7.327928014157341e-06,
+      "loss": 0.4168,
+      "step": 7823
+    },
+    {
+      "epoch": 1.1972456006120888,
+      "grad_norm": 2.1923521871806626,
+      "learning_rate": 7.3255400221508076e-06,
+      "loss": 0.4147,
+      "step": 7824
+    },
+    {
+      "epoch": 1.197398622800306,
+      "grad_norm": 1.9996996439024073,
+      "learning_rate": 7.323152194393326e-06,
+      "loss": 0.424,
+      "step": 7825
+    },
+    {
+      "epoch": 1.1975516449885233,
+      "grad_norm": 2.3176090940404004,
+      "learning_rate": 7.320764531031535e-06,
+      "loss": 0.4398,
+      "step": 7826
+    },
+    {
+      "epoch": 1.1977046671767406,
+      "grad_norm": 2.0145284165200503,
+      "learning_rate": 7.318377032212078e-06,
+      "loss": 0.361,
+      "step": 7827
+    },
+    {
+      "epoch": 1.197857689364958,
+      "grad_norm": 1.818639838520028,
+      "learning_rate": 7.315989698081576e-06,
+      "loss": 0.3775,
+      "step": 7828
+    },
+    {
+      "epoch": 1.1980107115531753,
+      "grad_norm": 2.1098549855251676,
+      "learning_rate": 7.313602528786649e-06,
+      "loss": 0.3689,
+      "step": 7829
+    },
+    {
+      "epoch": 1.1981637337413926,
+      "grad_norm": 2.128136439868593,
+      "learning_rate": 7.3112155244739e-06,
+      "loss": 0.4344,
+      "step": 7830
+    },
+    {
+      "epoch": 1.1983167559296097,
+      "grad_norm": 2.4135696579714314,
+      "learning_rate": 7.3088286852899224e-06,
+      "loss": 0.4493,
+      "step": 7831
+    },
+    {
+      "epoch": 1.198469778117827,
+      "grad_norm": 2.0006879555739894,
+      "learning_rate": 7.306442011381307e-06,
+      "loss": 0.4525,
+      "step": 7832
+    },
+    {
+      "epoch": 1.1986228003060444,
+      "grad_norm": 2.051025939461601,
+      "learning_rate": 7.3040555028946225e-06,
+      "loss": 0.4063,
+      "step": 7833
+    },
+    {
+      "epoch": 1.1987758224942617,
+      "grad_norm": 2.034625467920058,
+      "learning_rate": 7.301669159976434e-06,
+      "loss": 0.4244,
+      "step": 7834
+    },
+    {
+      "epoch": 1.198928844682479,
+      "grad_norm": 2.0888212266302326,
+      "learning_rate": 7.299282982773301e-06,
+      "loss": 0.3443,
+      "step": 7835
+    },
+    {
+      "epoch": 1.1990818668706962,
+      "grad_norm": 1.9813426666478027,
+      "learning_rate": 7.296896971431764e-06,
+      "loss": 0.3808,
+      "step": 7836
+    },
+    {
+      "epoch": 1.1992348890589135,
+      "grad_norm": 2.1295621512831446,
+      "learning_rate": 7.2945111260983584e-06,
+      "loss": 0.3701,
+      "step": 7837
+    },
+    {
+      "epoch": 1.1993879112471308,
+      "grad_norm": 2.0436205396614646,
+      "learning_rate": 7.292125446919611e-06,
+      "loss": 0.3835,
+      "step": 7838
+    },
+    {
+      "epoch": 1.1995409334353482,
+      "grad_norm": 2.119557743892499,
+      "learning_rate": 7.289739934042031e-06,
+      "loss": 0.3901,
+      "step": 7839
+    },
+    {
+      "epoch": 1.1996939556235655,
+      "grad_norm": 2.300357163359858,
+      "learning_rate": 7.287354587612123e-06,
+      "loss": 0.424,
+      "step": 7840
+    },
+    {
+      "epoch": 1.1998469778117826,
+      "grad_norm": 2.085956552889289,
+      "learning_rate": 7.284969407776382e-06,
+      "loss": 0.3756,
+      "step": 7841
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3081369285656725,
+      "learning_rate": 7.282584394681294e-06,
+      "loss": 0.4268,
+      "step": 7842
+    },
+    {
+      "epoch": 1.2001530221882173,
+      "grad_norm": 2.353996381043388,
+      "learning_rate": 7.280199548473328e-06,
+      "loss": 0.4335,
+      "step": 7843
+    },
+    {
+      "epoch": 1.2003060443764346,
+      "grad_norm": 2.414828978611619,
+      "learning_rate": 7.277814869298945e-06,
+      "loss": 0.4901,
+      "step": 7844
+    },
+    {
+      "epoch": 1.200459066564652,
+      "grad_norm": 2.3718399813628452,
+      "learning_rate": 7.275430357304604e-06,
+      "loss": 0.4031,
+      "step": 7845
+    },
+    {
+      "epoch": 1.200612088752869,
+      "grad_norm": 2.2121027430213163,
+      "learning_rate": 7.273046012636742e-06,
+      "loss": 0.4049,
+      "step": 7846
+    },
+    {
+      "epoch": 1.2007651109410864,
+      "grad_norm": 2.0996973379693897,
+      "learning_rate": 7.270661835441789e-06,
+      "loss": 0.3518,
+      "step": 7847
+    },
+    {
+      "epoch": 1.2009181331293037,
+      "grad_norm": 2.3792243056952844,
+      "learning_rate": 7.268277825866175e-06,
+      "loss": 0.4506,
+      "step": 7848
+    },
+    {
+      "epoch": 1.201071155317521,
+      "grad_norm": 2.2470123496158294,
+      "learning_rate": 7.265893984056302e-06,
+      "loss": 0.4264,
+      "step": 7849
+    },
+    {
+      "epoch": 1.2012241775057384,
+      "grad_norm": 2.6193999301938273,
+      "learning_rate": 7.263510310158577e-06,
+      "loss": 0.4304,
+      "step": 7850
+    },
+    {
+      "epoch": 1.2013771996939555,
+      "grad_norm": 1.9781200787965287,
+      "learning_rate": 7.261126804319391e-06,
+      "loss": 0.3284,
+      "step": 7851
+    },
+    {
+      "epoch": 1.2015302218821728,
+      "grad_norm": 2.0903308853535227,
+      "learning_rate": 7.258743466685119e-06,
+      "loss": 0.3898,
+      "step": 7852
+    },
+    {
+      "epoch": 1.2016832440703902,
+      "grad_norm": 2.0779995520765464,
+      "learning_rate": 7.2563602974021365e-06,
+      "loss": 0.3994,
+      "step": 7853
+    },
+    {
+      "epoch": 1.2018362662586075,
+      "grad_norm": 2.185208271217259,
+      "learning_rate": 7.253977296616802e-06,
+      "loss": 0.4303,
+      "step": 7854
+    },
+    {
+      "epoch": 1.2019892884468248,
+      "grad_norm": 2.093158167390441,
+      "learning_rate": 7.251594464475462e-06,
+      "loss": 0.4102,
+      "step": 7855
+    },
+    {
+      "epoch": 1.2021423106350422,
+      "grad_norm": 2.474345761675588,
+      "learning_rate": 7.249211801124459e-06,
+      "loss": 0.3867,
+      "step": 7856
+    },
+    {
+      "epoch": 1.2022953328232593,
+      "grad_norm": 2.354161944422409,
+      "learning_rate": 7.246829306710118e-06,
+      "loss": 0.4842,
+      "step": 7857
+    },
+    {
+      "epoch": 1.2024483550114766,
+      "grad_norm": 2.066474315946716,
+      "learning_rate": 7.244446981378764e-06,
+      "loss": 0.3929,
+      "step": 7858
+    },
+    {
+      "epoch": 1.202601377199694,
+      "grad_norm": 2.2841546405429742,
+      "learning_rate": 7.242064825276699e-06,
+      "loss": 0.3933,
+      "step": 7859
+    },
+    {
+      "epoch": 1.2027543993879113,
+      "grad_norm": 2.421878768693007,
+      "learning_rate": 7.239682838550219e-06,
+      "loss": 0.4667,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2029074215761286,
+      "grad_norm": 2.121335990190176,
+      "learning_rate": 7.237301021345621e-06,
+      "loss": 0.4164,
+      "step": 7861
+    },
+    {
+      "epoch": 1.203060443764346,
+      "grad_norm": 2.2477816426607347,
+      "learning_rate": 7.23491937380917e-06,
+      "loss": 0.4141,
+      "step": 7862
+    },
+    {
+      "epoch": 1.203213465952563,
+      "grad_norm": 2.0842774961175965,
+      "learning_rate": 7.232537896087138e-06,
+      "loss": 0.4256,
+      "step": 7863
+    },
+    {
+      "epoch": 1.2033664881407804,
+      "grad_norm": 2.3760441894624034,
+      "learning_rate": 7.230156588325783e-06,
+      "loss": 0.4749,
+      "step": 7864
+    },
+    {
+      "epoch": 1.2035195103289977,
+      "grad_norm": 2.8127024874878117,
+      "learning_rate": 7.2277754506713415e-06,
+      "loss": 0.3896,
+      "step": 7865
+    },
+    {
+      "epoch": 1.203672532517215,
+      "grad_norm": 2.3056681652953546,
+      "learning_rate": 7.22539448327006e-06,
+      "loss": 0.4713,
+      "step": 7866
+    },
+    {
+      "epoch": 1.2038255547054324,
+      "grad_norm": 2.038914674694339,
+      "learning_rate": 7.223013686268159e-06,
+      "loss": 0.3701,
+      "step": 7867
+    },
+    {
+      "epoch": 1.2039785768936495,
+      "grad_norm": 2.279184568859362,
+      "learning_rate": 7.2206330598118435e-06,
+      "loss": 0.5052,
+      "step": 7868
+    },
+    {
+      "epoch": 1.2041315990818668,
+      "grad_norm": 2.362094871311031,
+      "learning_rate": 7.218252604047331e-06,
+      "loss": 0.4475,
+      "step": 7869
+    },
+    {
+      "epoch": 1.2042846212700842,
+      "grad_norm": 2.2299992118588188,
+      "learning_rate": 7.215872319120809e-06,
+      "loss": 0.4853,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2044376434583015,
+      "grad_norm": 1.8836964001194594,
+      "learning_rate": 7.213492205178454e-06,
+      "loss": 0.2751,
+      "step": 7871
+    },
+    {
+      "epoch": 1.2045906656465188,
+      "grad_norm": 2.1731065858170515,
+      "learning_rate": 7.2111122623664486e-06,
+      "loss": 0.4221,
+      "step": 7872
+    },
+    {
+      "epoch": 1.204743687834736,
+      "grad_norm": 2.042405359275383,
+      "learning_rate": 7.208732490830945e-06,
+      "loss": 0.3539,
+      "step": 7873
+    },
+    {
+      "epoch": 1.2048967100229533,
+      "grad_norm": 1.916086959905358,
+      "learning_rate": 7.206352890718102e-06,
+      "loss": 0.3526,
+      "step": 7874
+    },
+    {
+      "epoch": 1.2050497322111706,
+      "grad_norm": 2.08298656195251,
+      "learning_rate": 7.203973462174059e-06,
+      "loss": 0.3444,
+      "step": 7875
+    },
+    {
+      "epoch": 1.205202754399388,
+      "grad_norm": 2.127087834921326,
+      "learning_rate": 7.201594205344937e-06,
+      "loss": 0.3774,
+      "step": 7876
+    },
+    {
+      "epoch": 1.2053557765876053,
+      "grad_norm": 1.8822819513462488,
+      "learning_rate": 7.199215120376866e-06,
+      "loss": 0.4222,
+      "step": 7877
+    },
+    {
+      "epoch": 1.2055087987758224,
+      "grad_norm": 2.0069738268178043,
+      "learning_rate": 7.19683620741595e-06,
+      "loss": 0.3617,
+      "step": 7878
+    },
+    {
+      "epoch": 1.2056618209640397,
+      "grad_norm": 2.347032617268608,
+      "learning_rate": 7.1944574666082925e-06,
+      "loss": 0.4167,
+      "step": 7879
+    },
+    {
+      "epoch": 1.205814843152257,
+      "grad_norm": 1.9841895774201872,
+      "learning_rate": 7.192078898099975e-06,
+      "loss": 0.3959,
+      "step": 7880
+    },
+    {
+      "epoch": 1.2059678653404744,
+      "grad_norm": 2.1545952125519254,
+      "learning_rate": 7.189700502037073e-06,
+      "loss": 0.3581,
+      "step": 7881
+    },
+    {
+      "epoch": 1.2061208875286917,
+      "grad_norm": 2.2942907440257345,
+      "learning_rate": 7.187322278565665e-06,
+      "loss": 0.4123,
+      "step": 7882
+    },
+    {
+      "epoch": 1.2062739097169088,
+      "grad_norm": 2.3337214188107076,
+      "learning_rate": 7.184944227831794e-06,
+      "loss": 0.4186,
+      "step": 7883
+    },
+    {
+      "epoch": 1.2064269319051262,
+      "grad_norm": 2.1105765004274546,
+      "learning_rate": 7.18256634998151e-06,
+      "loss": 0.3696,
+      "step": 7884
+    },
+    {
+      "epoch": 1.2065799540933435,
+      "grad_norm": 1.8495228363779856,
+      "learning_rate": 7.180188645160851e-06,
+      "loss": 0.3263,
+      "step": 7885
+    },
+    {
+      "epoch": 1.2067329762815608,
+      "grad_norm": 2.178466457135846,
+      "learning_rate": 7.1778111135158355e-06,
+      "loss": 0.4567,
+      "step": 7886
+    },
+    {
+      "epoch": 1.2068859984697782,
+      "grad_norm": 2.329996121904899,
+      "learning_rate": 7.175433755192483e-06,
+      "loss": 0.3929,
+      "step": 7887
+    },
+    {
+      "epoch": 1.2070390206579953,
+      "grad_norm": 2.7057684229341143,
+      "learning_rate": 7.173056570336794e-06,
+      "loss": 0.4107,
+      "step": 7888
+    },
+    {
+      "epoch": 1.2071920428462126,
+      "grad_norm": 2.2813650165867045,
+      "learning_rate": 7.170679559094755e-06,
+      "loss": 0.3918,
+      "step": 7889
+    },
+    {
+      "epoch": 1.20734506503443,
+      "grad_norm": 2.2649980429048617,
+      "learning_rate": 7.168302721612357e-06,
+      "loss": 0.3625,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2074980872226473,
+      "grad_norm": 1.9690419186555863,
+      "learning_rate": 7.165926058035566e-06,
+      "loss": 0.3964,
+      "step": 7891
+    },
+    {
+      "epoch": 1.2076511094108646,
+      "grad_norm": 2.402097955904389,
+      "learning_rate": 7.163549568510341e-06,
+      "loss": 0.4267,
+      "step": 7892
+    },
+    {
+      "epoch": 1.207804131599082,
+      "grad_norm": 2.241109535031182,
+      "learning_rate": 7.161173253182634e-06,
+      "loss": 0.3604,
+      "step": 7893
+    },
+    {
+      "epoch": 1.207957153787299,
+      "grad_norm": 2.3475139013875213,
+      "learning_rate": 7.1587971121983815e-06,
+      "loss": 0.4277,
+      "step": 7894
+    },
+    {
+      "epoch": 1.2081101759755164,
+      "grad_norm": 2.4465913526801844,
+      "learning_rate": 7.156421145703517e-06,
+      "loss": 0.3985,
+      "step": 7895
+    },
+    {
+      "epoch": 1.2082631981637337,
+      "grad_norm": 2.3863337251537318,
+      "learning_rate": 7.154045353843952e-06,
+      "loss": 0.3679,
+      "step": 7896
+    },
+    {
+      "epoch": 1.208416220351951,
+      "grad_norm": 2.126924455000003,
+      "learning_rate": 7.151669736765594e-06,
+      "loss": 0.3994,
+      "step": 7897
+    },
+    {
+      "epoch": 1.2085692425401684,
+      "grad_norm": 2.304629645464992,
+      "learning_rate": 7.149294294614344e-06,
+      "loss": 0.3881,
+      "step": 7898
+    },
+    {
+      "epoch": 1.2087222647283857,
+      "grad_norm": 2.10307375531413,
+      "learning_rate": 7.146919027536081e-06,
+      "loss": 0.407,
+      "step": 7899
+    },
+    {
+      "epoch": 1.2088752869166028,
+      "grad_norm": 2.1356508233173357,
+      "learning_rate": 7.14454393567668e-06,
+      "loss": 0.3948,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2090283091048202,
+      "grad_norm": 2.1661367713580146,
+      "learning_rate": 7.14216901918201e-06,
+      "loss": 0.369,
+      "step": 7901
+    },
+    {
+      "epoch": 1.2091813312930375,
+      "grad_norm": 2.073359287919676,
+      "learning_rate": 7.1397942781979165e-06,
+      "loss": 0.4001,
+      "step": 7902
+    },
+    {
+      "epoch": 1.2093343534812548,
+      "grad_norm": 2.285306774993513,
+      "learning_rate": 7.137419712870248e-06,
+      "loss": 0.4791,
+      "step": 7903
+    },
+    {
+      "epoch": 1.2094873756694722,
+      "grad_norm": 2.2565147922670272,
+      "learning_rate": 7.135045323344836e-06,
+      "loss": 0.4129,
+      "step": 7904
+    },
+    {
+      "epoch": 1.2096403978576893,
+      "grad_norm": 2.205771600598996,
+      "learning_rate": 7.132671109767493e-06,
+      "loss": 0.4151,
+      "step": 7905
+    },
+    {
+      "epoch": 1.2097934200459066,
+      "grad_norm": 2.3554198855954804,
+      "learning_rate": 7.130297072284037e-06,
+      "loss": 0.3735,
+      "step": 7906
+    },
+    {
+      "epoch": 1.209946442234124,
+      "grad_norm": 2.460816659992818,
+      "learning_rate": 7.127923211040267e-06,
+      "loss": 0.4421,
+      "step": 7907
+    },
+    {
+      "epoch": 1.2100994644223413,
+      "grad_norm": 2.2894306166581697,
+      "learning_rate": 7.125549526181963e-06,
+      "loss": 0.4287,
+      "step": 7908
+    },
+    {
+      "epoch": 1.2102524866105586,
+      "grad_norm": 2.2431820015839734,
+      "learning_rate": 7.123176017854909e-06,
+      "loss": 0.3583,
+      "step": 7909
+    },
+    {
+      "epoch": 1.2104055087987757,
+      "grad_norm": 2.3312818039925576,
+      "learning_rate": 7.120802686204869e-06,
+      "loss": 0.4208,
+      "step": 7910
+    },
+    {
+      "epoch": 1.210558530986993,
+      "grad_norm": 2.3032859188438737,
+      "learning_rate": 7.118429531377604e-06,
+      "loss": 0.4121,
+      "step": 7911
+    },
+    {
+      "epoch": 1.2107115531752104,
+      "grad_norm": 2.251670699371594,
+      "learning_rate": 7.116056553518851e-06,
+      "loss": 0.4033,
+      "step": 7912
+    },
+    {
+      "epoch": 1.2108645753634277,
+      "grad_norm": 2.0986050604327278,
+      "learning_rate": 7.113683752774345e-06,
+      "loss": 0.3579,
+      "step": 7913
+    },
+    {
+      "epoch": 1.211017597551645,
+      "grad_norm": 2.2418784167236545,
+      "learning_rate": 7.1113111292898174e-06,
+      "loss": 0.407,
+      "step": 7914
+    },
+    {
+      "epoch": 1.2111706197398622,
+      "grad_norm": 2.163620185778059,
+      "learning_rate": 7.108938683210968e-06,
+      "loss": 0.4055,
+      "step": 7915
+    },
+    {
+      "epoch": 1.2113236419280795,
+      "grad_norm": 2.355202682771081,
+      "learning_rate": 7.106566414683506e-06,
+      "loss": 0.4669,
+      "step": 7916
+    },
+    {
+      "epoch": 1.2114766641162968,
+      "grad_norm": 2.1271629737901625,
+      "learning_rate": 7.104194323853122e-06,
+      "loss": 0.3561,
+      "step": 7917
+    },
+    {
+      "epoch": 1.2116296863045142,
+      "grad_norm": 2.0575868876645096,
+      "learning_rate": 7.101822410865489e-06,
+      "loss": 0.3592,
+      "step": 7918
+    },
+    {
+      "epoch": 1.2117827084927315,
+      "grad_norm": 2.0538971546012696,
+      "learning_rate": 7.099450675866282e-06,
+      "loss": 0.4481,
+      "step": 7919
+    },
+    {
+      "epoch": 1.2119357306809486,
+      "grad_norm": 1.919676635657094,
+      "learning_rate": 7.0970791190011566e-06,
+      "loss": 0.3355,
+      "step": 7920
+    },
+    {
+      "epoch": 1.212088752869166,
+      "grad_norm": 2.204903650709108,
+      "learning_rate": 7.094707740415756e-06,
+      "loss": 0.4253,
+      "step": 7921
+    },
+    {
+      "epoch": 1.2122417750573833,
+      "grad_norm": 2.3272233967248335,
+      "learning_rate": 7.09233654025572e-06,
+      "loss": 0.3645,
+      "step": 7922
+    },
+    {
+      "epoch": 1.2123947972456006,
+      "grad_norm": 2.116304203277378,
+      "learning_rate": 7.08996551866667e-06,
+      "loss": 0.4008,
+      "step": 7923
+    },
+    {
+      "epoch": 1.212547819433818,
+      "grad_norm": 2.12004466283652,
+      "learning_rate": 7.087594675794226e-06,
+      "loss": 0.3908,
+      "step": 7924
+    },
+    {
+      "epoch": 1.2127008416220353,
+      "grad_norm": 2.3012142929784414,
+      "learning_rate": 7.085224011783984e-06,
+      "loss": 0.4094,
+      "step": 7925
+    },
+    {
+      "epoch": 1.2128538638102524,
+      "grad_norm": 2.225388844694236,
+      "learning_rate": 7.082853526781537e-06,
+      "loss": 0.3643,
+      "step": 7926
+    },
+    {
+      "epoch": 1.2130068859984697,
+      "grad_norm": 2.076324760976028,
+      "learning_rate": 7.080483220932469e-06,
+      "loss": 0.5358,
+      "step": 7927
+    },
+    {
+      "epoch": 1.213159908186687,
+      "grad_norm": 2.108445837070922,
+      "learning_rate": 7.078113094382348e-06,
+      "loss": 0.3903,
+      "step": 7928
+    },
+    {
+      "epoch": 1.2133129303749044,
+      "grad_norm": 2.541491652948375,
+      "learning_rate": 7.075743147276728e-06,
+      "loss": 0.4761,
+      "step": 7929
+    },
+    {
+      "epoch": 1.2134659525631217,
+      "grad_norm": 2.1264106715257003,
+      "learning_rate": 7.073373379761166e-06,
+      "loss": 0.4052,
+      "step": 7930
+    },
+    {
+      "epoch": 1.213618974751339,
+      "grad_norm": 2.500809227928582,
+      "learning_rate": 7.07100379198119e-06,
+      "loss": 0.4861,
+      "step": 7931
+    },
+    {
+      "epoch": 1.2137719969395562,
+      "grad_norm": 2.0731214372985245,
+      "learning_rate": 7.068634384082331e-06,
+      "loss": 0.3271,
+      "step": 7932
+    },
+    {
+      "epoch": 1.2139250191277735,
+      "grad_norm": 2.2473665302879544,
+      "learning_rate": 7.066265156210105e-06,
+      "loss": 0.3925,
+      "step": 7933
+    },
+    {
+      "epoch": 1.2140780413159908,
+      "grad_norm": 2.023127298800279,
+      "learning_rate": 7.063896108510008e-06,
+      "loss": 0.374,
+      "step": 7934
+    },
+    {
+      "epoch": 1.2142310635042082,
+      "grad_norm": 2.1057736936995712,
+      "learning_rate": 7.0615272411275395e-06,
+      "loss": 0.4494,
+      "step": 7935
+    },
+    {
+      "epoch": 1.2143840856924255,
+      "grad_norm": 2.1942854844397583,
+      "learning_rate": 7.05915855420818e-06,
+      "loss": 0.4072,
+      "step": 7936
+    },
+    {
+      "epoch": 1.2145371078806426,
+      "grad_norm": 2.16395490685311,
+      "learning_rate": 7.056790047897394e-06,
+      "loss": 0.3736,
+      "step": 7937
+    },
+    {
+      "epoch": 1.21469013006886,
+      "grad_norm": 2.2678730333273793,
+      "learning_rate": 7.054421722340647e-06,
+      "loss": 0.444,
+      "step": 7938
+    },
+    {
+      "epoch": 1.2148431522570773,
+      "grad_norm": 2.200405364638579,
+      "learning_rate": 7.052053577683384e-06,
+      "loss": 0.3841,
+      "step": 7939
+    },
+    {
+      "epoch": 1.2149961744452946,
+      "grad_norm": 2.277747587318131,
+      "learning_rate": 7.049685614071047e-06,
+      "loss": 0.476,
+      "step": 7940
+    },
+    {
+      "epoch": 1.215149196633512,
+      "grad_norm": 2.010732008758074,
+      "learning_rate": 7.047317831649056e-06,
+      "loss": 0.3571,
+      "step": 7941
+    },
+    {
+      "epoch": 1.215302218821729,
+      "grad_norm": 2.252647658079372,
+      "learning_rate": 7.044950230562826e-06,
+      "loss": 0.4632,
+      "step": 7942
+    },
+    {
+      "epoch": 1.2154552410099464,
+      "grad_norm": 2.3309241879798486,
+      "learning_rate": 7.042582810957767e-06,
+      "loss": 0.407,
+      "step": 7943
+    },
+    {
+      "epoch": 1.2156082631981637,
+      "grad_norm": 1.9771222565408002,
+      "learning_rate": 7.040215572979262e-06,
+      "loss": 0.3727,
+      "step": 7944
+    },
+    {
+      "epoch": 1.215761285386381,
+      "grad_norm": 2.2011196818146517,
+      "learning_rate": 7.0378485167727e-06,
+      "loss": 0.3456,
+      "step": 7945
+    },
+    {
+      "epoch": 1.2159143075745984,
+      "grad_norm": 2.0983866502398154,
+      "learning_rate": 7.035481642483451e-06,
+      "loss": 0.3567,
+      "step": 7946
+    },
+    {
+      "epoch": 1.2160673297628155,
+      "grad_norm": 2.391056276231594,
+      "learning_rate": 7.033114950256865e-06,
+      "loss": 0.4006,
+      "step": 7947
+    },
+    {
+      "epoch": 1.2162203519510328,
+      "grad_norm": 2.3512508079427223,
+      "learning_rate": 7.0307484402383015e-06,
+      "loss": 0.4517,
+      "step": 7948
+    },
+    {
+      "epoch": 1.2163733741392502,
+      "grad_norm": 2.1246473090309577,
+      "learning_rate": 7.028382112573093e-06,
+      "loss": 0.3766,
+      "step": 7949
+    },
+    {
+      "epoch": 1.2165263963274675,
+      "grad_norm": 2.075108528173804,
+      "learning_rate": 7.026015967406559e-06,
+      "loss": 0.3935,
+      "step": 7950
+    },
+    {
+      "epoch": 1.2166794185156848,
+      "grad_norm": 2.121208516187529,
+      "learning_rate": 7.023650004884024e-06,
+      "loss": 0.3355,
+      "step": 7951
+    },
+    {
+      "epoch": 1.216832440703902,
+      "grad_norm": 2.3563143997973084,
+      "learning_rate": 7.021284225150782e-06,
+      "loss": 0.398,
+      "step": 7952
+    },
+    {
+      "epoch": 1.2169854628921193,
+      "grad_norm": 2.073709554473588,
+      "learning_rate": 7.0189186283521295e-06,
+      "loss": 0.3789,
+      "step": 7953
+    },
+    {
+      "epoch": 1.2171384850803366,
+      "grad_norm": 2.528122947279484,
+      "learning_rate": 7.016553214633348e-06,
+      "loss": 0.4205,
+      "step": 7954
+    },
+    {
+      "epoch": 1.217291507268554,
+      "grad_norm": 2.2205808799791393,
+      "learning_rate": 7.014187984139702e-06,
+      "loss": 0.3803,
+      "step": 7955
+    },
+    {
+      "epoch": 1.2174445294567713,
+      "grad_norm": 2.0751870027824526,
+      "learning_rate": 7.0118229370164545e-06,
+      "loss": 0.3547,
+      "step": 7956
+    },
+    {
+      "epoch": 1.2175975516449886,
+      "grad_norm": 2.2315342903135393,
+      "learning_rate": 7.009458073408852e-06,
+      "loss": 0.4626,
+      "step": 7957
+    },
+    {
+      "epoch": 1.2177505738332057,
+      "grad_norm": 2.1810279357679,
+      "learning_rate": 7.007093393462124e-06,
+      "loss": 0.4244,
+      "step": 7958
+    },
+    {
+      "epoch": 1.217903596021423,
+      "grad_norm": 2.0422908944115594,
+      "learning_rate": 7.004728897321501e-06,
+      "loss": 0.5326,
+      "step": 7959
+    },
+    {
+      "epoch": 1.2180566182096404,
+      "grad_norm": 2.20462359397819,
+      "learning_rate": 7.002364585132192e-06,
+      "loss": 0.3102,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2182096403978577,
+      "grad_norm": 2.3536974554981125,
+      "learning_rate": 7.000000457039404e-06,
+      "loss": 0.4173,
+      "step": 7961
+    },
+    {
+      "epoch": 1.218362662586075,
+      "grad_norm": 2.038361961156172,
+      "learning_rate": 6.997636513188324e-06,
+      "loss": 0.39,
+      "step": 7962
+    },
+    {
+      "epoch": 1.2185156847742924,
+      "grad_norm": 2.210987365831342,
+      "learning_rate": 6.995272753724127e-06,
+      "loss": 0.4057,
+      "step": 7963
+    },
+    {
+      "epoch": 1.2186687069625095,
+      "grad_norm": 2.1544425173767907,
+      "learning_rate": 6.99290917879199e-06,
+      "loss": 0.3487,
+      "step": 7964
+    },
+    {
+      "epoch": 1.2188217291507268,
+      "grad_norm": 2.345572013946747,
+      "learning_rate": 6.990545788537062e-06,
+      "loss": 0.4078,
+      "step": 7965
+    },
+    {
+      "epoch": 1.2189747513389442,
+      "grad_norm": 2.0550826810477854,
+      "learning_rate": 6.988182583104488e-06,
+      "loss": 0.3286,
+      "step": 7966
+    },
+    {
+      "epoch": 1.2191277735271615,
+      "grad_norm": 2.474145020752616,
+      "learning_rate": 6.985819562639406e-06,
+      "loss": 0.3959,
+      "step": 7967
+    },
+    {
+      "epoch": 1.2192807957153788,
+      "grad_norm": 2.1499005174076764,
+      "learning_rate": 6.9834567272869345e-06,
+      "loss": 0.4703,
+      "step": 7968
+    },
+    {
+      "epoch": 1.219433817903596,
+      "grad_norm": 1.9949459289594664,
+      "learning_rate": 6.981094077192188e-06,
+      "loss": 0.4444,
+      "step": 7969
+    },
+    {
+      "epoch": 1.2195868400918133,
+      "grad_norm": 2.159287629056209,
+      "learning_rate": 6.978731612500266e-06,
+      "loss": 0.4424,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2197398622800306,
+      "grad_norm": 2.319018078003906,
+      "learning_rate": 6.97636933335625e-06,
+      "loss": 0.456,
+      "step": 7971
+    },
+    {
+      "epoch": 1.219892884468248,
+      "grad_norm": 2.2501379725026194,
+      "learning_rate": 6.974007239905225e-06,
+      "loss": 0.3941,
+      "step": 7972
+    },
+    {
+      "epoch": 1.2200459066564653,
+      "grad_norm": 2.376289275218683,
+      "learning_rate": 6.971645332292255e-06,
+      "loss": 0.479,
+      "step": 7973
+    },
+    {
+      "epoch": 1.2201989288446824,
+      "grad_norm": 2.4438939561516597,
+      "learning_rate": 6.969283610662387e-06,
+      "loss": 0.4189,
+      "step": 7974
+    },
+    {
+      "epoch": 1.2203519510328997,
+      "grad_norm": 2.0860769225758347,
+      "learning_rate": 6.966922075160671e-06,
+      "loss": 0.3817,
+      "step": 7975
+    },
+    {
+      "epoch": 1.220504973221117,
+      "grad_norm": 2.2571172531938952,
+      "learning_rate": 6.964560725932136e-06,
+      "loss": 0.3865,
+      "step": 7976
+    },
+    {
+      "epoch": 1.2206579954093344,
+      "grad_norm": 2.0870809930095686,
+      "learning_rate": 6.962199563121803e-06,
+      "loss": 0.4098,
+      "step": 7977
+    },
+    {
+      "epoch": 1.2208110175975517,
+      "grad_norm": 2.3337013901306065,
+      "learning_rate": 6.959838586874679e-06,
+      "loss": 0.4347,
+      "step": 7978
+    },
+    {
+      "epoch": 1.2209640397857688,
+      "grad_norm": 1.6879922077725298,
+      "learning_rate": 6.957477797335757e-06,
+      "loss": 0.3284,
+      "step": 7979
+    },
+    {
+      "epoch": 1.2211170619739862,
+      "grad_norm": 2.21500011103774,
+      "learning_rate": 6.955117194650029e-06,
+      "loss": 0.4315,
+      "step": 7980
+    },
+    {
+      "epoch": 1.2212700841622035,
+      "grad_norm": 2.365885263450686,
+      "learning_rate": 6.952756778962463e-06,
+      "loss": 0.4451,
+      "step": 7981
+    },
+    {
+      "epoch": 1.2214231063504208,
+      "grad_norm": 2.1970439077941983,
+      "learning_rate": 6.950396550418026e-06,
+      "loss": 0.3978,
+      "step": 7982
+    },
+    {
+      "epoch": 1.2215761285386382,
+      "grad_norm": 2.00539883100848,
+      "learning_rate": 6.9480365091616685e-06,
+      "loss": 0.3948,
+      "step": 7983
+    },
+    {
+      "epoch": 1.2217291507268553,
+      "grad_norm": 2.4169358456913175,
+      "learning_rate": 6.945676655338324e-06,
+      "loss": 0.4454,
+      "step": 7984
+    },
+    {
+      "epoch": 1.2218821729150726,
+      "grad_norm": 1.9858033554942487,
+      "learning_rate": 6.943316989092928e-06,
+      "loss": 0.3527,
+      "step": 7985
+    },
+    {
+      "epoch": 1.22203519510329,
+      "grad_norm": 2.432567266913346,
+      "learning_rate": 6.940957510570395e-06,
+      "loss": 0.4424,
+      "step": 7986
+    },
+    {
+      "epoch": 1.2221882172915073,
+      "grad_norm": 2.0487590024042253,
+      "learning_rate": 6.938598219915624e-06,
+      "loss": 0.4262,
+      "step": 7987
+    },
+    {
+      "epoch": 1.2223412394797246,
+      "grad_norm": 2.1278806090330153,
+      "learning_rate": 6.9362391172735155e-06,
+      "loss": 0.4153,
+      "step": 7988
+    },
+    {
+      "epoch": 1.2224942616679417,
+      "grad_norm": 1.9392705667390682,
+      "learning_rate": 6.933880202788945e-06,
+      "loss": 0.2756,
+      "step": 7989
+    },
+    {
+      "epoch": 1.222647283856159,
+      "grad_norm": 2.0982874308568746,
+      "learning_rate": 6.931521476606791e-06,
+      "loss": 0.3839,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2228003060443764,
+      "grad_norm": 1.9905106427152186,
+      "learning_rate": 6.929162938871905e-06,
+      "loss": 0.3439,
+      "step": 7991
+    },
+    {
+      "epoch": 1.2229533282325937,
+      "grad_norm": 2.137110660745122,
+      "learning_rate": 6.926804589729133e-06,
+      "loss": 0.3376,
+      "step": 7992
+    },
+    {
+      "epoch": 1.223106350420811,
+      "grad_norm": 2.0719186341939326,
+      "learning_rate": 6.924446429323318e-06,
+      "loss": 0.3714,
+      "step": 7993
+    },
+    {
+      "epoch": 1.2232593726090284,
+      "grad_norm": 1.9949333583357212,
+      "learning_rate": 6.922088457799278e-06,
+      "loss": 0.3984,
+      "step": 7994
+    },
+    {
+      "epoch": 1.2234123947972455,
+      "grad_norm": 2.0901002764905763,
+      "learning_rate": 6.919730675301824e-06,
+      "loss": 0.3392,
+      "step": 7995
+    },
+    {
+      "epoch": 1.2235654169854628,
+      "grad_norm": 2.0467946793168212,
+      "learning_rate": 6.917373081975764e-06,
+      "loss": 0.4335,
+      "step": 7996
+    },
+    {
+      "epoch": 1.2237184391736802,
+      "grad_norm": 2.099938498749769,
+      "learning_rate": 6.9150156779658775e-06,
+      "loss": 0.3334,
+      "step": 7997
+    },
+    {
+      "epoch": 1.2238714613618975,
+      "grad_norm": 2.315397070406997,
+      "learning_rate": 6.91265846341695e-06,
+      "loss": 0.4228,
+      "step": 7998
+    },
+    {
+      "epoch": 1.2240244835501148,
+      "grad_norm": 2.3280652559278465,
+      "learning_rate": 6.910301438473746e-06,
+      "loss": 0.4021,
+      "step": 7999
+    },
+    {
+      "epoch": 1.2241775057383322,
+      "grad_norm": 2.3985582799227076,
+      "learning_rate": 6.907944603281011e-06,
+      "loss": 0.3723,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2243305279265493,
+      "grad_norm": 2.130601512927696,
+      "learning_rate": 6.905587957983499e-06,
+      "loss": 0.3039,
+      "step": 8001
+    },
+    {
+      "epoch": 1.2244835501147666,
+      "grad_norm": 2.1247620729874748,
+      "learning_rate": 6.903231502725936e-06,
+      "loss": 0.4024,
+      "step": 8002
+    },
+    {
+      "epoch": 1.224636572302984,
+      "grad_norm": 2.497906614459289,
+      "learning_rate": 6.900875237653039e-06,
+      "loss": 0.4527,
+      "step": 8003
+    },
+    {
+      "epoch": 1.2247895944912013,
+      "grad_norm": 2.234719038902452,
+      "learning_rate": 6.8985191629095185e-06,
+      "loss": 0.3417,
+      "step": 8004
+    },
+    {
+      "epoch": 1.2249426166794186,
+      "grad_norm": 2.5449686391944337,
+      "learning_rate": 6.8961632786400665e-06,
+      "loss": 0.479,
+      "step": 8005
+    },
+    {
+      "epoch": 1.2250956388676357,
+      "grad_norm": 2.0912533506618227,
+      "learning_rate": 6.893807584989375e-06,
+      "loss": 0.3885,
+      "step": 8006
+    },
+    {
+      "epoch": 1.225248661055853,
+      "grad_norm": 2.061740186429149,
+      "learning_rate": 6.891452082102108e-06,
+      "loss": 0.3069,
+      "step": 8007
+    },
+    {
+      "epoch": 1.2254016832440704,
+      "grad_norm": 2.4195881426112176,
+      "learning_rate": 6.889096770122928e-06,
+      "loss": 0.3886,
+      "step": 8008
+    },
+    {
+      "epoch": 1.2255547054322877,
+      "grad_norm": 2.257160486240451,
+      "learning_rate": 6.88674164919649e-06,
+      "loss": 0.442,
+      "step": 8009
+    },
+    {
+      "epoch": 1.225707727620505,
+      "grad_norm": 1.9940025514019444,
+      "learning_rate": 6.8843867194674244e-06,
+      "loss": 0.3627,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2258607498087222,
+      "grad_norm": 2.0383222857719385,
+      "learning_rate": 6.882031981080355e-06,
+      "loss": 0.4049,
+      "step": 8011
+    },
+    {
+      "epoch": 1.2260137719969395,
+      "grad_norm": 2.0190329535880545,
+      "learning_rate": 6.879677434179904e-06,
+      "loss": 0.2951,
+      "step": 8012
+    },
+    {
+      "epoch": 1.2261667941851568,
+      "grad_norm": 2.055375099684583,
+      "learning_rate": 6.877323078910665e-06,
+      "loss": 0.3451,
+      "step": 8013
+    },
+    {
+      "epoch": 1.2263198163733742,
+      "grad_norm": 2.4333399428901563,
+      "learning_rate": 6.874968915417234e-06,
+      "loss": 0.3886,
+      "step": 8014
+    },
+    {
+      "epoch": 1.2264728385615915,
+      "grad_norm": 2.526027909918158,
+      "learning_rate": 6.872614943844189e-06,
+      "loss": 0.4255,
+      "step": 8015
+    },
+    {
+      "epoch": 1.2266258607498086,
+      "grad_norm": 2.1961298260330318,
+      "learning_rate": 6.870261164336089e-06,
+      "loss": 0.3917,
+      "step": 8016
+    },
+    {
+      "epoch": 1.226778882938026,
+      "grad_norm": 1.9173991175517395,
+      "learning_rate": 6.867907577037498e-06,
+      "loss": 0.3306,
+      "step": 8017
+    },
+    {
+      "epoch": 1.2269319051262433,
+      "grad_norm": 1.9365177251844081,
+      "learning_rate": 6.865554182092954e-06,
+      "loss": 0.3061,
+      "step": 8018
+    },
+    {
+      "epoch": 1.2270849273144606,
+      "grad_norm": 1.8093240085090097,
+      "learning_rate": 6.8632009796469935e-06,
+      "loss": 0.4069,
+      "step": 8019
+    },
+    {
+      "epoch": 1.227237949502678,
+      "grad_norm": 2.149860287286806,
+      "learning_rate": 6.860847969844129e-06,
+      "loss": 0.4024,
+      "step": 8020
+    },
+    {
+      "epoch": 1.227390971690895,
+      "grad_norm": 2.2115737004229428,
+      "learning_rate": 6.85849515282887e-06,
+      "loss": 0.3812,
+      "step": 8021
+    },
+    {
+      "epoch": 1.2275439938791124,
+      "grad_norm": 1.9495383178771113,
+      "learning_rate": 6.856142528745717e-06,
+      "loss": 0.3352,
+      "step": 8022
+    },
+    {
+      "epoch": 1.2276970160673297,
+      "grad_norm": 2.271878221084577,
+      "learning_rate": 6.853790097739148e-06,
+      "loss": 0.392,
+      "step": 8023
+    },
+    {
+      "epoch": 1.227850038255547,
+      "grad_norm": 2.0960449848601144,
+      "learning_rate": 6.851437859953636e-06,
+      "loss": 0.4094,
+      "step": 8024
+    },
+    {
+      "epoch": 1.2280030604437644,
+      "grad_norm": 2.1636547073328654,
+      "learning_rate": 6.849085815533644e-06,
+      "loss": 0.4111,
+      "step": 8025
+    },
+    {
+      "epoch": 1.2281560826319817,
+      "grad_norm": 2.4084541084097606,
+      "learning_rate": 6.846733964623618e-06,
+      "loss": 0.3485,
+      "step": 8026
+    },
+    {
+      "epoch": 1.2283091048201988,
+      "grad_norm": 1.9854369939079013,
+      "learning_rate": 6.844382307367994e-06,
+      "loss": 0.42,
+      "step": 8027
+    },
+    {
+      "epoch": 1.2284621270084162,
+      "grad_norm": 2.1105981109720138,
+      "learning_rate": 6.842030843911201e-06,
+      "loss": 0.3964,
+      "step": 8028
+    },
+    {
+      "epoch": 1.2286151491966335,
+      "grad_norm": 2.218758898126614,
+      "learning_rate": 6.839679574397641e-06,
+      "loss": 0.3978,
+      "step": 8029
+    },
+    {
+      "epoch": 1.2287681713848508,
+      "grad_norm": 2.2227044245916154,
+      "learning_rate": 6.837328498971727e-06,
+      "loss": 0.4151,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2289211935730682,
+      "grad_norm": 2.140680430133614,
+      "learning_rate": 6.834977617777844e-06,
+      "loss": 0.3219,
+      "step": 8031
+    },
+    {
+      "epoch": 1.2290742157612855,
+      "grad_norm": 1.9854371782404445,
+      "learning_rate": 6.83262693096036e-06,
+      "loss": 0.3803,
+      "step": 8032
+    },
+    {
+      "epoch": 1.2292272379495026,
+      "grad_norm": 2.205663362151382,
+      "learning_rate": 6.830276438663654e-06,
+      "loss": 0.3955,
+      "step": 8033
+    },
+    {
+      "epoch": 1.22938026013772,
+      "grad_norm": 2.10792051555875,
+      "learning_rate": 6.827926141032066e-06,
+      "loss": 0.3967,
+      "step": 8034
+    },
+    {
+      "epoch": 1.2295332823259373,
+      "grad_norm": 2.0734884900888386,
+      "learning_rate": 6.8255760382099465e-06,
+      "loss": 0.4093,
+      "step": 8035
+    },
+    {
+      "epoch": 1.2296863045141546,
+      "grad_norm": 2.2376128890445077,
+      "learning_rate": 6.823226130341623e-06,
+      "loss": 0.4023,
+      "step": 8036
+    },
+    {
+      "epoch": 1.229839326702372,
+      "grad_norm": 2.401436260509043,
+      "learning_rate": 6.820876417571405e-06,
+      "loss": 0.461,
+      "step": 8037
+    },
+    {
+      "epoch": 1.229992348890589,
+      "grad_norm": 2.0884246258149406,
+      "learning_rate": 6.818526900043606e-06,
+      "loss": 0.3659,
+      "step": 8038
+    },
+    {
+      "epoch": 1.2301453710788064,
+      "grad_norm": 2.0483232952456545,
+      "learning_rate": 6.816177577902518e-06,
+      "loss": 0.3641,
+      "step": 8039
+    },
+    {
+      "epoch": 1.2302983932670237,
+      "grad_norm": 2.382850648132974,
+      "learning_rate": 6.813828451292417e-06,
+      "loss": 0.4182,
+      "step": 8040
+    },
+    {
+      "epoch": 1.230451415455241,
+      "grad_norm": 2.2418447789270948,
+      "learning_rate": 6.811479520357576e-06,
+      "loss": 0.4607,
+      "step": 8041
+    },
+    {
+      "epoch": 1.2306044376434584,
+      "grad_norm": 1.9382911340909974,
+      "learning_rate": 6.80913078524225e-06,
+      "loss": 0.3438,
+      "step": 8042
+    },
+    {
+      "epoch": 1.2307574598316755,
+      "grad_norm": 1.8422515197448954,
+      "learning_rate": 6.806782246090688e-06,
+      "loss": 0.3331,
+      "step": 8043
+    },
+    {
+      "epoch": 1.2309104820198928,
+      "grad_norm": 2.225159881582689,
+      "learning_rate": 6.804433903047118e-06,
+      "loss": 0.3753,
+      "step": 8044
+    },
+    {
+      "epoch": 1.2310635042081102,
+      "grad_norm": 2.1226629248453763,
+      "learning_rate": 6.802085756255763e-06,
+      "loss": 0.357,
+      "step": 8045
+    },
+    {
+      "epoch": 1.2312165263963275,
+      "grad_norm": 2.281272129316277,
+      "learning_rate": 6.7997378058608355e-06,
+      "loss": 0.4267,
+      "step": 8046
+    },
+    {
+      "epoch": 1.2313695485845448,
+      "grad_norm": 2.409255082789223,
+      "learning_rate": 6.797390052006526e-06,
+      "loss": 0.4512,
+      "step": 8047
+    },
+    {
+      "epoch": 1.231522570772762,
+      "grad_norm": 2.022442772507196,
+      "learning_rate": 6.7950424948370205e-06,
+      "loss": 0.3694,
+      "step": 8048
+    },
+    {
+      "epoch": 1.2316755929609793,
+      "grad_norm": 2.048689723681348,
+      "learning_rate": 6.792695134496497e-06,
+      "loss": 0.3168,
+      "step": 8049
+    },
+    {
+      "epoch": 1.2318286151491966,
+      "grad_norm": 2.1639453889819347,
+      "learning_rate": 6.7903479711291095e-06,
+      "loss": 0.373,
+      "step": 8050
+    },
+    {
+      "epoch": 1.231981637337414,
+      "grad_norm": 2.306482688733974,
+      "learning_rate": 6.78800100487901e-06,
+      "loss": 0.41,
+      "step": 8051
+    },
+    {
+      "epoch": 1.2321346595256313,
+      "grad_norm": 2.1460361773704255,
+      "learning_rate": 6.785654235890338e-06,
+      "loss": 0.3688,
+      "step": 8052
+    },
+    {
+      "epoch": 1.2322876817138484,
+      "grad_norm": 2.5735775872854028,
+      "learning_rate": 6.7833076643072085e-06,
+      "loss": 0.4729,
+      "step": 8053
+    },
+    {
+      "epoch": 1.2324407039020657,
+      "grad_norm": 2.332483797336673,
+      "learning_rate": 6.78096129027374e-06,
+      "loss": 0.4682,
+      "step": 8054
+    },
+    {
+      "epoch": 1.232593726090283,
+      "grad_norm": 2.271105481517045,
+      "learning_rate": 6.778615113934031e-06,
+      "loss": 0.4461,
+      "step": 8055
+    },
+    {
+      "epoch": 1.2327467482785004,
+      "grad_norm": 2.1420714053603236,
+      "learning_rate": 6.776269135432174e-06,
+      "loss": 0.3699,
+      "step": 8056
+    },
+    {
+      "epoch": 1.2328997704667177,
+      "grad_norm": 2.333313250766813,
+      "learning_rate": 6.773923354912239e-06,
+      "loss": 0.349,
+      "step": 8057
+    },
+    {
+      "epoch": 1.233052792654935,
+      "grad_norm": 2.3201571423666594,
+      "learning_rate": 6.7715777725182875e-06,
+      "loss": 0.4587,
+      "step": 8058
+    },
+    {
+      "epoch": 1.2332058148431522,
+      "grad_norm": 2.350700286167741,
+      "learning_rate": 6.76923238839438e-06,
+      "loss": 0.4382,
+      "step": 8059
+    },
+    {
+      "epoch": 1.2333588370313695,
+      "grad_norm": 2.4481880445812005,
+      "learning_rate": 6.766887202684549e-06,
+      "loss": 0.4408,
+      "step": 8060
+    },
+    {
+      "epoch": 1.2335118592195868,
+      "grad_norm": 2.1630515616270882,
+      "learning_rate": 6.76454221553282e-06,
+      "loss": 0.3706,
+      "step": 8061
+    },
+    {
+      "epoch": 1.2336648814078042,
+      "grad_norm": 2.0092456944627766,
+      "learning_rate": 6.7621974270832145e-06,
+      "loss": 0.3611,
+      "step": 8062
+    },
+    {
+      "epoch": 1.2338179035960215,
+      "grad_norm": 2.1523204344513838,
+      "learning_rate": 6.7598528374797275e-06,
+      "loss": 0.386,
+      "step": 8063
+    },
+    {
+      "epoch": 1.2339709257842388,
+      "grad_norm": 2.3086250501937777,
+      "learning_rate": 6.757508446866357e-06,
+      "loss": 0.4011,
+      "step": 8064
+    },
+    {
+      "epoch": 1.234123947972456,
+      "grad_norm": 2.4251052079549575,
+      "learning_rate": 6.755164255387078e-06,
+      "loss": 0.3798,
+      "step": 8065
+    },
+    {
+      "epoch": 1.2342769701606733,
+      "grad_norm": 2.24168987184976,
+      "learning_rate": 6.752820263185851e-06,
+      "loss": 0.3947,
+      "step": 8066
+    },
+    {
+      "epoch": 1.2344299923488906,
+      "grad_norm": 2.097126542099647,
+      "learning_rate": 6.750476470406638e-06,
+      "loss": 0.4268,
+      "step": 8067
+    },
+    {
+      "epoch": 1.234583014537108,
+      "grad_norm": 2.289927186551288,
+      "learning_rate": 6.748132877193379e-06,
+      "loss": 0.3988,
+      "step": 8068
+    },
+    {
+      "epoch": 1.2347360367253253,
+      "grad_norm": 2.122315650716571,
+      "learning_rate": 6.745789483689997e-06,
+      "loss": 0.4257,
+      "step": 8069
+    },
+    {
+      "epoch": 1.2348890589135424,
+      "grad_norm": 1.9968670376809796,
+      "learning_rate": 6.743446290040417e-06,
+      "loss": 0.4458,
+      "step": 8070
+    },
+    {
+      "epoch": 1.2350420811017597,
+      "grad_norm": 1.8989966989084186,
+      "learning_rate": 6.7411032963885355e-06,
+      "loss": 0.411,
+      "step": 8071
+    },
+    {
+      "epoch": 1.235195103289977,
+      "grad_norm": 2.1196167992227637,
+      "learning_rate": 6.738760502878254e-06,
+      "loss": 0.3371,
+      "step": 8072
+    },
+    {
+      "epoch": 1.2353481254781944,
+      "grad_norm": 2.27319423233866,
+      "learning_rate": 6.736417909653448e-06,
+      "loss": 0.4286,
+      "step": 8073
+    },
+    {
+      "epoch": 1.2355011476664117,
+      "grad_norm": 2.1531922370964884,
+      "learning_rate": 6.734075516857981e-06,
+      "loss": 0.3935,
+      "step": 8074
+    },
+    {
+      "epoch": 1.2356541698546288,
+      "grad_norm": 2.169345191566362,
+      "learning_rate": 6.731733324635717e-06,
+      "loss": 0.3872,
+      "step": 8075
+    },
+    {
+      "epoch": 1.2358071920428462,
+      "grad_norm": 2.3456728853488804,
+      "learning_rate": 6.729391333130495e-06,
+      "loss": 0.4262,
+      "step": 8076
+    },
+    {
+      "epoch": 1.2359602142310635,
+      "grad_norm": 2.116438016922821,
+      "learning_rate": 6.727049542486142e-06,
+      "loss": 0.3507,
+      "step": 8077
+    },
+    {
+      "epoch": 1.2361132364192808,
+      "grad_norm": 2.2905850835193675,
+      "learning_rate": 6.724707952846485e-06,
+      "loss": 0.474,
+      "step": 8078
+    },
+    {
+      "epoch": 1.2362662586074982,
+      "grad_norm": 2.126345341606041,
+      "learning_rate": 6.722366564355322e-06,
+      "loss": 0.4478,
+      "step": 8079
+    },
+    {
+      "epoch": 1.2364192807957153,
+      "grad_norm": 2.2306359092945516,
+      "learning_rate": 6.720025377156453e-06,
+      "loss": 0.4082,
+      "step": 8080
+    },
+    {
+      "epoch": 1.2365723029839326,
+      "grad_norm": 2.4936280034611946,
+      "learning_rate": 6.7176843913936584e-06,
+      "loss": 0.3708,
+      "step": 8081
+    },
+    {
+      "epoch": 1.23672532517215,
+      "grad_norm": 2.060763607982872,
+      "learning_rate": 6.715343607210701e-06,
+      "loss": 0.4245,
+      "step": 8082
+    },
+    {
+      "epoch": 1.2368783473603673,
+      "grad_norm": 2.4140780179422334,
+      "learning_rate": 6.713003024751345e-06,
+      "loss": 0.4348,
+      "step": 8083
+    },
+    {
+      "epoch": 1.2370313695485846,
+      "grad_norm": 2.0437392400835552,
+      "learning_rate": 6.7106626441593315e-06,
+      "loss": 0.4317,
+      "step": 8084
+    },
+    {
+      "epoch": 1.2371843917368017,
+      "grad_norm": 2.166562578027147,
+      "learning_rate": 6.708322465578394e-06,
+      "loss": 0.412,
+      "step": 8085
+    },
+    {
+      "epoch": 1.237337413925019,
+      "grad_norm": 2.1138549681897807,
+      "learning_rate": 6.705982489152252e-06,
+      "loss": 0.4011,
+      "step": 8086
+    },
+    {
+      "epoch": 1.2374904361132364,
+      "grad_norm": 2.031131307063908,
+      "learning_rate": 6.703642715024608e-06,
+      "loss": 0.3462,
+      "step": 8087
+    },
+    {
+      "epoch": 1.2376434583014537,
+      "grad_norm": 2.1949317305374674,
+      "learning_rate": 6.701303143339164e-06,
+      "loss": 0.3736,
+      "step": 8088
+    },
+    {
+      "epoch": 1.237796480489671,
+      "grad_norm": 2.1993897401590603,
+      "learning_rate": 6.6989637742395975e-06,
+      "loss": 0.3788,
+      "step": 8089
+    },
+    {
+      "epoch": 1.2379495026778882,
+      "grad_norm": 2.5296632495849334,
+      "learning_rate": 6.696624607869576e-06,
+      "loss": 0.4256,
+      "step": 8090
+    },
+    {
+      "epoch": 1.2381025248661055,
+      "grad_norm": 2.2823327820853794,
+      "learning_rate": 6.694285644372765e-06,
+      "loss": 0.3772,
+      "step": 8091
+    },
+    {
+      "epoch": 1.2382555470543228,
+      "grad_norm": 2.0040278866366528,
+      "learning_rate": 6.691946883892802e-06,
+      "loss": 0.3082,
+      "step": 8092
+    },
+    {
+      "epoch": 1.2384085692425402,
+      "grad_norm": 1.928916255534059,
+      "learning_rate": 6.689608326573322e-06,
+      "loss": 0.4043,
+      "step": 8093
+    },
+    {
+      "epoch": 1.2385615914307575,
+      "grad_norm": 1.92290953923648,
+      "learning_rate": 6.6872699725579485e-06,
+      "loss": 0.3487,
+      "step": 8094
+    },
+    {
+      "epoch": 1.2387146136189748,
+      "grad_norm": 2.3340519002998414,
+      "learning_rate": 6.684931821990279e-06,
+      "loss": 0.4364,
+      "step": 8095
+    },
+    {
+      "epoch": 1.238867635807192,
+      "grad_norm": 2.212440455612955,
+      "learning_rate": 6.6825938750139185e-06,
+      "loss": 0.4273,
+      "step": 8096
+    },
+    {
+      "epoch": 1.2390206579954093,
+      "grad_norm": 2.111364367290438,
+      "learning_rate": 6.680256131772449e-06,
+      "loss": 0.3948,
+      "step": 8097
+    },
+    {
+      "epoch": 1.2391736801836266,
+      "grad_norm": 2.2003773037744505,
+      "learning_rate": 6.677918592409432e-06,
+      "loss": 0.4448,
+      "step": 8098
+    },
+    {
+      "epoch": 1.239326702371844,
+      "grad_norm": 1.8536781014593984,
+      "learning_rate": 6.675581257068433e-06,
+      "loss": 0.3131,
+      "step": 8099
+    },
+    {
+      "epoch": 1.2394797245600613,
+      "grad_norm": 2.0146176424945805,
+      "learning_rate": 6.673244125892991e-06,
+      "loss": 0.3832,
+      "step": 8100
+    },
+    {
+      "epoch": 1.2396327467482786,
+      "grad_norm": 2.346769756302333,
+      "learning_rate": 6.670907199026647e-06,
+      "loss": 0.4161,
+      "step": 8101
+    },
+    {
+      "epoch": 1.2397857689364957,
+      "grad_norm": 2.1642485454000058,
+      "learning_rate": 6.668570476612912e-06,
+      "loss": 0.344,
+      "step": 8102
+    },
+    {
+      "epoch": 1.239938791124713,
+      "grad_norm": 2.0988507243659997,
+      "learning_rate": 6.6662339587952966e-06,
+      "loss": 0.3651,
+      "step": 8103
+    },
+    {
+      "epoch": 1.2400918133129304,
+      "grad_norm": 1.9921147928691025,
+      "learning_rate": 6.6638976457172985e-06,
+      "loss": 0.3652,
+      "step": 8104
+    },
+    {
+      "epoch": 1.2402448355011477,
+      "grad_norm": 2.0682253549190404,
+      "learning_rate": 6.6615615375223945e-06,
+      "loss": 0.3633,
+      "step": 8105
+    },
+    {
+      "epoch": 1.240397857689365,
+      "grad_norm": 2.0394982635306493,
+      "learning_rate": 6.659225634354055e-06,
+      "loss": 0.3469,
+      "step": 8106
+    },
+    {
+      "epoch": 1.2405508798775822,
+      "grad_norm": 2.017362842165284,
+      "learning_rate": 6.656889936355742e-06,
+      "loss": 0.3196,
+      "step": 8107
+    },
+    {
+      "epoch": 1.2407039020657995,
+      "grad_norm": 2.1998384971210565,
+      "learning_rate": 6.654554443670893e-06,
+      "loss": 0.4183,
+      "step": 8108
+    },
+    {
+      "epoch": 1.2408569242540168,
+      "grad_norm": 2.194051345196769,
+      "learning_rate": 6.652219156442945e-06,
+      "loss": 0.3716,
+      "step": 8109
+    },
+    {
+      "epoch": 1.2410099464422342,
+      "grad_norm": 1.8924912286597517,
+      "learning_rate": 6.649884074815316e-06,
+      "loss": 0.3739,
+      "step": 8110
+    },
+    {
+      "epoch": 1.2411629686304515,
+      "grad_norm": 2.105527117813446,
+      "learning_rate": 6.647549198931407e-06,
+      "loss": 0.358,
+      "step": 8111
+    },
+    {
+      "epoch": 1.2413159908186686,
+      "grad_norm": 2.2597521686833333,
+      "learning_rate": 6.645214528934623e-06,
+      "loss": 0.3437,
+      "step": 8112
+    },
+    {
+      "epoch": 1.241469013006886,
+      "grad_norm": 2.200249191211589,
+      "learning_rate": 6.6428800649683374e-06,
+      "loss": 0.3841,
+      "step": 8113
+    },
+    {
+      "epoch": 1.2416220351951033,
+      "grad_norm": 2.3942357392351474,
+      "learning_rate": 6.640545807175914e-06,
+      "loss": 0.4409,
+      "step": 8114
+    },
+    {
+      "epoch": 1.2417750573833206,
+      "grad_norm": 2.2397158209349715,
+      "learning_rate": 6.638211755700722e-06,
+      "loss": 0.4461,
+      "step": 8115
+    },
+    {
+      "epoch": 1.241928079571538,
+      "grad_norm": 2.348508538360569,
+      "learning_rate": 6.6358779106860924e-06,
+      "loss": 0.4139,
+      "step": 8116
+    },
+    {
+      "epoch": 1.242081101759755,
+      "grad_norm": 2.189259772021628,
+      "learning_rate": 6.633544272275364e-06,
+      "loss": 0.4399,
+      "step": 8117
+    },
+    {
+      "epoch": 1.2422341239479724,
+      "grad_norm": 2.0453409050667197,
+      "learning_rate": 6.631210840611852e-06,
+      "loss": 0.4082,
+      "step": 8118
+    },
+    {
+      "epoch": 1.2423871461361897,
+      "grad_norm": 2.639988084591451,
+      "learning_rate": 6.628877615838859e-06,
+      "loss": 0.4131,
+      "step": 8119
+    },
+    {
+      "epoch": 1.242540168324407,
+      "grad_norm": 2.4274196063214677,
+      "learning_rate": 6.62654459809968e-06,
+      "loss": 0.4368,
+      "step": 8120
+    },
+    {
+      "epoch": 1.2426931905126244,
+      "grad_norm": 2.14359024994293,
+      "learning_rate": 6.624211787537594e-06,
+      "loss": 0.4156,
+      "step": 8121
+    },
+    {
+      "epoch": 1.2428462127008415,
+      "grad_norm": 2.292958397908944,
+      "learning_rate": 6.621879184295869e-06,
+      "loss": 0.4314,
+      "step": 8122
+    },
+    {
+      "epoch": 1.2429992348890588,
+      "grad_norm": 1.9339362064576864,
+      "learning_rate": 6.619546788517759e-06,
+      "loss": 0.3408,
+      "step": 8123
+    },
+    {
+      "epoch": 1.2431522570772762,
+      "grad_norm": 2.3692855461901927,
+      "learning_rate": 6.617214600346503e-06,
+      "loss": 0.4277,
+      "step": 8124
+    },
+    {
+      "epoch": 1.2433052792654935,
+      "grad_norm": 2.3900418782194226,
+      "learning_rate": 6.614882619925336e-06,
+      "loss": 0.4539,
+      "step": 8125
+    },
+    {
+      "epoch": 1.2434583014537108,
+      "grad_norm": 2.1281143454275444,
+      "learning_rate": 6.612550847397468e-06,
+      "loss": 0.3837,
+      "step": 8126
+    },
+    {
+      "epoch": 1.2436113236419282,
+      "grad_norm": 1.9287068128994407,
+      "learning_rate": 6.610219282906101e-06,
+      "loss": 0.391,
+      "step": 8127
+    },
+    {
+      "epoch": 1.2437643458301453,
+      "grad_norm": 2.2439153783777033,
+      "learning_rate": 6.607887926594434e-06,
+      "loss": 0.3952,
+      "step": 8128
+    },
+    {
+      "epoch": 1.2439173680183626,
+      "grad_norm": 2.4364819100764668,
+      "learning_rate": 6.605556778605635e-06,
+      "loss": 0.4225,
+      "step": 8129
+    },
+    {
+      "epoch": 1.24407039020658,
+      "grad_norm": 1.950810491447188,
+      "learning_rate": 6.603225839082875e-06,
+      "loss": 0.3582,
+      "step": 8130
+    },
+    {
+      "epoch": 1.2442234123947973,
+      "grad_norm": 2.1343110248130905,
+      "learning_rate": 6.600895108169308e-06,
+      "loss": 0.3882,
+      "step": 8131
+    },
+    {
+      "epoch": 1.2443764345830146,
+      "grad_norm": 2.133074845103394,
+      "learning_rate": 6.598564586008064e-06,
+      "loss": 0.3467,
+      "step": 8132
+    },
+    {
+      "epoch": 1.244529456771232,
+      "grad_norm": 2.4865499659321606,
+      "learning_rate": 6.596234272742278e-06,
+      "loss": 0.3426,
+      "step": 8133
+    },
+    {
+      "epoch": 1.244682478959449,
+      "grad_norm": 2.2551479642677403,
+      "learning_rate": 6.593904168515062e-06,
+      "loss": 0.3678,
+      "step": 8134
+    },
+    {
+      "epoch": 1.2448355011476664,
+      "grad_norm": 2.4241006823885494,
+      "learning_rate": 6.591574273469513e-06,
+      "loss": 0.3655,
+      "step": 8135
+    },
+    {
+      "epoch": 1.2449885233358837,
+      "grad_norm": 1.9992035540833695,
+      "learning_rate": 6.589244587748723e-06,
+      "loss": 0.3564,
+      "step": 8136
+    },
+    {
+      "epoch": 1.245141545524101,
+      "grad_norm": 2.3241941093092335,
+      "learning_rate": 6.586915111495764e-06,
+      "loss": 0.4287,
+      "step": 8137
+    },
+    {
+      "epoch": 1.2452945677123184,
+      "grad_norm": 2.287427281357684,
+      "learning_rate": 6.5845858448537045e-06,
+      "loss": 0.4589,
+      "step": 8138
+    },
+    {
+      "epoch": 1.2454475899005355,
+      "grad_norm": 1.857212005781437,
+      "learning_rate": 6.582256787965586e-06,
+      "loss": 0.3074,
+      "step": 8139
+    },
+    {
+      "epoch": 1.2456006120887528,
+      "grad_norm": 2.195369821049141,
+      "learning_rate": 6.579927940974449e-06,
+      "loss": 0.3611,
+      "step": 8140
+    },
+    {
+      "epoch": 1.2457536342769702,
+      "grad_norm": 2.199397634076349,
+      "learning_rate": 6.577599304023319e-06,
+      "loss": 0.3836,
+      "step": 8141
+    },
+    {
+      "epoch": 1.2459066564651875,
+      "grad_norm": 2.3932894708522023,
+      "learning_rate": 6.575270877255202e-06,
+      "loss": 0.4379,
+      "step": 8142
+    },
+    {
+      "epoch": 1.2460596786534048,
+      "grad_norm": 2.4586397278695524,
+      "learning_rate": 6.572942660813096e-06,
+      "loss": 0.4097,
+      "step": 8143
+    },
+    {
+      "epoch": 1.246212700841622,
+      "grad_norm": 2.2564384469631396,
+      "learning_rate": 6.570614654839993e-06,
+      "loss": 0.616,
+      "step": 8144
+    },
+    {
+      "epoch": 1.2463657230298393,
+      "grad_norm": 2.027426896315064,
+      "learning_rate": 6.568286859478854e-06,
+      "loss": 0.2986,
+      "step": 8145
+    },
+    {
+      "epoch": 1.2465187452180566,
+      "grad_norm": 1.9998258578563708,
+      "learning_rate": 6.565959274872648e-06,
+      "loss": 0.3829,
+      "step": 8146
+    },
+    {
+      "epoch": 1.246671767406274,
+      "grad_norm": 2.294962618851666,
+      "learning_rate": 6.563631901164318e-06,
+      "loss": 0.3712,
+      "step": 8147
+    },
+    {
+      "epoch": 1.2468247895944913,
+      "grad_norm": 2.2436925088267703,
+      "learning_rate": 6.561304738496792e-06,
+      "loss": 0.3324,
+      "step": 8148
+    },
+    {
+      "epoch": 1.2469778117827084,
+      "grad_norm": 2.4639587946632284,
+      "learning_rate": 6.558977787012997e-06,
+      "loss": 0.4681,
+      "step": 8149
+    },
+    {
+      "epoch": 1.2471308339709257,
+      "grad_norm": 2.381414016939421,
+      "learning_rate": 6.556651046855838e-06,
+      "loss": 0.3656,
+      "step": 8150
+    },
+    {
+      "epoch": 1.247283856159143,
+      "grad_norm": 1.8783872615246844,
+      "learning_rate": 6.554324518168206e-06,
+      "loss": 0.2745,
+      "step": 8151
+    },
+    {
+      "epoch": 1.2474368783473604,
+      "grad_norm": 2.259284316104315,
+      "learning_rate": 6.551998201092986e-06,
+      "loss": 0.44,
+      "step": 8152
+    },
+    {
+      "epoch": 1.2475899005355777,
+      "grad_norm": 2.0064902849502424,
+      "learning_rate": 6.549672095773043e-06,
+      "loss": 0.3091,
+      "step": 8153
+    },
+    {
+      "epoch": 1.2477429227237948,
+      "grad_norm": 2.0524252981887967,
+      "learning_rate": 6.547346202351239e-06,
+      "loss": 0.3426,
+      "step": 8154
+    },
+    {
+      "epoch": 1.2478959449120122,
+      "grad_norm": 2.0469596722487564,
+      "learning_rate": 6.54502052097041e-06,
+      "loss": 0.3511,
+      "step": 8155
+    },
+    {
+      "epoch": 1.2480489671002295,
+      "grad_norm": 2.2105301690125487,
+      "learning_rate": 6.542695051773384e-06,
+      "loss": 0.3652,
+      "step": 8156
+    },
+    {
+      "epoch": 1.2482019892884468,
+      "grad_norm": 2.0530242393989413,
+      "learning_rate": 6.540369794902984e-06,
+      "loss": 0.3917,
+      "step": 8157
+    },
+    {
+      "epoch": 1.2483550114766642,
+      "grad_norm": 1.936389285927461,
+      "learning_rate": 6.538044750502005e-06,
+      "loss": 0.3831,
+      "step": 8158
+    },
+    {
+      "epoch": 1.2485080336648815,
+      "grad_norm": 2.1162540345200056,
+      "learning_rate": 6.535719918713244e-06,
+      "loss": 0.3493,
+      "step": 8159
+    },
+    {
+      "epoch": 1.2486610558530986,
+      "grad_norm": 2.275333849051231,
+      "learning_rate": 6.533395299679475e-06,
+      "loss": 0.4457,
+      "step": 8160
+    },
+    {
+      "epoch": 1.248814078041316,
+      "grad_norm": 2.150509646064332,
+      "learning_rate": 6.531070893543461e-06,
+      "loss": 0.3997,
+      "step": 8161
+    },
+    {
+      "epoch": 1.2489671002295333,
+      "grad_norm": 2.0081462458065764,
+      "learning_rate": 6.5287467004479545e-06,
+      "loss": 0.3871,
+      "step": 8162
+    },
+    {
+      "epoch": 1.2491201224177506,
+      "grad_norm": 2.5256472370868592,
+      "learning_rate": 6.526422720535694e-06,
+      "loss": 0.4369,
+      "step": 8163
+    },
+    {
+      "epoch": 1.249273144605968,
+      "grad_norm": 2.264997586366189,
+      "learning_rate": 6.5240989539494e-06,
+      "loss": 0.3933,
+      "step": 8164
+    },
+    {
+      "epoch": 1.2494261667941853,
+      "grad_norm": 2.0935312426949495,
+      "learning_rate": 6.521775400831789e-06,
+      "loss": 0.4138,
+      "step": 8165
+    },
+    {
+      "epoch": 1.2495791889824024,
+      "grad_norm": 1.993012753242544,
+      "learning_rate": 6.519452061325555e-06,
+      "loss": 0.3732,
+      "step": 8166
+    },
+    {
+      "epoch": 1.2497322111706197,
+      "grad_norm": 2.0981709407547404,
+      "learning_rate": 6.517128935573389e-06,
+      "loss": 0.3587,
+      "step": 8167
+    },
+    {
+      "epoch": 1.249885233358837,
+      "grad_norm": 2.126075996875917,
+      "learning_rate": 6.51480602371796e-06,
+      "loss": 0.4134,
+      "step": 8168
+    },
+    {
+      "epoch": 1.2500382555470544,
+      "grad_norm": 2.2515728182518178,
+      "learning_rate": 6.5124833259019235e-06,
+      "loss": 0.4175,
+      "step": 8169
+    },
+    {
+      "epoch": 1.2501912777352717,
+      "grad_norm": 2.1486259337174594,
+      "learning_rate": 6.510160842267935e-06,
+      "loss": 0.3988,
+      "step": 8170
+    },
+    {
+      "epoch": 1.2503442999234888,
+      "grad_norm": 2.1867889880458597,
+      "learning_rate": 6.507838572958619e-06,
+      "loss": 0.3892,
+      "step": 8171
+    },
+    {
+      "epoch": 1.2504973221117062,
+      "grad_norm": 2.2552520050421427,
+      "learning_rate": 6.505516518116595e-06,
+      "loss": 0.3685,
+      "step": 8172
+    },
+    {
+      "epoch": 1.2506503442999235,
+      "grad_norm": 2.1887592404297695,
+      "learning_rate": 6.503194677884477e-06,
+      "loss": 0.4032,
+      "step": 8173
+    },
+    {
+      "epoch": 1.2508033664881408,
+      "grad_norm": 2.124074218625281,
+      "learning_rate": 6.500873052404848e-06,
+      "loss": 0.364,
+      "step": 8174
+    },
+    {
+      "epoch": 1.2509563886763582,
+      "grad_norm": 2.333056411610224,
+      "learning_rate": 6.4985516418202965e-06,
+      "loss": 0.3767,
+      "step": 8175
+    },
+    {
+      "epoch": 1.2511094108645753,
+      "grad_norm": 2.213655759545077,
+      "learning_rate": 6.496230446273389e-06,
+      "loss": 0.3687,
+      "step": 8176
+    },
+    {
+      "epoch": 1.2512624330527926,
+      "grad_norm": 2.192929258564723,
+      "learning_rate": 6.493909465906671e-06,
+      "loss": 0.3897,
+      "step": 8177
+    },
+    {
+      "epoch": 1.25141545524101,
+      "grad_norm": 1.9952666186839858,
+      "learning_rate": 6.491588700862692e-06,
+      "loss": 0.3551,
+      "step": 8178
+    },
+    {
+      "epoch": 1.2515684774292273,
+      "grad_norm": 2.2618334986409523,
+      "learning_rate": 6.489268151283978e-06,
+      "loss": 0.3391,
+      "step": 8179
+    },
+    {
+      "epoch": 1.2517214996174446,
+      "grad_norm": 2.0576909482129793,
+      "learning_rate": 6.486947817313036e-06,
+      "loss": 0.3226,
+      "step": 8180
+    },
+    {
+      "epoch": 1.2518745218056617,
+      "grad_norm": 2.7975000133563808,
+      "learning_rate": 6.484627699092374e-06,
+      "loss": 0.4377,
+      "step": 8181
+    },
+    {
+      "epoch": 1.252027543993879,
+      "grad_norm": 2.2709472936432453,
+      "learning_rate": 6.482307796764476e-06,
+      "loss": 0.3881,
+      "step": 8182
+    },
+    {
+      "epoch": 1.2521805661820964,
+      "grad_norm": 2.331279784622257,
+      "learning_rate": 6.4799881104718194e-06,
+      "loss": 0.4555,
+      "step": 8183
+    },
+    {
+      "epoch": 1.2523335883703137,
+      "grad_norm": 1.9791757619354962,
+      "learning_rate": 6.477668640356862e-06,
+      "loss": 0.3695,
+      "step": 8184
+    },
+    {
+      "epoch": 1.252486610558531,
+      "grad_norm": 2.2578902118790785,
+      "learning_rate": 6.475349386562052e-06,
+      "loss": 0.3684,
+      "step": 8185
+    },
+    {
+      "epoch": 1.2526396327467482,
+      "grad_norm": 2.0440767072571435,
+      "learning_rate": 6.4730303492298276e-06,
+      "loss": 0.3873,
+      "step": 8186
+    },
+    {
+      "epoch": 1.2527926549349655,
+      "grad_norm": 2.2598153800347975,
+      "learning_rate": 6.4707115285026065e-06,
+      "loss": 0.4313,
+      "step": 8187
+    },
+    {
+      "epoch": 1.2529456771231828,
+      "grad_norm": 2.2472633970388163,
+      "learning_rate": 6.468392924522793e-06,
+      "loss": 0.3577,
+      "step": 8188
+    },
+    {
+      "epoch": 1.2530986993114002,
+      "grad_norm": 2.000066195188573,
+      "learning_rate": 6.4660745374327916e-06,
+      "loss": 0.3561,
+      "step": 8189
+    },
+    {
+      "epoch": 1.2532517214996175,
+      "grad_norm": 1.998668284331149,
+      "learning_rate": 6.46375636737497e-06,
+      "loss": 0.3622,
+      "step": 8190
+    },
+    {
+      "epoch": 1.2534047436878346,
+      "grad_norm": 2.2063495642394155,
+      "learning_rate": 6.461438414491712e-06,
+      "loss": 0.4527,
+      "step": 8191
+    },
+    {
+      "epoch": 1.2535577658760522,
+      "grad_norm": 2.246557833214377,
+      "learning_rate": 6.459120678925361e-06,
+      "loss": 0.4203,
+      "step": 8192
+    },
+    {
+      "epoch": 1.2537107880642693,
+      "grad_norm": 2.2063828436285715,
+      "learning_rate": 6.456803160818257e-06,
+      "loss": 0.3825,
+      "step": 8193
+    },
+    {
+      "epoch": 1.2538638102524866,
+      "grad_norm": 1.9092101174408787,
+      "learning_rate": 6.454485860312738e-06,
+      "loss": 0.3348,
+      "step": 8194
+    },
+    {
+      "epoch": 1.254016832440704,
+      "grad_norm": 2.0977756057796055,
+      "learning_rate": 6.452168777551108e-06,
+      "loss": 0.4035,
+      "step": 8195
+    },
+    {
+      "epoch": 1.2541698546289213,
+      "grad_norm": 2.1005447185578676,
+      "learning_rate": 6.449851912675674e-06,
+      "loss": 0.3674,
+      "step": 8196
+    },
+    {
+      "epoch": 1.2543228768171386,
+      "grad_norm": 2.558298193490183,
+      "learning_rate": 6.447535265828726e-06,
+      "loss": 0.4957,
+      "step": 8197
+    },
+    {
+      "epoch": 1.2544758990053557,
+      "grad_norm": 2.024092000979385,
+      "learning_rate": 6.44521883715253e-06,
+      "loss": 0.3693,
+      "step": 8198
+    },
+    {
+      "epoch": 1.254628921193573,
+      "grad_norm": 2.2401108910393646,
+      "learning_rate": 6.4429026267893515e-06,
+      "loss": 0.3757,
+      "step": 8199
+    },
+    {
+      "epoch": 1.2547819433817904,
+      "grad_norm": 2.2399359164291486,
+      "learning_rate": 6.440586634881443e-06,
+      "loss": 0.3903,
+      "step": 8200
+    },
+    {
+      "epoch": 1.2549349655700077,
+      "grad_norm": 2.231638113223591,
+      "learning_rate": 6.438270861571028e-06,
+      "loss": 0.3898,
+      "step": 8201
+    },
+    {
+      "epoch": 1.255087987758225,
+      "grad_norm": 2.3003887351119485,
+      "learning_rate": 6.435955307000334e-06,
+      "loss": 0.4559,
+      "step": 8202
+    },
+    {
+      "epoch": 1.2552410099464422,
+      "grad_norm": 2.342065540275981,
+      "learning_rate": 6.4336399713115675e-06,
+      "loss": 0.4731,
+      "step": 8203
+    },
+    {
+      "epoch": 1.2553940321346595,
+      "grad_norm": 2.159649433148729,
+      "learning_rate": 6.431324854646925e-06,
+      "loss": 0.4246,
+      "step": 8204
+    },
+    {
+      "epoch": 1.2555470543228768,
+      "grad_norm": 2.1847647079301815,
+      "learning_rate": 6.429009957148582e-06,
+      "loss": 0.4069,
+      "step": 8205
+    },
+    {
+      "epoch": 1.2557000765110942,
+      "grad_norm": 2.1882504529272904,
+      "learning_rate": 6.426695278958706e-06,
+      "loss": 0.3894,
+      "step": 8206
+    },
+    {
+      "epoch": 1.2558530986993115,
+      "grad_norm": 2.09420529387032,
+      "learning_rate": 6.424380820219455e-06,
+      "loss": 0.3681,
+      "step": 8207
+    },
+    {
+      "epoch": 1.2560061208875286,
+      "grad_norm": 2.428772300102065,
+      "learning_rate": 6.422066581072964e-06,
+      "loss": 0.4136,
+      "step": 8208
+    },
+    {
+      "epoch": 1.256159143075746,
+      "grad_norm": 2.213077828037295,
+      "learning_rate": 6.419752561661358e-06,
+      "loss": 0.3422,
+      "step": 8209
+    },
+    {
+      "epoch": 1.2563121652639633,
+      "grad_norm": 2.0985586609316913,
+      "learning_rate": 6.41743876212676e-06,
+      "loss": 0.4081,
+      "step": 8210
+    },
+    {
+      "epoch": 1.2564651874521806,
+      "grad_norm": 2.2509752229595543,
+      "learning_rate": 6.415125182611257e-06,
+      "loss": 0.4167,
+      "step": 8211
+    },
+    {
+      "epoch": 1.256618209640398,
+      "grad_norm": 1.9046730532077025,
+      "learning_rate": 6.412811823256942e-06,
+      "loss": 0.3838,
+      "step": 8212
+    },
+    {
+      "epoch": 1.256771231828615,
+      "grad_norm": 2.2792001975357454,
+      "learning_rate": 6.410498684205889e-06,
+      "loss": 0.4713,
+      "step": 8213
+    },
+    {
+      "epoch": 1.2569242540168324,
+      "grad_norm": 2.3477391524234297,
+      "learning_rate": 6.40818576560015e-06,
+      "loss": 0.4213,
+      "step": 8214
+    },
+    {
+      "epoch": 1.2570772762050497,
+      "grad_norm": 2.1809286043256364,
+      "learning_rate": 6.405873067581776e-06,
+      "loss": 0.3181,
+      "step": 8215
+    },
+    {
+      "epoch": 1.257230298393267,
+      "grad_norm": 2.2438204832034225,
+      "learning_rate": 6.4035605902927985e-06,
+      "loss": 0.3751,
+      "step": 8216
+    },
+    {
+      "epoch": 1.2573833205814844,
+      "grad_norm": 2.2567114261781756,
+      "learning_rate": 6.401248333875232e-06,
+      "loss": 0.3622,
+      "step": 8217
+    },
+    {
+      "epoch": 1.2575363427697015,
+      "grad_norm": 1.9982564455204277,
+      "learning_rate": 6.398936298471086e-06,
+      "loss": 0.3942,
+      "step": 8218
+    },
+    {
+      "epoch": 1.2576893649579188,
+      "grad_norm": 2.2711651757471576,
+      "learning_rate": 6.3966244842223455e-06,
+      "loss": 0.4798,
+      "step": 8219
+    },
+    {
+      "epoch": 1.2578423871461362,
+      "grad_norm": 2.6053006150189892,
+      "learning_rate": 6.394312891270997e-06,
+      "loss": 0.4886,
+      "step": 8220
+    },
+    {
+      "epoch": 1.2579954093343535,
+      "grad_norm": 2.1850007261040543,
+      "learning_rate": 6.392001519758997e-06,
+      "loss": 0.3153,
+      "step": 8221
+    },
+    {
+      "epoch": 1.2581484315225708,
+      "grad_norm": 1.9624789585008473,
+      "learning_rate": 6.389690369828297e-06,
+      "loss": 0.3293,
+      "step": 8222
+    },
+    {
+      "epoch": 1.258301453710788,
+      "grad_norm": 1.8446939371406101,
+      "learning_rate": 6.387379441620838e-06,
+      "loss": 0.3596,
+      "step": 8223
+    },
+    {
+      "epoch": 1.2584544758990053,
+      "grad_norm": 1.9329320390958946,
+      "learning_rate": 6.3850687352785365e-06,
+      "loss": 0.3903,
+      "step": 8224
+    },
+    {
+      "epoch": 1.2586074980872226,
+      "grad_norm": 2.18800037719509,
+      "learning_rate": 6.3827582509433085e-06,
+      "loss": 0.4227,
+      "step": 8225
+    },
+    {
+      "epoch": 1.25876052027544,
+      "grad_norm": 1.9241949681406423,
+      "learning_rate": 6.380447988757048e-06,
+      "loss": 0.3641,
+      "step": 8226
+    },
+    {
+      "epoch": 1.2589135424636573,
+      "grad_norm": 2.172606220056416,
+      "learning_rate": 6.378137948861633e-06,
+      "loss": 0.3112,
+      "step": 8227
+    },
+    {
+      "epoch": 1.2590665646518744,
+      "grad_norm": 2.5231403921304647,
+      "learning_rate": 6.375828131398937e-06,
+      "loss": 0.4162,
+      "step": 8228
+    },
+    {
+      "epoch": 1.259219586840092,
+      "grad_norm": 2.196984189818374,
+      "learning_rate": 6.373518536510815e-06,
+      "loss": 0.4081,
+      "step": 8229
+    },
+    {
+      "epoch": 1.259372609028309,
+      "grad_norm": 2.262628444670184,
+      "learning_rate": 6.371209164339103e-06,
+      "loss": 0.3758,
+      "step": 8230
+    },
+    {
+      "epoch": 1.2595256312165264,
+      "grad_norm": 2.1348336472126883,
+      "learning_rate": 6.3689000150256345e-06,
+      "loss": 0.4025,
+      "step": 8231
+    },
+    {
+      "epoch": 1.2596786534047437,
+      "grad_norm": 2.2223035828990434,
+      "learning_rate": 6.3665910887122175e-06,
+      "loss": 0.3868,
+      "step": 8232
+    },
+    {
+      "epoch": 1.259831675592961,
+      "grad_norm": 2.3435045332318207,
+      "learning_rate": 6.364282385540661e-06,
+      "loss": 0.3535,
+      "step": 8233
+    },
+    {
+      "epoch": 1.2599846977811784,
+      "grad_norm": 2.110334468447516,
+      "learning_rate": 6.361973905652743e-06,
+      "loss": 0.3774,
+      "step": 8234
+    },
+    {
+      "epoch": 1.2601377199693955,
+      "grad_norm": 2.3897983562604272,
+      "learning_rate": 6.359665649190239e-06,
+      "loss": 0.4594,
+      "step": 8235
+    },
+    {
+      "epoch": 1.2602907421576128,
+      "grad_norm": 2.402575115221835,
+      "learning_rate": 6.357357616294912e-06,
+      "loss": 0.4075,
+      "step": 8236
+    },
+    {
+      "epoch": 1.2604437643458302,
+      "grad_norm": 2.263508452347503,
+      "learning_rate": 6.3550498071085025e-06,
+      "loss": 0.3954,
+      "step": 8237
+    },
+    {
+      "epoch": 1.2605967865340475,
+      "grad_norm": 2.2037969854889,
+      "learning_rate": 6.3527422217727405e-06,
+      "loss": 0.427,
+      "step": 8238
+    },
+    {
+      "epoch": 1.2607498087222648,
+      "grad_norm": 2.2239380934976687,
+      "learning_rate": 6.350434860429351e-06,
+      "loss": 0.4106,
+      "step": 8239
+    },
+    {
+      "epoch": 1.260902830910482,
+      "grad_norm": 2.468996984247626,
+      "learning_rate": 6.34812772322003e-06,
+      "loss": 0.4408,
+      "step": 8240
+    },
+    {
+      "epoch": 1.2610558530986993,
+      "grad_norm": 2.0541356098332915,
+      "learning_rate": 6.3458208102864735e-06,
+      "loss": 0.3487,
+      "step": 8241
+    },
+    {
+      "epoch": 1.2612088752869166,
+      "grad_norm": 2.30173111222071,
+      "learning_rate": 6.343514121770358e-06,
+      "loss": 0.4123,
+      "step": 8242
+    },
+    {
+      "epoch": 1.261361897475134,
+      "grad_norm": 2.4270570534290963,
+      "learning_rate": 6.341207657813342e-06,
+      "loss": 0.4424,
+      "step": 8243
+    },
+    {
+      "epoch": 1.2615149196633513,
+      "grad_norm": 2.399051121363175,
+      "learning_rate": 6.338901418557078e-06,
+      "loss": 0.4351,
+      "step": 8244
+    },
+    {
+      "epoch": 1.2616679418515684,
+      "grad_norm": 2.0292806752648724,
+      "learning_rate": 6.336595404143202e-06,
+      "loss": 0.3252,
+      "step": 8245
+    },
+    {
+      "epoch": 1.2618209640397857,
+      "grad_norm": 2.3769670869429755,
+      "learning_rate": 6.334289614713329e-06,
+      "loss": 0.4465,
+      "step": 8246
+    },
+    {
+      "epoch": 1.261973986228003,
+      "grad_norm": 2.217733053161381,
+      "learning_rate": 6.331984050409075e-06,
+      "loss": 0.3758,
+      "step": 8247
+    },
+    {
+      "epoch": 1.2621270084162204,
+      "grad_norm": 2.142845566249004,
+      "learning_rate": 6.329678711372025e-06,
+      "loss": 0.3559,
+      "step": 8248
+    },
+    {
+      "epoch": 1.2622800306044377,
+      "grad_norm": 2.2208892699743004,
+      "learning_rate": 6.32737359774377e-06,
+      "loss": 0.4055,
+      "step": 8249
+    },
+    {
+      "epoch": 1.2624330527926548,
+      "grad_norm": 2.1323922539303015,
+      "learning_rate": 6.325068709665867e-06,
+      "loss": 0.3995,
+      "step": 8250
+    },
+    {
+      "epoch": 1.2625860749808722,
+      "grad_norm": 1.8538505473912326,
+      "learning_rate": 6.322764047279869e-06,
+      "loss": 0.3345,
+      "step": 8251
+    },
+    {
+      "epoch": 1.2627390971690895,
+      "grad_norm": 2.086198855742557,
+      "learning_rate": 6.320459610727321e-06,
+      "loss": 0.3561,
+      "step": 8252
+    },
+    {
+      "epoch": 1.2628921193573068,
+      "grad_norm": 2.293317461350589,
+      "learning_rate": 6.318155400149739e-06,
+      "loss": 0.4263,
+      "step": 8253
+    },
+    {
+      "epoch": 1.2630451415455242,
+      "grad_norm": 2.057507239735757,
+      "learning_rate": 6.315851415688637e-06,
+      "loss": 0.3783,
+      "step": 8254
+    },
+    {
+      "epoch": 1.2631981637337413,
+      "grad_norm": 2.0939084818628655,
+      "learning_rate": 6.313547657485516e-06,
+      "loss": 0.4935,
+      "step": 8255
+    },
+    {
+      "epoch": 1.2633511859219586,
+      "grad_norm": 2.0975012438371197,
+      "learning_rate": 6.31124412568185e-06,
+      "loss": 0.4144,
+      "step": 8256
+    },
+    {
+      "epoch": 1.263504208110176,
+      "grad_norm": 2.212720946192638,
+      "learning_rate": 6.3089408204191146e-06,
+      "loss": 0.4442,
+      "step": 8257
+    },
+    {
+      "epoch": 1.2636572302983933,
+      "grad_norm": 2.1433392764607047,
+      "learning_rate": 6.306637741838766e-06,
+      "loss": 0.4074,
+      "step": 8258
+    },
+    {
+      "epoch": 1.2638102524866106,
+      "grad_norm": 1.9780751620760415,
+      "learning_rate": 6.304334890082238e-06,
+      "loss": 0.347,
+      "step": 8259
+    },
+    {
+      "epoch": 1.2639632746748277,
+      "grad_norm": 1.8717079822783924,
+      "learning_rate": 6.302032265290963e-06,
+      "loss": 0.3133,
+      "step": 8260
+    },
+    {
+      "epoch": 1.2641162968630453,
+      "grad_norm": 1.991064644715254,
+      "learning_rate": 6.299729867606352e-06,
+      "loss": 0.4103,
+      "step": 8261
+    },
+    {
+      "epoch": 1.2642693190512624,
+      "grad_norm": 2.0474793196900722,
+      "learning_rate": 6.297427697169809e-06,
+      "loss": 0.3109,
+      "step": 8262
+    },
+    {
+      "epoch": 1.2644223412394797,
+      "grad_norm": 2.0837694344823885,
+      "learning_rate": 6.295125754122713e-06,
+      "loss": 0.3487,
+      "step": 8263
+    },
+    {
+      "epoch": 1.264575363427697,
+      "grad_norm": 2.0638674344607497,
+      "learning_rate": 6.292824038606437e-06,
+      "loss": 0.3289,
+      "step": 8264
+    },
+    {
+      "epoch": 1.2647283856159144,
+      "grad_norm": 2.322091446527354,
+      "learning_rate": 6.290522550762343e-06,
+      "loss": 0.4061,
+      "step": 8265
+    },
+    {
+      "epoch": 1.2648814078041317,
+      "grad_norm": 2.4768178687633124,
+      "learning_rate": 6.288221290731769e-06,
+      "loss": 0.4325,
+      "step": 8266
+    },
+    {
+      "epoch": 1.2650344299923488,
+      "grad_norm": 2.193177524623937,
+      "learning_rate": 6.285920258656043e-06,
+      "loss": 0.4204,
+      "step": 8267
+    },
+    {
+      "epoch": 1.2651874521805662,
+      "grad_norm": 2.428941265384786,
+      "learning_rate": 6.283619454676489e-06,
+      "loss": 0.4369,
+      "step": 8268
+    },
+    {
+      "epoch": 1.2653404743687835,
+      "grad_norm": 2.3254280713571265,
+      "learning_rate": 6.281318878934397e-06,
+      "loss": 0.3982,
+      "step": 8269
+    },
+    {
+      "epoch": 1.2654934965570008,
+      "grad_norm": 2.272008199503931,
+      "learning_rate": 6.279018531571064e-06,
+      "loss": 0.4191,
+      "step": 8270
+    },
+    {
+      "epoch": 1.2656465187452182,
+      "grad_norm": 2.305652085924085,
+      "learning_rate": 6.27671841272776e-06,
+      "loss": 0.3955,
+      "step": 8271
+    },
+    {
+      "epoch": 1.2657995409334353,
+      "grad_norm": 1.9353755654140936,
+      "learning_rate": 6.27441852254574e-06,
+      "loss": 0.3637,
+      "step": 8272
+    },
+    {
+      "epoch": 1.2659525631216526,
+      "grad_norm": 1.8828491506877458,
+      "learning_rate": 6.2721188611662575e-06,
+      "loss": 0.2892,
+      "step": 8273
+    },
+    {
+      "epoch": 1.26610558530987,
+      "grad_norm": 2.1776025532535623,
+      "learning_rate": 6.269819428730539e-06,
+      "loss": 0.4541,
+      "step": 8274
+    },
+    {
+      "epoch": 1.2662586074980873,
+      "grad_norm": 1.999490178432348,
+      "learning_rate": 6.267520225379797e-06,
+      "loss": 0.3681,
+      "step": 8275
+    },
+    {
+      "epoch": 1.2664116296863046,
+      "grad_norm": 2.253935070804977,
+      "learning_rate": 6.265221251255245e-06,
+      "loss": 0.3812,
+      "step": 8276
+    },
+    {
+      "epoch": 1.2665646518745217,
+      "grad_norm": 2.002987514155831,
+      "learning_rate": 6.262922506498063e-06,
+      "loss": 0.3807,
+      "step": 8277
+    },
+    {
+      "epoch": 1.266717674062739,
+      "grad_norm": 2.0505934288097167,
+      "learning_rate": 6.26062399124943e-06,
+      "loss": 0.3383,
+      "step": 8278
+    },
+    {
+      "epoch": 1.2668706962509564,
+      "grad_norm": 2.0558883228802864,
+      "learning_rate": 6.25832570565051e-06,
+      "loss": 0.3349,
+      "step": 8279
+    },
+    {
+      "epoch": 1.2670237184391737,
+      "grad_norm": 2.1637451051227834,
+      "learning_rate": 6.25602764984244e-06,
+      "loss": 0.3656,
+      "step": 8280
+    },
+    {
+      "epoch": 1.267176740627391,
+      "grad_norm": 2.101792748438887,
+      "learning_rate": 6.2537298239663615e-06,
+      "loss": 0.3983,
+      "step": 8281
+    },
+    {
+      "epoch": 1.2673297628156082,
+      "grad_norm": 1.9893459953979338,
+      "learning_rate": 6.251432228163392e-06,
+      "loss": 0.3411,
+      "step": 8282
+    },
+    {
+      "epoch": 1.2674827850038255,
+      "grad_norm": 2.2123349417488734,
+      "learning_rate": 6.249134862574631e-06,
+      "loss": 0.4035,
+      "step": 8283
+    },
+    {
+      "epoch": 1.2676358071920428,
+      "grad_norm": 1.9567029487218803,
+      "learning_rate": 6.2468377273411715e-06,
+      "loss": 0.3217,
+      "step": 8284
+    },
+    {
+      "epoch": 1.2677888293802602,
+      "grad_norm": 2.25711180116385,
+      "learning_rate": 6.244540822604088e-06,
+      "loss": 0.3533,
+      "step": 8285
+    },
+    {
+      "epoch": 1.2679418515684775,
+      "grad_norm": 2.1290431639737477,
+      "learning_rate": 6.24224414850445e-06,
+      "loss": 0.4025,
+      "step": 8286
+    },
+    {
+      "epoch": 1.2680948737566946,
+      "grad_norm": 2.4206966590733434,
+      "learning_rate": 6.239947705183296e-06,
+      "loss": 0.4232,
+      "step": 8287
+    },
+    {
+      "epoch": 1.268247895944912,
+      "grad_norm": 2.2201454182364677,
+      "learning_rate": 6.23765149278166e-06,
+      "loss": 0.3474,
+      "step": 8288
+    },
+    {
+      "epoch": 1.2684009181331293,
+      "grad_norm": 2.0819815759252935,
+      "learning_rate": 6.2353555114405695e-06,
+      "loss": 0.3625,
+      "step": 8289
+    },
+    {
+      "epoch": 1.2685539403213466,
+      "grad_norm": 2.6327659066634244,
+      "learning_rate": 6.233059761301022e-06,
+      "loss": 0.3831,
+      "step": 8290
+    },
+    {
+      "epoch": 1.268706962509564,
+      "grad_norm": 2.358386566791711,
+      "learning_rate": 6.2307642425040085e-06,
+      "loss": 0.4258,
+      "step": 8291
+    },
+    {
+      "epoch": 1.268859984697781,
+      "grad_norm": 2.406902397025713,
+      "learning_rate": 6.228468955190512e-06,
+      "loss": 0.4622,
+      "step": 8292
+    },
+    {
+      "epoch": 1.2690130068859986,
+      "grad_norm": 2.1879106838578375,
+      "learning_rate": 6.2261738995014885e-06,
+      "loss": 0.4255,
+      "step": 8293
+    },
+    {
+      "epoch": 1.2691660290742157,
+      "grad_norm": 2.0277613512524058,
+      "learning_rate": 6.22387907557789e-06,
+      "loss": 0.4056,
+      "step": 8294
+    },
+    {
+      "epoch": 1.269319051262433,
+      "grad_norm": 2.172511714493054,
+      "learning_rate": 6.221584483560652e-06,
+      "loss": 0.3701,
+      "step": 8295
+    },
+    {
+      "epoch": 1.2694720734506504,
+      "grad_norm": 1.9385020661538073,
+      "learning_rate": 6.219290123590689e-06,
+      "loss": 0.3819,
+      "step": 8296
+    },
+    {
+      "epoch": 1.2696250956388677,
+      "grad_norm": 2.0184288428167183,
+      "learning_rate": 6.21699599580891e-06,
+      "loss": 0.3895,
+      "step": 8297
+    },
+    {
+      "epoch": 1.269778117827085,
+      "grad_norm": 2.275096189344347,
+      "learning_rate": 6.214702100356205e-06,
+      "loss": 0.3442,
+      "step": 8298
+    },
+    {
+      "epoch": 1.2699311400153022,
+      "grad_norm": 2.6762797079326064,
+      "learning_rate": 6.212408437373456e-06,
+      "loss": 0.3961,
+      "step": 8299
+    },
+    {
+      "epoch": 1.2700841622035195,
+      "grad_norm": 2.2741365286187025,
+      "learning_rate": 6.21011500700152e-06,
+      "loss": 0.4131,
+      "step": 8300
+    },
+    {
+      "epoch": 1.2702371843917368,
+      "grad_norm": 1.9190830243481853,
+      "learning_rate": 6.207821809381247e-06,
+      "loss": 0.2969,
+      "step": 8301
+    },
+    {
+      "epoch": 1.2703902065799542,
+      "grad_norm": 2.2401792695423497,
+      "learning_rate": 6.205528844653474e-06,
+      "loss": 0.3656,
+      "step": 8302
+    },
+    {
+      "epoch": 1.2705432287681715,
+      "grad_norm": 1.9741605060912868,
+      "learning_rate": 6.2032361129590165e-06,
+      "loss": 0.3255,
+      "step": 8303
+    },
+    {
+      "epoch": 1.2706962509563886,
+      "grad_norm": 1.9776819900089568,
+      "learning_rate": 6.200943614438681e-06,
+      "loss": 0.3396,
+      "step": 8304
+    },
+    {
+      "epoch": 1.270849273144606,
+      "grad_norm": 2.4727351802146096,
+      "learning_rate": 6.1986513492332645e-06,
+      "loss": 0.4541,
+      "step": 8305
+    },
+    {
+      "epoch": 1.2710022953328233,
+      "grad_norm": 1.762131739082361,
+      "learning_rate": 6.196359317483536e-06,
+      "loss": 0.2889,
+      "step": 8306
+    },
+    {
+      "epoch": 1.2711553175210406,
+      "grad_norm": 2.077140126012181,
+      "learning_rate": 6.1940675193302625e-06,
+      "loss": 0.3449,
+      "step": 8307
+    },
+    {
+      "epoch": 1.271308339709258,
+      "grad_norm": 1.966671711976856,
+      "learning_rate": 6.191775954914195e-06,
+      "loss": 0.3855,
+      "step": 8308
+    },
+    {
+      "epoch": 1.271461361897475,
+      "grad_norm": 1.9752626408364398,
+      "learning_rate": 6.189484624376059e-06,
+      "loss": 0.321,
+      "step": 8309
+    },
+    {
+      "epoch": 1.2716143840856924,
+      "grad_norm": 2.1434276190951,
+      "learning_rate": 6.187193527856581e-06,
+      "loss": 0.4265,
+      "step": 8310
+    },
+    {
+      "epoch": 1.2717674062739097,
+      "grad_norm": 2.145563444426424,
+      "learning_rate": 6.184902665496466e-06,
+      "loss": 0.34,
+      "step": 8311
+    },
+    {
+      "epoch": 1.271920428462127,
+      "grad_norm": 1.9283577771018996,
+      "learning_rate": 6.182612037436399e-06,
+      "loss": 0.3911,
+      "step": 8312
+    },
+    {
+      "epoch": 1.2720734506503444,
+      "grad_norm": 2.3122756860692784,
+      "learning_rate": 6.180321643817062e-06,
+      "loss": 0.3537,
+      "step": 8313
+    },
+    {
+      "epoch": 1.2722264728385615,
+      "grad_norm": 2.164697408819195,
+      "learning_rate": 6.178031484779115e-06,
+      "loss": 0.353,
+      "step": 8314
+    },
+    {
+      "epoch": 1.2723794950267788,
+      "grad_norm": 2.0017443136895654,
+      "learning_rate": 6.1757415604632084e-06,
+      "loss": 0.3517,
+      "step": 8315
+    },
+    {
+      "epoch": 1.2725325172149962,
+      "grad_norm": 2.4725242851848463,
+      "learning_rate": 6.173451871009974e-06,
+      "loss": 0.3811,
+      "step": 8316
+    },
+    {
+      "epoch": 1.2726855394032135,
+      "grad_norm": 1.9614402340989665,
+      "learning_rate": 6.1711624165600255e-06,
+      "loss": 0.3094,
+      "step": 8317
+    },
+    {
+      "epoch": 1.2728385615914308,
+      "grad_norm": 2.4254988450390393,
+      "learning_rate": 6.168873197253975e-06,
+      "loss": 0.3652,
+      "step": 8318
+    },
+    {
+      "epoch": 1.272991583779648,
+      "grad_norm": 2.3981313138203686,
+      "learning_rate": 6.1665842132324095e-06,
+      "loss": 0.4186,
+      "step": 8319
+    },
+    {
+      "epoch": 1.2731446059678653,
+      "grad_norm": 2.1409247382783057,
+      "learning_rate": 6.164295464635899e-06,
+      "loss": 0.3752,
+      "step": 8320
+    },
+    {
+      "epoch": 1.2732976281560826,
+      "grad_norm": 2.1396974459941616,
+      "learning_rate": 6.162006951605015e-06,
+      "loss": 0.3566,
+      "step": 8321
+    },
+    {
+      "epoch": 1.2734506503443,
+      "grad_norm": 1.9810680981670772,
+      "learning_rate": 6.159718674280294e-06,
+      "loss": 0.3446,
+      "step": 8322
+    },
+    {
+      "epoch": 1.2736036725325173,
+      "grad_norm": 2.049489820508499,
+      "learning_rate": 6.157430632802275e-06,
+      "loss": 0.3551,
+      "step": 8323
+    },
+    {
+      "epoch": 1.2737566947207344,
+      "grad_norm": 2.1250524285136674,
+      "learning_rate": 6.1551428273114745e-06,
+      "loss": 0.3756,
+      "step": 8324
+    },
+    {
+      "epoch": 1.2739097169089517,
+      "grad_norm": 2.0211598009609637,
+      "learning_rate": 6.15285525794839e-06,
+      "loss": 0.3541,
+      "step": 8325
+    },
+    {
+      "epoch": 1.274062739097169,
+      "grad_norm": 2.1857253921429525,
+      "learning_rate": 6.150567924853516e-06,
+      "loss": 0.4467,
+      "step": 8326
+    },
+    {
+      "epoch": 1.2742157612853864,
+      "grad_norm": 2.4665073319611563,
+      "learning_rate": 6.1482808281673275e-06,
+      "loss": 0.4236,
+      "step": 8327
+    },
+    {
+      "epoch": 1.2743687834736037,
+      "grad_norm": 2.0913011820808998,
+      "learning_rate": 6.1459939680302774e-06,
+      "loss": 0.3339,
+      "step": 8328
+    },
+    {
+      "epoch": 1.2745218056618208,
+      "grad_norm": 1.952019493038919,
+      "learning_rate": 6.143707344582817e-06,
+      "loss": 0.3398,
+      "step": 8329
+    },
+    {
+      "epoch": 1.2746748278500384,
+      "grad_norm": 2.2720660090661906,
+      "learning_rate": 6.141420957965372e-06,
+      "loss": 0.4193,
+      "step": 8330
+    },
+    {
+      "epoch": 1.2748278500382555,
+      "grad_norm": 1.9846585102425593,
+      "learning_rate": 6.1391348083183635e-06,
+      "loss": 0.2835,
+      "step": 8331
+    },
+    {
+      "epoch": 1.2749808722264728,
+      "grad_norm": 2.2036433802674287,
+      "learning_rate": 6.13684889578219e-06,
+      "loss": 0.3277,
+      "step": 8332
+    },
+    {
+      "epoch": 1.2751338944146902,
+      "grad_norm": 1.9806487568109303,
+      "learning_rate": 6.134563220497236e-06,
+      "loss": 0.2871,
+      "step": 8333
+    },
+    {
+      "epoch": 1.2752869166029075,
+      "grad_norm": 1.9513522399246421,
+      "learning_rate": 6.132277782603879e-06,
+      "loss": 0.3365,
+      "step": 8334
+    },
+    {
+      "epoch": 1.2754399387911248,
+      "grad_norm": 2.165274496595406,
+      "learning_rate": 6.129992582242472e-06,
+      "loss": 0.3678,
+      "step": 8335
+    },
+    {
+      "epoch": 1.275592960979342,
+      "grad_norm": 1.9969987826577444,
+      "learning_rate": 6.127707619553361e-06,
+      "loss": 0.4157,
+      "step": 8336
+    },
+    {
+      "epoch": 1.2757459831675593,
+      "grad_norm": 2.265249391828096,
+      "learning_rate": 6.125422894676876e-06,
+      "loss": 0.394,
+      "step": 8337
+    },
+    {
+      "epoch": 1.2758990053557766,
+      "grad_norm": 2.116122460738184,
+      "learning_rate": 6.123138407753324e-06,
+      "loss": 0.3288,
+      "step": 8338
+    },
+    {
+      "epoch": 1.276052027543994,
+      "grad_norm": 1.8821468642653791,
+      "learning_rate": 6.120854158923011e-06,
+      "loss": 0.3307,
+      "step": 8339
+    },
+    {
+      "epoch": 1.2762050497322113,
+      "grad_norm": 1.9513607257134291,
+      "learning_rate": 6.118570148326222e-06,
+      "loss": 0.2738,
+      "step": 8340
+    },
+    {
+      "epoch": 1.2763580719204284,
+      "grad_norm": 1.9936208113589038,
+      "learning_rate": 6.11628637610322e-06,
+      "loss": 0.3224,
+      "step": 8341
+    },
+    {
+      "epoch": 1.2765110941086457,
+      "grad_norm": 2.3496749593098785,
+      "learning_rate": 6.114002842394267e-06,
+      "loss": 0.5041,
+      "step": 8342
+    },
+    {
+      "epoch": 1.276664116296863,
+      "grad_norm": 2.0713929832468616,
+      "learning_rate": 6.1117195473395985e-06,
+      "loss": 0.3629,
+      "step": 8343
+    },
+    {
+      "epoch": 1.2768171384850804,
+      "grad_norm": 2.2931273716137355,
+      "learning_rate": 6.109436491079449e-06,
+      "loss": 0.401,
+      "step": 8344
+    },
+    {
+      "epoch": 1.2769701606732977,
+      "grad_norm": 1.9863701102849063,
+      "learning_rate": 6.1071536737540225e-06,
+      "loss": 0.3444,
+      "step": 8345
+    },
+    {
+      "epoch": 1.2771231828615148,
+      "grad_norm": 1.7369545129822166,
+      "learning_rate": 6.104871095503515e-06,
+      "loss": 0.3197,
+      "step": 8346
+    },
+    {
+      "epoch": 1.2772762050497322,
+      "grad_norm": 2.077382461800277,
+      "learning_rate": 6.102588756468115e-06,
+      "loss": 0.382,
+      "step": 8347
+    },
+    {
+      "epoch": 1.2774292272379495,
+      "grad_norm": 2.494916291650205,
+      "learning_rate": 6.100306656787986e-06,
+      "loss": 0.3721,
+      "step": 8348
+    },
+    {
+      "epoch": 1.2775822494261668,
+      "grad_norm": 1.9543398821885236,
+      "learning_rate": 6.098024796603277e-06,
+      "loss": 0.374,
+      "step": 8349
+    },
+    {
+      "epoch": 1.2777352716143842,
+      "grad_norm": 2.0822556782993784,
+      "learning_rate": 6.095743176054134e-06,
+      "loss": 0.3969,
+      "step": 8350
+    },
+    {
+      "epoch": 1.2778882938026013,
+      "grad_norm": 2.0266313057284635,
+      "learning_rate": 6.093461795280673e-06,
+      "loss": 0.3629,
+      "step": 8351
+    },
+    {
+      "epoch": 1.2780413159908186,
+      "grad_norm": 2.2597584533701247,
+      "learning_rate": 6.091180654423008e-06,
+      "loss": 0.3956,
+      "step": 8352
+    },
+    {
+      "epoch": 1.278194338179036,
+      "grad_norm": 2.156520273496613,
+      "learning_rate": 6.0888997536212314e-06,
+      "loss": 0.3296,
+      "step": 8353
+    },
+    {
+      "epoch": 1.2783473603672533,
+      "grad_norm": 2.0934949462261416,
+      "learning_rate": 6.086619093015416e-06,
+      "loss": 0.3407,
+      "step": 8354
+    },
+    {
+      "epoch": 1.2785003825554706,
+      "grad_norm": 2.169281922759286,
+      "learning_rate": 6.0843386727456355e-06,
+      "loss": 0.3667,
+      "step": 8355
+    },
+    {
+      "epoch": 1.2786534047436877,
+      "grad_norm": 2.22918601294946,
+      "learning_rate": 6.0820584929519375e-06,
+      "loss": 0.3654,
+      "step": 8356
+    },
+    {
+      "epoch": 1.278806426931905,
+      "grad_norm": 2.6905374498040087,
+      "learning_rate": 6.079778553774347e-06,
+      "loss": 0.4124,
+      "step": 8357
+    },
+    {
+      "epoch": 1.2789594491201224,
+      "grad_norm": 2.2477897591567015,
+      "learning_rate": 6.077498855352899e-06,
+      "loss": 0.4542,
+      "step": 8358
+    },
+    {
+      "epoch": 1.2791124713083397,
+      "grad_norm": 2.2985892832882864,
+      "learning_rate": 6.075219397827585e-06,
+      "loss": 0.3577,
+      "step": 8359
+    },
+    {
+      "epoch": 1.279265493496557,
+      "grad_norm": 2.1801673874314336,
+      "learning_rate": 6.072940181338406e-06,
+      "loss": 0.4154,
+      "step": 8360
+    },
+    {
+      "epoch": 1.2794185156847742,
+      "grad_norm": 2.069130573415728,
+      "learning_rate": 6.070661206025334e-06,
+      "loss": 0.3956,
+      "step": 8361
+    },
+    {
+      "epoch": 1.2795715378729917,
+      "grad_norm": 2.0048130565920155,
+      "learning_rate": 6.068382472028325e-06,
+      "loss": 0.3135,
+      "step": 8362
+    },
+    {
+      "epoch": 1.2797245600612088,
+      "grad_norm": 1.88601860046552,
+      "learning_rate": 6.066103979487332e-06,
+      "loss": 0.3595,
+      "step": 8363
+    },
+    {
+      "epoch": 1.2798775822494262,
+      "grad_norm": 2.308966911833846,
+      "learning_rate": 6.0638257285422845e-06,
+      "loss": 0.3559,
+      "step": 8364
+    },
+    {
+      "epoch": 1.2800306044376435,
+      "grad_norm": 1.9954425438063566,
+      "learning_rate": 6.061547719333092e-06,
+      "loss": 0.4064,
+      "step": 8365
+    },
+    {
+      "epoch": 1.2801836266258608,
+      "grad_norm": 2.2874589176202775,
+      "learning_rate": 6.059269951999665e-06,
+      "loss": 0.3284,
+      "step": 8366
+    },
+    {
+      "epoch": 1.2803366488140782,
+      "grad_norm": 2.281622447058958,
+      "learning_rate": 6.0569924266818844e-06,
+      "loss": 0.4108,
+      "step": 8367
+    },
+    {
+      "epoch": 1.2804896710022953,
+      "grad_norm": 2.206075216573966,
+      "learning_rate": 6.054715143519627e-06,
+      "loss": 0.366,
+      "step": 8368
+    },
+    {
+      "epoch": 1.2806426931905126,
+      "grad_norm": 2.2776572125639842,
+      "learning_rate": 6.052438102652746e-06,
+      "loss": 0.4196,
+      "step": 8369
+    },
+    {
+      "epoch": 1.28079571537873,
+      "grad_norm": 1.7869163264626549,
+      "learning_rate": 6.05016130422108e-06,
+      "loss": 0.3237,
+      "step": 8370
+    },
+    {
+      "epoch": 1.2809487375669473,
+      "grad_norm": 2.1341648244510423,
+      "learning_rate": 6.047884748364467e-06,
+      "loss": 0.3737,
+      "step": 8371
+    },
+    {
+      "epoch": 1.2811017597551646,
+      "grad_norm": 2.18049711268148,
+      "learning_rate": 6.0456084352227065e-06,
+      "loss": 0.409,
+      "step": 8372
+    },
+    {
+      "epoch": 1.2812547819433817,
+      "grad_norm": 2.1175669065339093,
+      "learning_rate": 6.043332364935603e-06,
+      "loss": 0.3558,
+      "step": 8373
+    },
+    {
+      "epoch": 1.281407804131599,
+      "grad_norm": 2.173177324049422,
+      "learning_rate": 6.041056537642942e-06,
+      "loss": 0.336,
+      "step": 8374
+    },
+    {
+      "epoch": 1.2815608263198164,
+      "grad_norm": 2.0800240094069697,
+      "learning_rate": 6.0387809534844795e-06,
+      "loss": 0.3442,
+      "step": 8375
+    },
+    {
+      "epoch": 1.2817138485080337,
+      "grad_norm": 1.906890683579724,
+      "learning_rate": 6.03650561259998e-06,
+      "loss": 0.3571,
+      "step": 8376
+    },
+    {
+      "epoch": 1.281866870696251,
+      "grad_norm": 1.989073502947194,
+      "learning_rate": 6.0342305151291755e-06,
+      "loss": 0.3221,
+      "step": 8377
+    },
+    {
+      "epoch": 1.2820198928844682,
+      "grad_norm": 1.816191193612736,
+      "learning_rate": 6.031955661211788e-06,
+      "loss": 0.3345,
+      "step": 8378
+    },
+    {
+      "epoch": 1.2821729150726855,
+      "grad_norm": 2.0046077691510362,
+      "learning_rate": 6.029681050987526e-06,
+      "loss": 0.3253,
+      "step": 8379
+    },
+    {
+      "epoch": 1.2823259372609028,
+      "grad_norm": 1.998886441065847,
+      "learning_rate": 6.027406684596082e-06,
+      "loss": 0.358,
+      "step": 8380
+    },
+    {
+      "epoch": 1.2824789594491202,
+      "grad_norm": 2.301187711651734,
+      "learning_rate": 6.025132562177139e-06,
+      "loss": 0.3304,
+      "step": 8381
+    },
+    {
+      "epoch": 1.2826319816373375,
+      "grad_norm": 2.1476577961675627,
+      "learning_rate": 6.022858683870351e-06,
+      "loss": 0.3894,
+      "step": 8382
+    },
+    {
+      "epoch": 1.2827850038255546,
+      "grad_norm": 2.245599970159706,
+      "learning_rate": 6.020585049815368e-06,
+      "loss": 0.402,
+      "step": 8383
+    },
+    {
+      "epoch": 1.282938026013772,
+      "grad_norm": 1.8652376913870043,
+      "learning_rate": 6.01831166015183e-06,
+      "loss": 0.3539,
+      "step": 8384
+    },
+    {
+      "epoch": 1.2830910482019893,
+      "grad_norm": 2.091989916493354,
+      "learning_rate": 6.016038515019347e-06,
+      "loss": 0.3117,
+      "step": 8385
+    },
+    {
+      "epoch": 1.2832440703902066,
+      "grad_norm": 2.2039189929629255,
+      "learning_rate": 6.013765614557522e-06,
+      "loss": 0.4243,
+      "step": 8386
+    },
+    {
+      "epoch": 1.283397092578424,
+      "grad_norm": 2.2239495888325886,
+      "learning_rate": 6.011492958905949e-06,
+      "loss": 0.3114,
+      "step": 8387
+    },
+    {
+      "epoch": 1.283550114766641,
+      "grad_norm": 2.1535282910124747,
+      "learning_rate": 6.009220548204193e-06,
+      "loss": 0.383,
+      "step": 8388
+    },
+    {
+      "epoch": 1.2837031369548584,
+      "grad_norm": 1.8290585796153886,
+      "learning_rate": 6.006948382591816e-06,
+      "loss": 0.2985,
+      "step": 8389
+    },
+    {
+      "epoch": 1.2838561591430757,
+      "grad_norm": 2.023290060036757,
+      "learning_rate": 6.004676462208363e-06,
+      "loss": 0.3981,
+      "step": 8390
+    },
+    {
+      "epoch": 1.284009181331293,
+      "grad_norm": 1.914516702960148,
+      "learning_rate": 6.002404787193352e-06,
+      "loss": 0.2963,
+      "step": 8391
+    },
+    {
+      "epoch": 1.2841622035195104,
+      "grad_norm": 2.1996232859634914,
+      "learning_rate": 6.000133357686306e-06,
+      "loss": 0.3343,
+      "step": 8392
+    },
+    {
+      "epoch": 1.2843152257077275,
+      "grad_norm": 2.190271452100412,
+      "learning_rate": 5.99786217382672e-06,
+      "loss": 0.3885,
+      "step": 8393
+    },
+    {
+      "epoch": 1.284468247895945,
+      "grad_norm": 2.3640070841460195,
+      "learning_rate": 5.995591235754069e-06,
+      "loss": 0.4181,
+      "step": 8394
+    },
+    {
+      "epoch": 1.2846212700841622,
+      "grad_norm": 2.0388498245605,
+      "learning_rate": 5.993320543607828e-06,
+      "loss": 0.3608,
+      "step": 8395
+    },
+    {
+      "epoch": 1.2847742922723795,
+      "grad_norm": 2.2357999558982247,
+      "learning_rate": 5.9910500975274456e-06,
+      "loss": 0.3957,
+      "step": 8396
+    },
+    {
+      "epoch": 1.2849273144605968,
+      "grad_norm": 2.041737865466858,
+      "learning_rate": 5.988779897652364e-06,
+      "loss": 0.4002,
+      "step": 8397
+    },
+    {
+      "epoch": 1.2850803366488142,
+      "grad_norm": 2.239729576172599,
+      "learning_rate": 5.986509944121996e-06,
+      "loss": 0.3372,
+      "step": 8398
+    },
+    {
+      "epoch": 1.2852333588370315,
+      "grad_norm": 2.2708251624696763,
+      "learning_rate": 5.984240237075754e-06,
+      "loss": 0.3929,
+      "step": 8399
+    },
+    {
+      "epoch": 1.2853863810252486,
+      "grad_norm": 1.9576468807793188,
+      "learning_rate": 5.981970776653032e-06,
+      "loss": 0.3,
+      "step": 8400
+    },
+    {
+      "epoch": 1.285539403213466,
+      "grad_norm": 2.348142708232122,
+      "learning_rate": 5.979701562993199e-06,
+      "loss": 0.4128,
+      "step": 8401
+    },
+    {
+      "epoch": 1.2856924254016833,
+      "grad_norm": 2.225493045946411,
+      "learning_rate": 5.977432596235623e-06,
+      "loss": 0.4226,
+      "step": 8402
+    },
+    {
+      "epoch": 1.2858454475899006,
+      "grad_norm": 2.0902073882347283,
+      "learning_rate": 5.9751638765196494e-06,
+      "loss": 0.3488,
+      "step": 8403
+    },
+    {
+      "epoch": 1.285998469778118,
+      "grad_norm": 2.0223018514828786,
+      "learning_rate": 5.972895403984603e-06,
+      "loss": 0.3408,
+      "step": 8404
+    },
+    {
+      "epoch": 1.286151491966335,
+      "grad_norm": 2.171673961595489,
+      "learning_rate": 5.970627178769806e-06,
+      "loss": 0.4067,
+      "step": 8405
+    },
+    {
+      "epoch": 1.2863045141545524,
+      "grad_norm": 2.216297180724609,
+      "learning_rate": 5.96835920101456e-06,
+      "loss": 0.3865,
+      "step": 8406
+    },
+    {
+      "epoch": 1.2864575363427697,
+      "grad_norm": 2.224480053286395,
+      "learning_rate": 5.966091470858142e-06,
+      "loss": 0.3895,
+      "step": 8407
+    },
+    {
+      "epoch": 1.286610558530987,
+      "grad_norm": 2.0159160285134994,
+      "learning_rate": 5.96382398843983e-06,
+      "loss": 0.3411,
+      "step": 8408
+    },
+    {
+      "epoch": 1.2867635807192044,
+      "grad_norm": 1.9803037406134543,
+      "learning_rate": 5.961556753898874e-06,
+      "loss": 0.3131,
+      "step": 8409
+    },
+    {
+      "epoch": 1.2869166029074215,
+      "grad_norm": 1.9246640134321054,
+      "learning_rate": 5.959289767374521e-06,
+      "loss": 0.2991,
+      "step": 8410
+    },
+    {
+      "epoch": 1.2870696250956388,
+      "grad_norm": 2.29223275943616,
+      "learning_rate": 5.957023029005988e-06,
+      "loss": 0.3997,
+      "step": 8411
+    },
+    {
+      "epoch": 1.2872226472838562,
+      "grad_norm": 2.0759714800081297,
+      "learning_rate": 5.9547565389324865e-06,
+      "loss": 0.3184,
+      "step": 8412
+    },
+    {
+      "epoch": 1.2873756694720735,
+      "grad_norm": 2.345956476444365,
+      "learning_rate": 5.952490297293214e-06,
+      "loss": 0.406,
+      "step": 8413
+    },
+    {
+      "epoch": 1.2875286916602908,
+      "grad_norm": 2.1234583404676997,
+      "learning_rate": 5.950224304227345e-06,
+      "loss": 0.3872,
+      "step": 8414
+    },
+    {
+      "epoch": 1.287681713848508,
+      "grad_norm": 2.3341118330349926,
+      "learning_rate": 5.947958559874041e-06,
+      "loss": 0.3964,
+      "step": 8415
+    },
+    {
+      "epoch": 1.2878347360367253,
+      "grad_norm": 2.346952170027564,
+      "learning_rate": 5.945693064372457e-06,
+      "loss": 0.3331,
+      "step": 8416
+    },
+    {
+      "epoch": 1.2879877582249426,
+      "grad_norm": 2.0502283032629234,
+      "learning_rate": 5.94342781786172e-06,
+      "loss": 0.3677,
+      "step": 8417
+    },
+    {
+      "epoch": 1.28814078041316,
+      "grad_norm": 2.3096833762404616,
+      "learning_rate": 5.941162820480952e-06,
+      "loss": 0.3422,
+      "step": 8418
+    },
+    {
+      "epoch": 1.2882938026013773,
+      "grad_norm": 2.138229007243028,
+      "learning_rate": 5.9388980723692545e-06,
+      "loss": 0.3992,
+      "step": 8419
+    },
+    {
+      "epoch": 1.2884468247895944,
+      "grad_norm": 2.262793774066718,
+      "learning_rate": 5.936633573665711e-06,
+      "loss": 0.3245,
+      "step": 8420
+    },
+    {
+      "epoch": 1.2885998469778117,
+      "grad_norm": 1.9477409894816895,
+      "learning_rate": 5.934369324509397e-06,
+      "loss": 0.2831,
+      "step": 8421
+    },
+    {
+      "epoch": 1.288752869166029,
+      "grad_norm": 2.2281573276173394,
+      "learning_rate": 5.93210532503937e-06,
+      "loss": 0.3967,
+      "step": 8422
+    },
+    {
+      "epoch": 1.2889058913542464,
+      "grad_norm": 2.22880743175084,
+      "learning_rate": 5.929841575394663e-06,
+      "loss": 0.3822,
+      "step": 8423
+    },
+    {
+      "epoch": 1.2890589135424637,
+      "grad_norm": 1.965550070449436,
+      "learning_rate": 5.9275780757143105e-06,
+      "loss": 0.3124,
+      "step": 8424
+    },
+    {
+      "epoch": 1.2892119357306808,
+      "grad_norm": 2.027513364688824,
+      "learning_rate": 5.9253148261373185e-06,
+      "loss": 0.3242,
+      "step": 8425
+    },
+    {
+      "epoch": 1.2893649579188982,
+      "grad_norm": 2.2444218795153486,
+      "learning_rate": 5.923051826802686e-06,
+      "loss": 0.3393,
+      "step": 8426
+    },
+    {
+      "epoch": 1.2895179801071155,
+      "grad_norm": 2.4148438662955822,
+      "learning_rate": 5.92078907784939e-06,
+      "loss": 0.3669,
+      "step": 8427
+    },
+    {
+      "epoch": 1.2896710022953328,
+      "grad_norm": 2.2021012890604483,
+      "learning_rate": 5.9185265794163905e-06,
+      "loss": 0.3546,
+      "step": 8428
+    },
+    {
+      "epoch": 1.2898240244835502,
+      "grad_norm": 2.1678300062988622,
+      "learning_rate": 5.916264331642645e-06,
+      "loss": 0.3763,
+      "step": 8429
+    },
+    {
+      "epoch": 1.2899770466717673,
+      "grad_norm": 2.070445823425188,
+      "learning_rate": 5.9140023346670815e-06,
+      "loss": 0.3678,
+      "step": 8430
+    },
+    {
+      "epoch": 1.2901300688599848,
+      "grad_norm": 2.2804935664834383,
+      "learning_rate": 5.911740588628616e-06,
+      "loss": 0.379,
+      "step": 8431
+    },
+    {
+      "epoch": 1.290283091048202,
+      "grad_norm": 2.0648217974475305,
+      "learning_rate": 5.9094790936661594e-06,
+      "loss": 0.3711,
+      "step": 8432
+    },
+    {
+      "epoch": 1.2904361132364193,
+      "grad_norm": 2.133264133751307,
+      "learning_rate": 5.907217849918588e-06,
+      "loss": 0.4078,
+      "step": 8433
+    },
+    {
+      "epoch": 1.2905891354246366,
+      "grad_norm": 2.0061436737092135,
+      "learning_rate": 5.904956857524784e-06,
+      "loss": 0.4,
+      "step": 8434
+    },
+    {
+      "epoch": 1.290742157612854,
+      "grad_norm": 1.9897296088214493,
+      "learning_rate": 5.902696116623599e-06,
+      "loss": 0.3381,
+      "step": 8435
+    },
+    {
+      "epoch": 1.2908951798010713,
+      "grad_norm": 1.9672303748095752,
+      "learning_rate": 5.900435627353868e-06,
+      "loss": 0.3526,
+      "step": 8436
+    },
+    {
+      "epoch": 1.2910482019892884,
+      "grad_norm": 2.3082399010119192,
+      "learning_rate": 5.898175389854431e-06,
+      "loss": 0.4296,
+      "step": 8437
+    },
+    {
+      "epoch": 1.2912012241775057,
+      "grad_norm": 2.043754552160689,
+      "learning_rate": 5.8959154042640855e-06,
+      "loss": 0.3501,
+      "step": 8438
+    },
+    {
+      "epoch": 1.291354246365723,
+      "grad_norm": 2.0111371417155053,
+      "learning_rate": 5.893655670721632e-06,
+      "loss": 0.3295,
+      "step": 8439
+    },
+    {
+      "epoch": 1.2915072685539404,
+      "grad_norm": 2.1845344064148677,
+      "learning_rate": 5.891396189365849e-06,
+      "loss": 0.3611,
+      "step": 8440
+    },
+    {
+      "epoch": 1.2916602907421577,
+      "grad_norm": 2.3597698023948865,
+      "learning_rate": 5.889136960335496e-06,
+      "loss": 0.3903,
+      "step": 8441
+    },
+    {
+      "epoch": 1.2918133129303748,
+      "grad_norm": 2.2960601544175567,
+      "learning_rate": 5.886877983769324e-06,
+      "loss": 0.3677,
+      "step": 8442
+    },
+    {
+      "epoch": 1.2919663351185922,
+      "grad_norm": 2.4107761408220028,
+      "learning_rate": 5.884619259806069e-06,
+      "loss": 0.357,
+      "step": 8443
+    },
+    {
+      "epoch": 1.2921193573068095,
+      "grad_norm": 2.211778674971672,
+      "learning_rate": 5.882360788584441e-06,
+      "loss": 0.3633,
+      "step": 8444
+    },
+    {
+      "epoch": 1.2922723794950268,
+      "grad_norm": 2.210377762764795,
+      "learning_rate": 5.880102570243147e-06,
+      "loss": 0.4351,
+      "step": 8445
+    },
+    {
+      "epoch": 1.2924254016832442,
+      "grad_norm": 2.0353077772274246,
+      "learning_rate": 5.877844604920869e-06,
+      "loss": 0.3138,
+      "step": 8446
+    },
+    {
+      "epoch": 1.2925784238714613,
+      "grad_norm": 2.1796447244885417,
+      "learning_rate": 5.875586892756284e-06,
+      "loss": 0.4046,
+      "step": 8447
+    },
+    {
+      "epoch": 1.2927314460596786,
+      "grad_norm": 2.1991555410920003,
+      "learning_rate": 5.873329433888042e-06,
+      "loss": 0.3785,
+      "step": 8448
+    },
+    {
+      "epoch": 1.292884468247896,
+      "grad_norm": 2.214367249698276,
+      "learning_rate": 5.87107222845478e-06,
+      "loss": 0.3908,
+      "step": 8449
+    },
+    {
+      "epoch": 1.2930374904361133,
+      "grad_norm": 2.261557704663619,
+      "learning_rate": 5.8688152765951274e-06,
+      "loss": 0.4244,
+      "step": 8450
+    },
+    {
+      "epoch": 1.2931905126243306,
+      "grad_norm": 2.099925902912094,
+      "learning_rate": 5.86655857844769e-06,
+      "loss": 0.359,
+      "step": 8451
+    },
+    {
+      "epoch": 1.2933435348125477,
+      "grad_norm": 2.5580065217380406,
+      "learning_rate": 5.8643021341510576e-06,
+      "loss": 0.3966,
+      "step": 8452
+    },
+    {
+      "epoch": 1.293496557000765,
+      "grad_norm": 2.334103559025696,
+      "learning_rate": 5.8620459438438145e-06,
+      "loss": 0.3454,
+      "step": 8453
+    },
+    {
+      "epoch": 1.2936495791889824,
+      "grad_norm": 2.208197675578791,
+      "learning_rate": 5.8597900076645135e-06,
+      "loss": 0.4068,
+      "step": 8454
+    },
+    {
+      "epoch": 1.2938026013771997,
+      "grad_norm": 2.158637856805305,
+      "learning_rate": 5.857534325751706e-06,
+      "loss": 0.4008,
+      "step": 8455
+    },
+    {
+      "epoch": 1.293955623565417,
+      "grad_norm": 2.076159021580232,
+      "learning_rate": 5.8552788982439234e-06,
+      "loss": 0.3684,
+      "step": 8456
+    },
+    {
+      "epoch": 1.2941086457536342,
+      "grad_norm": 2.2118198370758497,
+      "learning_rate": 5.853023725279675e-06,
+      "loss": 0.4319,
+      "step": 8457
+    },
+    {
+      "epoch": 1.2942616679418515,
+      "grad_norm": 2.0053899150292107,
+      "learning_rate": 5.850768806997465e-06,
+      "loss": 0.3374,
+      "step": 8458
+    },
+    {
+      "epoch": 1.2944146901300688,
+      "grad_norm": 2.1800900300923005,
+      "learning_rate": 5.848514143535774e-06,
+      "loss": 0.4282,
+      "step": 8459
+    },
+    {
+      "epoch": 1.2945677123182862,
+      "grad_norm": 2.2040001586963505,
+      "learning_rate": 5.846259735033068e-06,
+      "loss": 0.3899,
+      "step": 8460
+    },
+    {
+      "epoch": 1.2947207345065035,
+      "grad_norm": 2.14628589047775,
+      "learning_rate": 5.844005581627803e-06,
+      "loss": 0.3827,
+      "step": 8461
+    },
+    {
+      "epoch": 1.2948737566947206,
+      "grad_norm": 1.955815144083365,
+      "learning_rate": 5.84175168345841e-06,
+      "loss": 0.2907,
+      "step": 8462
+    },
+    {
+      "epoch": 1.2950267788829382,
+      "grad_norm": 2.2990986634509496,
+      "learning_rate": 5.8394980406633185e-06,
+      "loss": 0.3976,
+      "step": 8463
+    },
+    {
+      "epoch": 1.2951798010711553,
+      "grad_norm": 2.3551470293805488,
+      "learning_rate": 5.837244653380929e-06,
+      "loss": 0.4031,
+      "step": 8464
+    },
+    {
+      "epoch": 1.2953328232593726,
+      "grad_norm": 2.0455913786984032,
+      "learning_rate": 5.834991521749623e-06,
+      "loss": 0.3854,
+      "step": 8465
+    },
+    {
+      "epoch": 1.29548584544759,
+      "grad_norm": 2.295691500876195,
+      "learning_rate": 5.832738645907786e-06,
+      "loss": 0.4205,
+      "step": 8466
+    },
+    {
+      "epoch": 1.2956388676358073,
+      "grad_norm": 2.0712979508795573,
+      "learning_rate": 5.830486025993775e-06,
+      "loss": 0.3819,
+      "step": 8467
+    },
+    {
+      "epoch": 1.2957918898240246,
+      "grad_norm": 2.1216828024914287,
+      "learning_rate": 5.82823366214592e-06,
+      "loss": 0.3732,
+      "step": 8468
+    },
+    {
+      "epoch": 1.2959449120122417,
+      "grad_norm": 1.850364780767729,
+      "learning_rate": 5.82598155450256e-06,
+      "loss": 0.3449,
+      "step": 8469
+    },
+    {
+      "epoch": 1.296097934200459,
+      "grad_norm": 2.0162996634036245,
+      "learning_rate": 5.823729703202001e-06,
+      "loss": 0.3521,
+      "step": 8470
+    },
+    {
+      "epoch": 1.2962509563886764,
+      "grad_norm": 1.8674820045225895,
+      "learning_rate": 5.82147810838254e-06,
+      "loss": 0.3005,
+      "step": 8471
+    },
+    {
+      "epoch": 1.2964039785768937,
+      "grad_norm": 2.190393768245424,
+      "learning_rate": 5.819226770182453e-06,
+      "loss": 0.4191,
+      "step": 8472
+    },
+    {
+      "epoch": 1.296557000765111,
+      "grad_norm": 2.494076811857044,
+      "learning_rate": 5.816975688740007e-06,
+      "loss": 0.4447,
+      "step": 8473
+    },
+    {
+      "epoch": 1.2967100229533282,
+      "grad_norm": 2.113249050025513,
+      "learning_rate": 5.814724864193449e-06,
+      "loss": 0.3346,
+      "step": 8474
+    },
+    {
+      "epoch": 1.2968630451415455,
+      "grad_norm": 2.3310835110243144,
+      "learning_rate": 5.8124742966810075e-06,
+      "loss": 0.5078,
+      "step": 8475
+    },
+    {
+      "epoch": 1.2970160673297628,
+      "grad_norm": 2.083303137693278,
+      "learning_rate": 5.810223986340901e-06,
+      "loss": 0.3528,
+      "step": 8476
+    },
+    {
+      "epoch": 1.2971690895179802,
+      "grad_norm": 2.081771653631222,
+      "learning_rate": 5.807973933311332e-06,
+      "loss": 0.3542,
+      "step": 8477
+    },
+    {
+      "epoch": 1.2973221117061975,
+      "grad_norm": 2.2464382632521436,
+      "learning_rate": 5.805724137730482e-06,
+      "loss": 0.3513,
+      "step": 8478
+    },
+    {
+      "epoch": 1.2974751338944146,
+      "grad_norm": 2.0902690420170305,
+      "learning_rate": 5.803474599736521e-06,
+      "loss": 0.334,
+      "step": 8479
+    },
+    {
+      "epoch": 1.297628156082632,
+      "grad_norm": 1.977981981638016,
+      "learning_rate": 5.801225319467602e-06,
+      "loss": 0.35,
+      "step": 8480
+    },
+    {
+      "epoch": 1.2977811782708493,
+      "grad_norm": 2.6637236988962694,
+      "learning_rate": 5.798976297061861e-06,
+      "loss": 0.4043,
+      "step": 8481
+    },
+    {
+      "epoch": 1.2979342004590666,
+      "grad_norm": 2.2270672736281747,
+      "learning_rate": 5.79672753265742e-06,
+      "loss": 0.3494,
+      "step": 8482
+    },
+    {
+      "epoch": 1.298087222647284,
+      "grad_norm": 2.0626813380311,
+      "learning_rate": 5.794479026392381e-06,
+      "loss": 0.3174,
+      "step": 8483
+    },
+    {
+      "epoch": 1.298240244835501,
+      "grad_norm": 2.392483617852613,
+      "learning_rate": 5.792230778404846e-06,
+      "loss": 0.3716,
+      "step": 8484
+    },
+    {
+      "epoch": 1.2983932670237184,
+      "grad_norm": 1.9548439126657664,
+      "learning_rate": 5.789982788832875e-06,
+      "loss": 0.3328,
+      "step": 8485
+    },
+    {
+      "epoch": 1.2985462892119357,
+      "grad_norm": 2.2712771203487097,
+      "learning_rate": 5.7877350578145265e-06,
+      "loss": 0.4683,
+      "step": 8486
+    },
+    {
+      "epoch": 1.298699311400153,
+      "grad_norm": 2.1349081647901396,
+      "learning_rate": 5.785487585487855e-06,
+      "loss": 0.3355,
+      "step": 8487
+    },
+    {
+      "epoch": 1.2988523335883704,
+      "grad_norm": 2.0144446923386585,
+      "learning_rate": 5.7832403719908735e-06,
+      "loss": 0.3535,
+      "step": 8488
+    },
+    {
+      "epoch": 1.2990053557765875,
+      "grad_norm": 2.323913427366677,
+      "learning_rate": 5.7809934174615935e-06,
+      "loss": 0.3846,
+      "step": 8489
+    },
+    {
+      "epoch": 1.2991583779648048,
+      "grad_norm": 2.2065277565739647,
+      "learning_rate": 5.77874672203802e-06,
+      "loss": 0.405,
+      "step": 8490
+    },
+    {
+      "epoch": 1.2993114001530222,
+      "grad_norm": 2.0701072367879694,
+      "learning_rate": 5.776500285858115e-06,
+      "loss": 0.3591,
+      "step": 8491
+    },
+    {
+      "epoch": 1.2994644223412395,
+      "grad_norm": 2.026467391159173,
+      "learning_rate": 5.774254109059856e-06,
+      "loss": 0.3801,
+      "step": 8492
+    },
+    {
+      "epoch": 1.2996174445294568,
+      "grad_norm": 2.03636713492077,
+      "learning_rate": 5.772008191781187e-06,
+      "loss": 0.4068,
+      "step": 8493
+    },
+    {
+      "epoch": 1.299770466717674,
+      "grad_norm": 2.235728964125411,
+      "learning_rate": 5.769762534160026e-06,
+      "loss": 0.4005,
+      "step": 8494
+    },
+    {
+      "epoch": 1.2999234889058915,
+      "grad_norm": 2.2226179990971597,
+      "learning_rate": 5.7675171363343e-06,
+      "loss": 0.3848,
+      "step": 8495
+    },
+    {
+      "epoch": 1.3000765110941086,
+      "grad_norm": 1.8548276674323887,
+      "learning_rate": 5.765271998441909e-06,
+      "loss": 0.2832,
+      "step": 8496
+    },
+    {
+      "epoch": 1.300229533282326,
+      "grad_norm": 2.058654357219985,
+      "learning_rate": 5.763027120620723e-06,
+      "loss": 0.3759,
+      "step": 8497
+    },
+    {
+      "epoch": 1.3003825554705433,
+      "grad_norm": 2.1669192089657128,
+      "learning_rate": 5.760782503008622e-06,
+      "loss": 0.3865,
+      "step": 8498
+    },
+    {
+      "epoch": 1.3005355776587606,
+      "grad_norm": 2.2800499466539854,
+      "learning_rate": 5.758538145743448e-06,
+      "loss": 0.4721,
+      "step": 8499
+    },
+    {
+      "epoch": 1.300688599846978,
+      "grad_norm": 1.8692911785721809,
+      "learning_rate": 5.7562940489630424e-06,
+      "loss": 0.3094,
+      "step": 8500
+    },
+    {
+      "epoch": 1.300841622035195,
+      "grad_norm": 2.3481643066715803,
+      "learning_rate": 5.75405021280522e-06,
+      "loss": 0.3394,
+      "step": 8501
+    },
+    {
+      "epoch": 1.3009946442234124,
+      "grad_norm": 2.3364862272013154,
+      "learning_rate": 5.751806637407783e-06,
+      "loss": 0.4305,
+      "step": 8502
+    },
+    {
+      "epoch": 1.3011476664116297,
+      "grad_norm": 2.106806897595338,
+      "learning_rate": 5.7495633229085205e-06,
+      "loss": 0.3885,
+      "step": 8503
+    },
+    {
+      "epoch": 1.301300688599847,
+      "grad_norm": 2.131244287327404,
+      "learning_rate": 5.747320269445203e-06,
+      "loss": 0.3757,
+      "step": 8504
+    },
+    {
+      "epoch": 1.3014537107880644,
+      "grad_norm": 1.9915545641442498,
+      "learning_rate": 5.745077477155582e-06,
+      "loss": 0.3304,
+      "step": 8505
+    },
+    {
+      "epoch": 1.3016067329762815,
+      "grad_norm": 1.932676633650167,
+      "learning_rate": 5.7428349461774e-06,
+      "loss": 0.3865,
+      "step": 8506
+    },
+    {
+      "epoch": 1.3017597551644988,
+      "grad_norm": 2.3799151262160496,
+      "learning_rate": 5.740592676648377e-06,
+      "loss": 0.366,
+      "step": 8507
+    },
+    {
+      "epoch": 1.3019127773527162,
+      "grad_norm": 2.394625708639643,
+      "learning_rate": 5.7383506687062215e-06,
+      "loss": 0.4086,
+      "step": 8508
+    },
+    {
+      "epoch": 1.3020657995409335,
+      "grad_norm": 1.9055085009689208,
+      "learning_rate": 5.736108922488624e-06,
+      "loss": 0.3348,
+      "step": 8509
+    },
+    {
+      "epoch": 1.3022188217291508,
+      "grad_norm": 2.053303277029518,
+      "learning_rate": 5.733867438133256e-06,
+      "loss": 0.3045,
+      "step": 8510
+    },
+    {
+      "epoch": 1.302371843917368,
+      "grad_norm": 2.169386364248647,
+      "learning_rate": 5.731626215777779e-06,
+      "loss": 0.3458,
+      "step": 8511
+    },
+    {
+      "epoch": 1.3025248661055853,
+      "grad_norm": 2.127208555931068,
+      "learning_rate": 5.72938525555983e-06,
+      "loss": 0.3355,
+      "step": 8512
+    },
+    {
+      "epoch": 1.3026778882938026,
+      "grad_norm": 2.0541456236284463,
+      "learning_rate": 5.727144557617047e-06,
+      "loss": 0.4235,
+      "step": 8513
+    },
+    {
+      "epoch": 1.30283091048202,
+      "grad_norm": 2.1306174022427053,
+      "learning_rate": 5.724904122087028e-06,
+      "loss": 0.3497,
+      "step": 8514
+    },
+    {
+      "epoch": 1.3029839326702373,
+      "grad_norm": 1.9520284290277785,
+      "learning_rate": 5.722663949107368e-06,
+      "loss": 0.4199,
+      "step": 8515
+    },
+    {
+      "epoch": 1.3031369548584544,
+      "grad_norm": 2.319002474956079,
+      "learning_rate": 5.720424038815655e-06,
+      "loss": 0.3666,
+      "step": 8516
+    },
+    {
+      "epoch": 1.3032899770466717,
+      "grad_norm": 2.193066933790192,
+      "learning_rate": 5.71818439134944e-06,
+      "loss": 0.3682,
+      "step": 8517
+    },
+    {
+      "epoch": 1.303442999234889,
+      "grad_norm": 2.064281725208092,
+      "learning_rate": 5.715945006846269e-06,
+      "loss": 0.3089,
+      "step": 8518
+    },
+    {
+      "epoch": 1.3035960214231064,
+      "grad_norm": 2.165584682981705,
+      "learning_rate": 5.713705885443681e-06,
+      "loss": 0.4026,
+      "step": 8519
+    },
+    {
+      "epoch": 1.3037490436113237,
+      "grad_norm": 2.1493436478964374,
+      "learning_rate": 5.711467027279174e-06,
+      "loss": 0.342,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3039020657995408,
+      "grad_norm": 1.7198630036611893,
+      "learning_rate": 5.709228432490259e-06,
+      "loss": 0.3119,
+      "step": 8521
+    },
+    {
+      "epoch": 1.3040550879877582,
+      "grad_norm": 2.053429899418811,
+      "learning_rate": 5.7069901012144155e-06,
+      "loss": 0.3461,
+      "step": 8522
+    },
+    {
+      "epoch": 1.3042081101759755,
+      "grad_norm": 2.191895371424377,
+      "learning_rate": 5.704752033589096e-06,
+      "loss": 0.3762,
+      "step": 8523
+    },
+    {
+      "epoch": 1.3043611323641928,
+      "grad_norm": 2.176273855972572,
+      "learning_rate": 5.702514229751761e-06,
+      "loss": 0.3912,
+      "step": 8524
+    },
+    {
+      "epoch": 1.3045141545524102,
+      "grad_norm": 2.1364346441788316,
+      "learning_rate": 5.70027668983984e-06,
+      "loss": 0.3702,
+      "step": 8525
+    },
+    {
+      "epoch": 1.3046671767406273,
+      "grad_norm": 2.1394636414100354,
+      "learning_rate": 5.698039413990748e-06,
+      "loss": 0.3959,
+      "step": 8526
+    },
+    {
+      "epoch": 1.3048201989288448,
+      "grad_norm": 2.0727055307523643,
+      "learning_rate": 5.6958024023418834e-06,
+      "loss": 0.3547,
+      "step": 8527
+    },
+    {
+      "epoch": 1.304973221117062,
+      "grad_norm": 2.044965766447203,
+      "learning_rate": 5.6935656550306326e-06,
+      "loss": 0.3701,
+      "step": 8528
+    },
+    {
+      "epoch": 1.3051262433052793,
+      "grad_norm": 2.259607204648534,
+      "learning_rate": 5.691329172194362e-06,
+      "loss": 0.3761,
+      "step": 8529
+    },
+    {
+      "epoch": 1.3052792654934966,
+      "grad_norm": 2.1227844121777326,
+      "learning_rate": 5.689092953970421e-06,
+      "loss": 0.4002,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3054322876817137,
+      "grad_norm": 2.2885565622270008,
+      "learning_rate": 5.6868570004961474e-06,
+      "loss": 0.3782,
+      "step": 8531
+    },
+    {
+      "epoch": 1.3055853098699313,
+      "grad_norm": 2.2684685389350303,
+      "learning_rate": 5.684621311908859e-06,
+      "loss": 0.3547,
+      "step": 8532
+    },
+    {
+      "epoch": 1.3057383320581484,
+      "grad_norm": 2.2468869954835675,
+      "learning_rate": 5.682385888345856e-06,
+      "loss": 0.315,
+      "step": 8533
+    },
+    {
+      "epoch": 1.3058913542463657,
+      "grad_norm": 2.2885821903709296,
+      "learning_rate": 5.680150729944427e-06,
+      "loss": 0.3492,
+      "step": 8534
+    },
+    {
+      "epoch": 1.306044376434583,
+      "grad_norm": 2.2283745517249582,
+      "learning_rate": 5.677915836841842e-06,
+      "loss": 0.4227,
+      "step": 8535
+    },
+    {
+      "epoch": 1.3061973986228004,
+      "grad_norm": 2.10231696253404,
+      "learning_rate": 5.675681209175349e-06,
+      "loss": 0.3635,
+      "step": 8536
+    },
+    {
+      "epoch": 1.3063504208110177,
+      "grad_norm": 2.233717020395182,
+      "learning_rate": 5.673446847082197e-06,
+      "loss": 0.4036,
+      "step": 8537
+    },
+    {
+      "epoch": 1.3065034429992348,
+      "grad_norm": 1.9310872009083184,
+      "learning_rate": 5.671212750699597e-06,
+      "loss": 0.2954,
+      "step": 8538
+    },
+    {
+      "epoch": 1.3066564651874522,
+      "grad_norm": 2.030845760878608,
+      "learning_rate": 5.6689789201647505e-06,
+      "loss": 0.3745,
+      "step": 8539
+    },
+    {
+      "epoch": 1.3068094873756695,
+      "grad_norm": 1.8189427337342248,
+      "learning_rate": 5.66674535561486e-06,
+      "loss": 0.2948,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3069625095638868,
+      "grad_norm": 2.0469441672632027,
+      "learning_rate": 5.664512057187081e-06,
+      "loss": 0.3323,
+      "step": 8541
+    },
+    {
+      "epoch": 1.3071155317521042,
+      "grad_norm": 2.0924553585486403,
+      "learning_rate": 5.662279025018582e-06,
+      "loss": 0.407,
+      "step": 8542
+    },
+    {
+      "epoch": 1.3072685539403213,
+      "grad_norm": 2.3896033125336005,
+      "learning_rate": 5.660046259246501e-06,
+      "loss": 0.4156,
+      "step": 8543
+    },
+    {
+      "epoch": 1.3074215761285386,
+      "grad_norm": 2.1963052778718253,
+      "learning_rate": 5.65781376000795e-06,
+      "loss": 0.3667,
+      "step": 8544
+    },
+    {
+      "epoch": 1.307574598316756,
+      "grad_norm": 2.2404897511730404,
+      "learning_rate": 5.6555815274400464e-06,
+      "loss": 0.3412,
+      "step": 8545
+    },
+    {
+      "epoch": 1.3077276205049733,
+      "grad_norm": 1.976985343614817,
+      "learning_rate": 5.65334956167988e-06,
+      "loss": 0.3942,
+      "step": 8546
+    },
+    {
+      "epoch": 1.3078806426931906,
+      "grad_norm": 2.260212423328555,
+      "learning_rate": 5.651117862864512e-06,
+      "loss": 0.3259,
+      "step": 8547
+    },
+    {
+      "epoch": 1.3080336648814077,
+      "grad_norm": 2.164028887318763,
+      "learning_rate": 5.6488864311310155e-06,
+      "loss": 0.3614,
+      "step": 8548
+    },
+    {
+      "epoch": 1.308186687069625,
+      "grad_norm": 2.30626799094876,
+      "learning_rate": 5.646655266616423e-06,
+      "loss": 0.4431,
+      "step": 8549
+    },
+    {
+      "epoch": 1.3083397092578424,
+      "grad_norm": 1.9692214406859057,
+      "learning_rate": 5.6444243694577615e-06,
+      "loss": 0.4161,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3084927314460597,
+      "grad_norm": 2.1853722471880594,
+      "learning_rate": 5.642193739792039e-06,
+      "loss": 0.3402,
+      "step": 8551
+    },
+    {
+      "epoch": 1.308645753634277,
+      "grad_norm": 2.0970827282400992,
+      "learning_rate": 5.639963377756247e-06,
+      "loss": 0.4279,
+      "step": 8552
+    },
+    {
+      "epoch": 1.3087987758224942,
+      "grad_norm": 2.177870812573683,
+      "learning_rate": 5.63773328348736e-06,
+      "loss": 0.3626,
+      "step": 8553
+    },
+    {
+      "epoch": 1.3089517980107115,
+      "grad_norm": 2.591283101525571,
+      "learning_rate": 5.635503457122338e-06,
+      "loss": 0.4168,
+      "step": 8554
+    },
+    {
+      "epoch": 1.3091048201989288,
+      "grad_norm": 1.9694916240067648,
+      "learning_rate": 5.633273898798122e-06,
+      "loss": 0.3234,
+      "step": 8555
+    },
+    {
+      "epoch": 1.3092578423871462,
+      "grad_norm": 2.290182102724067,
+      "learning_rate": 5.631044608651639e-06,
+      "loss": 0.3949,
+      "step": 8556
+    },
+    {
+      "epoch": 1.3094108645753635,
+      "grad_norm": 2.276801072822145,
+      "learning_rate": 5.628815586819797e-06,
+      "loss": 0.3722,
+      "step": 8557
+    },
+    {
+      "epoch": 1.3095638867635806,
+      "grad_norm": 2.1922935542275392,
+      "learning_rate": 5.626586833439491e-06,
+      "loss": 0.2876,
+      "step": 8558
+    },
+    {
+      "epoch": 1.309716908951798,
+      "grad_norm": 2.125655685633201,
+      "learning_rate": 5.6243583486475965e-06,
+      "loss": 0.3184,
+      "step": 8559
+    },
+    {
+      "epoch": 1.3098699311400153,
+      "grad_norm": 2.290520817699174,
+      "learning_rate": 5.622130132580973e-06,
+      "loss": 0.4141,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3100229533282326,
+      "grad_norm": 2.162194717394431,
+      "learning_rate": 5.619902185376465e-06,
+      "loss": 0.4075,
+      "step": 8561
+    },
+    {
+      "epoch": 1.31017597551645,
+      "grad_norm": 2.300845062414474,
+      "learning_rate": 5.617674507170899e-06,
+      "loss": 0.3332,
+      "step": 8562
+    },
+    {
+      "epoch": 1.310328997704667,
+      "grad_norm": 2.2539854405846516,
+      "learning_rate": 5.615447098101083e-06,
+      "loss": 0.3666,
+      "step": 8563
+    },
+    {
+      "epoch": 1.3104820198928846,
+      "grad_norm": 2.0871045714190015,
+      "learning_rate": 5.613219958303815e-06,
+      "loss": 0.3525,
+      "step": 8564
+    },
+    {
+      "epoch": 1.3106350420811017,
+      "grad_norm": 2.2476746690566896,
+      "learning_rate": 5.6109930879158656e-06,
+      "loss": 0.3941,
+      "step": 8565
+    },
+    {
+      "epoch": 1.310788064269319,
+      "grad_norm": 2.111502301320883,
+      "learning_rate": 5.60876648707401e-06,
+      "loss": 0.3222,
+      "step": 8566
+    },
+    {
+      "epoch": 1.3109410864575364,
+      "grad_norm": 2.217927916834475,
+      "learning_rate": 5.606540155914978e-06,
+      "loss": 0.3817,
+      "step": 8567
+    },
+    {
+      "epoch": 1.3110941086457537,
+      "grad_norm": 2.1948887182227312,
+      "learning_rate": 5.604314094575498e-06,
+      "loss": 0.3711,
+      "step": 8568
+    },
+    {
+      "epoch": 1.311247130833971,
+      "grad_norm": 2.376746522845046,
+      "learning_rate": 5.602088303192295e-06,
+      "loss": 0.4375,
+      "step": 8569
+    },
+    {
+      "epoch": 1.3114001530221882,
+      "grad_norm": 2.0035516123023993,
+      "learning_rate": 5.599862781902049e-06,
+      "loss": 0.3363,
+      "step": 8570
+    },
+    {
+      "epoch": 1.3115531752104055,
+      "grad_norm": 2.0738010508442755,
+      "learning_rate": 5.59763753084144e-06,
+      "loss": 0.3137,
+      "step": 8571
+    },
+    {
+      "epoch": 1.3117061973986228,
+      "grad_norm": 2.0745210501379217,
+      "learning_rate": 5.595412550147139e-06,
+      "loss": 0.3249,
+      "step": 8572
+    },
+    {
+      "epoch": 1.3118592195868402,
+      "grad_norm": 2.267526336468913,
+      "learning_rate": 5.593187839955776e-06,
+      "loss": 0.3783,
+      "step": 8573
+    },
+    {
+      "epoch": 1.3120122417750575,
+      "grad_norm": 2.3694203084709935,
+      "learning_rate": 5.590963400403993e-06,
+      "loss": 0.3513,
+      "step": 8574
+    },
+    {
+      "epoch": 1.3121652639632746,
+      "grad_norm": 2.012157483197845,
+      "learning_rate": 5.588739231628397e-06,
+      "loss": 0.2914,
+      "step": 8575
+    },
+    {
+      "epoch": 1.312318286151492,
+      "grad_norm": 2.0659451699838502,
+      "learning_rate": 5.586515333765577e-06,
+      "loss": 0.3701,
+      "step": 8576
+    },
+    {
+      "epoch": 1.3124713083397093,
+      "grad_norm": 1.921731703819784,
+      "learning_rate": 5.584291706952119e-06,
+      "loss": 0.2673,
+      "step": 8577
+    },
+    {
+      "epoch": 1.3126243305279266,
+      "grad_norm": 2.1435100630609893,
+      "learning_rate": 5.58206835132458e-06,
+      "loss": 0.3065,
+      "step": 8578
+    },
+    {
+      "epoch": 1.312777352716144,
+      "grad_norm": 2.178374984629396,
+      "learning_rate": 5.579845267019508e-06,
+      "loss": 0.3687,
+      "step": 8579
+    },
+    {
+      "epoch": 1.312930374904361,
+      "grad_norm": 2.4781271076321207,
+      "learning_rate": 5.577622454173429e-06,
+      "loss": 0.4722,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3130833970925784,
+      "grad_norm": 2.085509746046495,
+      "learning_rate": 5.5753999129228565e-06,
+      "loss": 0.3132,
+      "step": 8581
+    },
+    {
+      "epoch": 1.3132364192807957,
+      "grad_norm": 1.9262893935758099,
+      "learning_rate": 5.5731776434042835e-06,
+      "loss": 0.3986,
+      "step": 8582
+    },
+    {
+      "epoch": 1.313389441469013,
+      "grad_norm": 1.9838375146009155,
+      "learning_rate": 5.57095564575419e-06,
+      "loss": 0.2953,
+      "step": 8583
+    },
+    {
+      "epoch": 1.3135424636572304,
+      "grad_norm": 2.10472370597914,
+      "learning_rate": 5.568733920109038e-06,
+      "loss": 0.3327,
+      "step": 8584
+    },
+    {
+      "epoch": 1.3136954858454475,
+      "grad_norm": 1.9445119541733116,
+      "learning_rate": 5.5665124666052695e-06,
+      "loss": 0.3065,
+      "step": 8585
+    },
+    {
+      "epoch": 1.3138485080336648,
+      "grad_norm": 2.603442112632636,
+      "learning_rate": 5.564291285379314e-06,
+      "loss": 0.3505,
+      "step": 8586
+    },
+    {
+      "epoch": 1.3140015302218822,
+      "grad_norm": 2.0445384664789885,
+      "learning_rate": 5.562070376567585e-06,
+      "loss": 0.3704,
+      "step": 8587
+    },
+    {
+      "epoch": 1.3141545524100995,
+      "grad_norm": 2.1157344605778827,
+      "learning_rate": 5.559849740306474e-06,
+      "loss": 0.3614,
+      "step": 8588
+    },
+    {
+      "epoch": 1.3143075745983168,
+      "grad_norm": 2.13504060924164,
+      "learning_rate": 5.5576293767323606e-06,
+      "loss": 0.3657,
+      "step": 8589
+    },
+    {
+      "epoch": 1.314460596786534,
+      "grad_norm": 2.289321737350508,
+      "learning_rate": 5.555409285981605e-06,
+      "loss": 0.3704,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3146136189747513,
+      "grad_norm": 2.1213736767351055,
+      "learning_rate": 5.553189468190554e-06,
+      "loss": 0.3293,
+      "step": 8591
+    },
+    {
+      "epoch": 1.3147666411629686,
+      "grad_norm": 2.2923070600542075,
+      "learning_rate": 5.550969923495533e-06,
+      "loss": 0.3757,
+      "step": 8592
+    },
+    {
+      "epoch": 1.314919663351186,
+      "grad_norm": 1.9832544875286664,
+      "learning_rate": 5.548750652032855e-06,
+      "loss": 0.2989,
+      "step": 8593
+    },
+    {
+      "epoch": 1.3150726855394033,
+      "grad_norm": 2.1527441947882116,
+      "learning_rate": 5.546531653938807e-06,
+      "loss": 0.3381,
+      "step": 8594
+    },
+    {
+      "epoch": 1.3152257077276204,
+      "grad_norm": 2.212591146712386,
+      "learning_rate": 5.5443129293496825e-06,
+      "loss": 0.3609,
+      "step": 8595
+    },
+    {
+      "epoch": 1.315378729915838,
+      "grad_norm": 2.2950267556502757,
+      "learning_rate": 5.542094478401725e-06,
+      "loss": 0.4125,
+      "step": 8596
+    },
+    {
+      "epoch": 1.315531752104055,
+      "grad_norm": 2.0985948379512593,
+      "learning_rate": 5.539876301231182e-06,
+      "loss": 0.3706,
+      "step": 8597
+    },
+    {
+      "epoch": 1.3156847742922724,
+      "grad_norm": 2.167223206670299,
+      "learning_rate": 5.537658397974293e-06,
+      "loss": 0.3895,
+      "step": 8598
+    },
+    {
+      "epoch": 1.3158377964804897,
+      "grad_norm": 2.4820509629216074,
+      "learning_rate": 5.535440768767253e-06,
+      "loss": 0.3807,
+      "step": 8599
+    },
+    {
+      "epoch": 1.315990818668707,
+      "grad_norm": 2.2159112092687896,
+      "learning_rate": 5.533223413746256e-06,
+      "loss": 0.3477,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3161438408569244,
+      "grad_norm": 2.2364380030208326,
+      "learning_rate": 5.531006333047493e-06,
+      "loss": 0.3696,
+      "step": 8601
+    },
+    {
+      "epoch": 1.3162968630451415,
+      "grad_norm": 1.829042244844607,
+      "learning_rate": 5.528789526807105e-06,
+      "loss": 0.2988,
+      "step": 8602
+    },
+    {
+      "epoch": 1.3164498852333588,
+      "grad_norm": 2.061590745235742,
+      "learning_rate": 5.526572995161248e-06,
+      "loss": 0.3311,
+      "step": 8603
+    },
+    {
+      "epoch": 1.3166029074215762,
+      "grad_norm": 2.133150535391964,
+      "learning_rate": 5.524356738246048e-06,
+      "loss": 0.3052,
+      "step": 8604
+    },
+    {
+      "epoch": 1.3167559296097935,
+      "grad_norm": 2.0168388184124106,
+      "learning_rate": 5.5221407561976e-06,
+      "loss": 0.3656,
+      "step": 8605
+    },
+    {
+      "epoch": 1.3169089517980108,
+      "grad_norm": 2.2695540004967096,
+      "learning_rate": 5.519925049152011e-06,
+      "loss": 0.3959,
+      "step": 8606
+    },
+    {
+      "epoch": 1.317061973986228,
+      "grad_norm": 2.305770870691185,
+      "learning_rate": 5.51770961724535e-06,
+      "loss": 0.4459,
+      "step": 8607
+    },
+    {
+      "epoch": 1.3172149961744453,
+      "grad_norm": 2.2371126516169797,
+      "learning_rate": 5.515494460613678e-06,
+      "loss": 0.3856,
+      "step": 8608
+    },
+    {
+      "epoch": 1.3173680183626626,
+      "grad_norm": 2.2476766777179322,
+      "learning_rate": 5.513279579393034e-06,
+      "loss": 0.3237,
+      "step": 8609
+    },
+    {
+      "epoch": 1.31752104055088,
+      "grad_norm": 2.010718411909031,
+      "learning_rate": 5.511064973719442e-06,
+      "loss": 0.4544,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3176740627390973,
+      "grad_norm": 1.8265150494873308,
+      "learning_rate": 5.508850643728911e-06,
+      "loss": 0.2927,
+      "step": 8611
+    },
+    {
+      "epoch": 1.3178270849273144,
+      "grad_norm": 2.1922469791566797,
+      "learning_rate": 5.506636589557433e-06,
+      "loss": 0.3636,
+      "step": 8612
+    },
+    {
+      "epoch": 1.3179801071155317,
+      "grad_norm": 1.9522370911195468,
+      "learning_rate": 5.504422811340977e-06,
+      "loss": 0.345,
+      "step": 8613
+    },
+    {
+      "epoch": 1.318133129303749,
+      "grad_norm": 1.8985775896312533,
+      "learning_rate": 5.502209309215505e-06,
+      "loss": 0.3437,
+      "step": 8614
+    },
+    {
+      "epoch": 1.3182861514919664,
+      "grad_norm": 2.531908097820924,
+      "learning_rate": 5.499996083316952e-06,
+      "loss": 0.3961,
+      "step": 8615
+    },
+    {
+      "epoch": 1.3184391736801837,
+      "grad_norm": 2.300599284977995,
+      "learning_rate": 5.49778313378125e-06,
+      "loss": 0.3706,
+      "step": 8616
+    },
+    {
+      "epoch": 1.3185921958684008,
+      "grad_norm": 1.9660351647994214,
+      "learning_rate": 5.495570460744295e-06,
+      "loss": 0.3349,
+      "step": 8617
+    },
+    {
+      "epoch": 1.3187452180566182,
+      "grad_norm": 2.320828623988891,
+      "learning_rate": 5.4933580643419744e-06,
+      "loss": 0.3809,
+      "step": 8618
+    },
+    {
+      "epoch": 1.3188982402448355,
+      "grad_norm": 2.2384342613355557,
+      "learning_rate": 5.491145944710174e-06,
+      "loss": 0.3401,
+      "step": 8619
+    },
+    {
+      "epoch": 1.3190512624330528,
+      "grad_norm": 1.8784193029515464,
+      "learning_rate": 5.488934101984737e-06,
+      "loss": 0.3359,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3192042846212702,
+      "grad_norm": 2.22885101285974,
+      "learning_rate": 5.486722536301499e-06,
+      "loss": 0.3814,
+      "step": 8621
+    },
+    {
+      "epoch": 1.3193573068094873,
+      "grad_norm": 2.518679488884114,
+      "learning_rate": 5.4845112477962936e-06,
+      "loss": 0.4485,
+      "step": 8622
+    },
+    {
+      "epoch": 1.3195103289977046,
+      "grad_norm": 2.2918568209225616,
+      "learning_rate": 5.482300236604909e-06,
+      "loss": 0.3204,
+      "step": 8623
+    },
+    {
+      "epoch": 1.319663351185922,
+      "grad_norm": 2.483464519209941,
+      "learning_rate": 5.4800895028631466e-06,
+      "loss": 0.3592,
+      "step": 8624
+    },
+    {
+      "epoch": 1.3198163733741393,
+      "grad_norm": 2.2598966067125104,
+      "learning_rate": 5.477879046706771e-06,
+      "loss": 0.373,
+      "step": 8625
+    },
+    {
+      "epoch": 1.3199693955623566,
+      "grad_norm": 2.162251025832532,
+      "learning_rate": 5.475668868271527e-06,
+      "loss": 0.3023,
+      "step": 8626
+    },
+    {
+      "epoch": 1.3201224177505737,
+      "grad_norm": 2.055830955710422,
+      "learning_rate": 5.473458967693162e-06,
+      "loss": 0.3665,
+      "step": 8627
+    },
+    {
+      "epoch": 1.3202754399387913,
+      "grad_norm": 2.3676870900519495,
+      "learning_rate": 5.471249345107392e-06,
+      "loss": 0.3866,
+      "step": 8628
+    },
+    {
+      "epoch": 1.3204284621270084,
+      "grad_norm": 2.2301256287820377,
+      "learning_rate": 5.469040000649909e-06,
+      "loss": 0.45,
+      "step": 8629
+    },
+    {
+      "epoch": 1.3205814843152257,
+      "grad_norm": 2.250482208887682,
+      "learning_rate": 5.466830934456409e-06,
+      "loss": 0.2956,
+      "step": 8630
+    },
+    {
+      "epoch": 1.320734506503443,
+      "grad_norm": 1.9602055867205517,
+      "learning_rate": 5.464622146662555e-06,
+      "loss": 0.3221,
+      "step": 8631
+    },
+    {
+      "epoch": 1.3208875286916604,
+      "grad_norm": 1.991138809691683,
+      "learning_rate": 5.462413637403997e-06,
+      "loss": 0.3425,
+      "step": 8632
+    },
+    {
+      "epoch": 1.3210405508798777,
+      "grad_norm": 2.265706180431464,
+      "learning_rate": 5.4602054068163705e-06,
+      "loss": 0.3718,
+      "step": 8633
+    },
+    {
+      "epoch": 1.3211935730680948,
+      "grad_norm": 2.437289055001441,
+      "learning_rate": 5.457997455035286e-06,
+      "loss": 0.3938,
+      "step": 8634
+    },
+    {
+      "epoch": 1.3213465952563122,
+      "grad_norm": 2.093157026239112,
+      "learning_rate": 5.45578978219635e-06,
+      "loss": 0.3366,
+      "step": 8635
+    },
+    {
+      "epoch": 1.3214996174445295,
+      "grad_norm": 2.113182376884977,
+      "learning_rate": 5.453582388435139e-06,
+      "loss": 0.3187,
+      "step": 8636
+    },
+    {
+      "epoch": 1.3216526396327468,
+      "grad_norm": 2.3073790380831802,
+      "learning_rate": 5.451375273887219e-06,
+      "loss": 0.3507,
+      "step": 8637
+    },
+    {
+      "epoch": 1.3218056618209642,
+      "grad_norm": 2.0134841418393075,
+      "learning_rate": 5.449168438688138e-06,
+      "loss": 0.3791,
+      "step": 8638
+    },
+    {
+      "epoch": 1.3219586840091813,
+      "grad_norm": 2.1274435392661935,
+      "learning_rate": 5.446961882973427e-06,
+      "loss": 0.3967,
+      "step": 8639
+    },
+    {
+      "epoch": 1.3221117061973986,
+      "grad_norm": 2.2861859533824056,
+      "learning_rate": 5.444755606878598e-06,
+      "loss": 0.3898,
+      "step": 8640
+    },
+    {
+      "epoch": 1.322264728385616,
+      "grad_norm": 2.4796974235976386,
+      "learning_rate": 5.442549610539146e-06,
+      "loss": 0.428,
+      "step": 8641
+    },
+    {
+      "epoch": 1.3224177505738333,
+      "grad_norm": 2.338041372087975,
+      "learning_rate": 5.440343894090552e-06,
+      "loss": 0.3597,
+      "step": 8642
+    },
+    {
+      "epoch": 1.3225707727620506,
+      "grad_norm": 2.2580701341379497,
+      "learning_rate": 5.438138457668277e-06,
+      "loss": 0.4203,
+      "step": 8643
+    },
+    {
+      "epoch": 1.3227237949502677,
+      "grad_norm": 2.1502655499210204,
+      "learning_rate": 5.435933301407765e-06,
+      "loss": 0.2846,
+      "step": 8644
+    },
+    {
+      "epoch": 1.322876817138485,
+      "grad_norm": 2.3728101853670616,
+      "learning_rate": 5.433728425444444e-06,
+      "loss": 0.3727,
+      "step": 8645
+    },
+    {
+      "epoch": 1.3230298393267024,
+      "grad_norm": 2.45813606253095,
+      "learning_rate": 5.431523829913725e-06,
+      "loss": 0.4305,
+      "step": 8646
+    },
+    {
+      "epoch": 1.3231828615149197,
+      "grad_norm": 1.9190440302219542,
+      "learning_rate": 5.4293195149509945e-06,
+      "loss": 0.3219,
+      "step": 8647
+    },
+    {
+      "epoch": 1.323335883703137,
+      "grad_norm": 2.0409914653263064,
+      "learning_rate": 5.427115480691642e-06,
+      "loss": 0.33,
+      "step": 8648
+    },
+    {
+      "epoch": 1.3234889058913542,
+      "grad_norm": 2.264838190331384,
+      "learning_rate": 5.424911727271012e-06,
+      "loss": 0.3573,
+      "step": 8649
+    },
+    {
+      "epoch": 1.3236419280795715,
+      "grad_norm": 2.1410706263548462,
+      "learning_rate": 5.422708254824447e-06,
+      "loss": 0.3522,
+      "step": 8650
+    },
+    {
+      "epoch": 1.3237949502677888,
+      "grad_norm": 2.4540891731226884,
+      "learning_rate": 5.420505063487282e-06,
+      "loss": 0.4584,
+      "step": 8651
+    },
+    {
+      "epoch": 1.3239479724560062,
+      "grad_norm": 2.248809644345026,
+      "learning_rate": 5.418302153394809e-06,
+      "loss": 0.2998,
+      "step": 8652
+    },
+    {
+      "epoch": 1.3241009946442235,
+      "grad_norm": 2.0834229656833516,
+      "learning_rate": 5.4160995246823275e-06,
+      "loss": 0.3484,
+      "step": 8653
+    },
+    {
+      "epoch": 1.3242540168324406,
+      "grad_norm": 2.017276455572411,
+      "learning_rate": 5.413897177485112e-06,
+      "loss": 0.3551,
+      "step": 8654
+    },
+    {
+      "epoch": 1.324407039020658,
+      "grad_norm": 2.1939725955162674,
+      "learning_rate": 5.411695111938402e-06,
+      "loss": 0.3737,
+      "step": 8655
+    },
+    {
+      "epoch": 1.3245600612088753,
+      "grad_norm": 2.5776750204395222,
+      "learning_rate": 5.409493328177451e-06,
+      "loss": 0.3604,
+      "step": 8656
+    },
+    {
+      "epoch": 1.3247130833970926,
+      "grad_norm": 2.1881840047583956,
+      "learning_rate": 5.407291826337475e-06,
+      "loss": 0.3046,
+      "step": 8657
+    },
+    {
+      "epoch": 1.32486610558531,
+      "grad_norm": 2.282370930865902,
+      "learning_rate": 5.405090606553667e-06,
+      "loss": 0.3554,
+      "step": 8658
+    },
+    {
+      "epoch": 1.325019127773527,
+      "grad_norm": 2.366658092343715,
+      "learning_rate": 5.402889668961225e-06,
+      "loss": 0.4061,
+      "step": 8659
+    },
+    {
+      "epoch": 1.3251721499617444,
+      "grad_norm": 2.232569419388663,
+      "learning_rate": 5.400689013695311e-06,
+      "loss": 0.3978,
+      "step": 8660
+    },
+    {
+      "epoch": 1.3253251721499617,
+      "grad_norm": 2.129469851628214,
+      "learning_rate": 5.398488640891079e-06,
+      "loss": 0.3334,
+      "step": 8661
+    },
+    {
+      "epoch": 1.325478194338179,
+      "grad_norm": 2.4309004407524863,
+      "learning_rate": 5.3962885506836594e-06,
+      "loss": 0.415,
+      "step": 8662
+    },
+    {
+      "epoch": 1.3256312165263964,
+      "grad_norm": 2.0085523161120835,
+      "learning_rate": 5.39408874320817e-06,
+      "loss": 0.3572,
+      "step": 8663
+    },
+    {
+      "epoch": 1.3257842387146135,
+      "grad_norm": 2.117619652748608,
+      "learning_rate": 5.391889218599711e-06,
+      "loss": 0.3793,
+      "step": 8664
+    },
+    {
+      "epoch": 1.325937260902831,
+      "grad_norm": 2.451783913784382,
+      "learning_rate": 5.38968997699336e-06,
+      "loss": 0.4311,
+      "step": 8665
+    },
+    {
+      "epoch": 1.3260902830910481,
+      "grad_norm": 2.1107048901080963,
+      "learning_rate": 5.387491018524184e-06,
+      "loss": 0.3542,
+      "step": 8666
+    },
+    {
+      "epoch": 1.3262433052792655,
+      "grad_norm": 2.220269332344632,
+      "learning_rate": 5.385292343327229e-06,
+      "loss": 0.3762,
+      "step": 8667
+    },
+    {
+      "epoch": 1.3263963274674828,
+      "grad_norm": 2.3319220579245328,
+      "learning_rate": 5.383093951537524e-06,
+      "loss": 0.4135,
+      "step": 8668
+    },
+    {
+      "epoch": 1.3265493496557001,
+      "grad_norm": 2.1507183519468236,
+      "learning_rate": 5.380895843290079e-06,
+      "loss": 0.3431,
+      "step": 8669
+    },
+    {
+      "epoch": 1.3267023718439175,
+      "grad_norm": 2.010715555568859,
+      "learning_rate": 5.3786980187198925e-06,
+      "loss": 0.3584,
+      "step": 8670
+    },
+    {
+      "epoch": 1.3268553940321346,
+      "grad_norm": 2.335760014272626,
+      "learning_rate": 5.376500477961939e-06,
+      "loss": 0.3866,
+      "step": 8671
+    },
+    {
+      "epoch": 1.327008416220352,
+      "grad_norm": 2.3548091082451705,
+      "learning_rate": 5.374303221151178e-06,
+      "loss": 0.4641,
+      "step": 8672
+    },
+    {
+      "epoch": 1.3271614384085693,
+      "grad_norm": 2.077535323157751,
+      "learning_rate": 5.372106248422551e-06,
+      "loss": 0.3462,
+      "step": 8673
+    },
+    {
+      "epoch": 1.3273144605967866,
+      "grad_norm": 2.150316112354047,
+      "learning_rate": 5.369909559910985e-06,
+      "loss": 0.3316,
+      "step": 8674
+    },
+    {
+      "epoch": 1.327467482785004,
+      "grad_norm": 2.0408943761398035,
+      "learning_rate": 5.367713155751386e-06,
+      "loss": 0.2938,
+      "step": 8675
+    },
+    {
+      "epoch": 1.327620504973221,
+      "grad_norm": 2.0607758587739786,
+      "learning_rate": 5.36551703607864e-06,
+      "loss": 0.3099,
+      "step": 8676
+    },
+    {
+      "epoch": 1.3277735271614384,
+      "grad_norm": 2.173805194803525,
+      "learning_rate": 5.3633212010276295e-06,
+      "loss": 0.3144,
+      "step": 8677
+    },
+    {
+      "epoch": 1.3279265493496557,
+      "grad_norm": 2.1294183601816563,
+      "learning_rate": 5.3611256507332e-06,
+      "loss": 0.3466,
+      "step": 8678
+    },
+    {
+      "epoch": 1.328079571537873,
+      "grad_norm": 2.0559929103742736,
+      "learning_rate": 5.358930385330188e-06,
+      "loss": 0.3814,
+      "step": 8679
+    },
+    {
+      "epoch": 1.3282325937260904,
+      "grad_norm": 2.164937780009964,
+      "learning_rate": 5.356735404953424e-06,
+      "loss": 0.336,
+      "step": 8680
+    },
+    {
+      "epoch": 1.3283856159143075,
+      "grad_norm": 2.0680556793090505,
+      "learning_rate": 5.354540709737696e-06,
+      "loss": 0.3517,
+      "step": 8681
+    },
+    {
+      "epoch": 1.3285386381025248,
+      "grad_norm": 2.2848427066316033,
+      "learning_rate": 5.352346299817799e-06,
+      "loss": 0.3799,
+      "step": 8682
+    },
+    {
+      "epoch": 1.3286916602907421,
+      "grad_norm": 1.862547574716012,
+      "learning_rate": 5.350152175328502e-06,
+      "loss": 0.3432,
+      "step": 8683
+    },
+    {
+      "epoch": 1.3288446824789595,
+      "grad_norm": 2.1285348894501404,
+      "learning_rate": 5.347958336404544e-06,
+      "loss": 0.3473,
+      "step": 8684
+    },
+    {
+      "epoch": 1.3289977046671768,
+      "grad_norm": 2.15008899321762,
+      "learning_rate": 5.345764783180665e-06,
+      "loss": 0.356,
+      "step": 8685
+    },
+    {
+      "epoch": 1.329150726855394,
+      "grad_norm": 2.3855541790397847,
+      "learning_rate": 5.343571515791584e-06,
+      "loss": 0.4541,
+      "step": 8686
+    },
+    {
+      "epoch": 1.3293037490436113,
+      "grad_norm": 2.0715480983178383,
+      "learning_rate": 5.3413785343719835e-06,
+      "loss": 0.3749,
+      "step": 8687
+    },
+    {
+      "epoch": 1.3294567712318286,
+      "grad_norm": 2.2500551363113463,
+      "learning_rate": 5.339185839056555e-06,
+      "loss": 0.3332,
+      "step": 8688
+    },
+    {
+      "epoch": 1.329609793420046,
+      "grad_norm": 2.391261775337038,
+      "learning_rate": 5.336993429979958e-06,
+      "loss": 0.4095,
+      "step": 8689
+    },
+    {
+      "epoch": 1.3297628156082633,
+      "grad_norm": 2.1836965124937944,
+      "learning_rate": 5.3348013072768365e-06,
+      "loss": 0.3513,
+      "step": 8690
+    },
+    {
+      "epoch": 1.3299158377964804,
+      "grad_norm": 2.3000977897605535,
+      "learning_rate": 5.332609471081818e-06,
+      "loss": 0.4085,
+      "step": 8691
+    },
+    {
+      "epoch": 1.3300688599846977,
+      "grad_norm": 1.9898087771318698,
+      "learning_rate": 5.330417921529509e-06,
+      "loss": 0.3931,
+      "step": 8692
+    },
+    {
+      "epoch": 1.330221882172915,
+      "grad_norm": 2.323604812423909,
+      "learning_rate": 5.328226658754503e-06,
+      "loss": 0.3886,
+      "step": 8693
+    },
+    {
+      "epoch": 1.3303749043611324,
+      "grad_norm": 1.9935416913349375,
+      "learning_rate": 5.326035682891375e-06,
+      "loss": 0.2802,
+      "step": 8694
+    },
+    {
+      "epoch": 1.3305279265493497,
+      "grad_norm": 2.0321753727504985,
+      "learning_rate": 5.3238449940746805e-06,
+      "loss": 0.3618,
+      "step": 8695
+    },
+    {
+      "epoch": 1.3306809487375668,
+      "grad_norm": 2.543698718100647,
+      "learning_rate": 5.321654592438958e-06,
+      "loss": 0.3664,
+      "step": 8696
+    },
+    {
+      "epoch": 1.3308339709257844,
+      "grad_norm": 2.321863025318773,
+      "learning_rate": 5.319464478118723e-06,
+      "loss": 0.3786,
+      "step": 8697
+    },
+    {
+      "epoch": 1.3309869931140015,
+      "grad_norm": 2.1807540364309705,
+      "learning_rate": 5.317274651248495e-06,
+      "loss": 0.3745,
+      "step": 8698
+    },
+    {
+      "epoch": 1.3311400153022188,
+      "grad_norm": 1.9410874369549784,
+      "learning_rate": 5.3150851119627445e-06,
+      "loss": 0.2741,
+      "step": 8699
+    },
+    {
+      "epoch": 1.3312930374904361,
+      "grad_norm": 2.2986422088732272,
+      "learning_rate": 5.3128958603959415e-06,
+      "loss": 0.3781,
+      "step": 8700
+    },
+    {
+      "epoch": 1.3314460596786535,
+      "grad_norm": 1.8028486611526706,
+      "learning_rate": 5.310706896682547e-06,
+      "loss": 0.3409,
+      "step": 8701
+    },
+    {
+      "epoch": 1.3315990818668708,
+      "grad_norm": 1.9787947600298408,
+      "learning_rate": 5.308518220956983e-06,
+      "loss": 0.2612,
+      "step": 8702
+    },
+    {
+      "epoch": 1.331752104055088,
+      "grad_norm": 1.9880430933707562,
+      "learning_rate": 5.306329833353664e-06,
+      "loss": 0.3417,
+      "step": 8703
+    },
+    {
+      "epoch": 1.3319051262433053,
+      "grad_norm": 2.2725659984139157,
+      "learning_rate": 5.3041417340070005e-06,
+      "loss": 0.3574,
+      "step": 8704
+    },
+    {
+      "epoch": 1.3320581484315226,
+      "grad_norm": 2.1827794106466465,
+      "learning_rate": 5.301953923051354e-06,
+      "loss": 0.4218,
+      "step": 8705
+    },
+    {
+      "epoch": 1.33221117061974,
+      "grad_norm": 2.2996389614314876,
+      "learning_rate": 5.2997664006211e-06,
+      "loss": 0.3736,
+      "step": 8706
+    },
+    {
+      "epoch": 1.3323641928079573,
+      "grad_norm": 2.250697167924539,
+      "learning_rate": 5.297579166850584e-06,
+      "loss": 0.3826,
+      "step": 8707
+    },
+    {
+      "epoch": 1.3325172149961744,
+      "grad_norm": 2.0961534983528147,
+      "learning_rate": 5.295392221874118e-06,
+      "loss": 0.4577,
+      "step": 8708
+    },
+    {
+      "epoch": 1.3326702371843917,
+      "grad_norm": 2.178073703599122,
+      "learning_rate": 5.293205565826025e-06,
+      "loss": 0.3955,
+      "step": 8709
+    },
+    {
+      "epoch": 1.332823259372609,
+      "grad_norm": 2.582183841226845,
+      "learning_rate": 5.2910191988405966e-06,
+      "loss": 0.429,
+      "step": 8710
+    },
+    {
+      "epoch": 1.3329762815608264,
+      "grad_norm": 2.6280302076259265,
+      "learning_rate": 5.28883312105209e-06,
+      "loss": 0.4491,
+      "step": 8711
+    },
+    {
+      "epoch": 1.3331293037490437,
+      "grad_norm": 2.4395766994785677,
+      "learning_rate": 5.286647332594779e-06,
+      "loss": 0.4561,
+      "step": 8712
+    },
+    {
+      "epoch": 1.3332823259372608,
+      "grad_norm": 2.257750469790304,
+      "learning_rate": 5.284461833602892e-06,
+      "loss": 0.3722,
+      "step": 8713
+    },
+    {
+      "epoch": 1.3334353481254781,
+      "grad_norm": 2.4393338162992078,
+      "learning_rate": 5.2822766242106505e-06,
+      "loss": 0.3524,
+      "step": 8714
+    },
+    {
+      "epoch": 1.3335883703136955,
+      "grad_norm": 2.0470819190945777,
+      "learning_rate": 5.280091704552257e-06,
+      "loss": 0.3516,
+      "step": 8715
+    },
+    {
+      "epoch": 1.3337413925019128,
+      "grad_norm": 1.8879596223256987,
+      "learning_rate": 5.2779070747618985e-06,
+      "loss": 0.3157,
+      "step": 8716
+    },
+    {
+      "epoch": 1.3338944146901301,
+      "grad_norm": 2.1578376231813072,
+      "learning_rate": 5.275722734973739e-06,
+      "loss": 0.3593,
+      "step": 8717
+    },
+    {
+      "epoch": 1.3340474368783473,
+      "grad_norm": 2.3849493596618565,
+      "learning_rate": 5.273538685321926e-06,
+      "loss": 0.3537,
+      "step": 8718
+    },
+    {
+      "epoch": 1.3342004590665646,
+      "grad_norm": 2.0641129751115175,
+      "learning_rate": 5.2713549259405936e-06,
+      "loss": 0.3433,
+      "step": 8719
+    },
+    {
+      "epoch": 1.334353481254782,
+      "grad_norm": 2.0207189663188747,
+      "learning_rate": 5.269171456963853e-06,
+      "loss": 0.504,
+      "step": 8720
+    },
+    {
+      "epoch": 1.3345065034429993,
+      "grad_norm": 1.9309222928308578,
+      "learning_rate": 5.266988278525802e-06,
+      "loss": 0.3243,
+      "step": 8721
+    },
+    {
+      "epoch": 1.3346595256312166,
+      "grad_norm": 2.417786904966103,
+      "learning_rate": 5.264805390760515e-06,
+      "loss": 0.3591,
+      "step": 8722
+    },
+    {
+      "epoch": 1.3348125478194337,
+      "grad_norm": 2.0559938066081114,
+      "learning_rate": 5.262622793802055e-06,
+      "loss": 0.325,
+      "step": 8723
+    },
+    {
+      "epoch": 1.334965570007651,
+      "grad_norm": 2.154604476772168,
+      "learning_rate": 5.26044048778446e-06,
+      "loss": 0.3335,
+      "step": 8724
+    },
+    {
+      "epoch": 1.3351185921958684,
+      "grad_norm": 2.0975545420310717,
+      "learning_rate": 5.258258472841755e-06,
+      "loss": 0.4099,
+      "step": 8725
+    },
+    {
+      "epoch": 1.3352716143840857,
+      "grad_norm": 1.898966736601991,
+      "learning_rate": 5.256076749107946e-06,
+      "loss": 0.3336,
+      "step": 8726
+    },
+    {
+      "epoch": 1.335424636572303,
+      "grad_norm": 2.126852100718994,
+      "learning_rate": 5.253895316717028e-06,
+      "loss": 0.3528,
+      "step": 8727
+    },
+    {
+      "epoch": 1.3355776587605201,
+      "grad_norm": 2.0082127986354785,
+      "learning_rate": 5.251714175802962e-06,
+      "loss": 0.349,
+      "step": 8728
+    },
+    {
+      "epoch": 1.3357306809487377,
+      "grad_norm": 2.424781858285472,
+      "learning_rate": 5.249533326499701e-06,
+      "loss": 0.3659,
+      "step": 8729
+    },
+    {
+      "epoch": 1.3358837031369548,
+      "grad_norm": 2.3398283834939924,
+      "learning_rate": 5.247352768941189e-06,
+      "loss": 0.3618,
+      "step": 8730
+    },
+    {
+      "epoch": 1.3360367253251721,
+      "grad_norm": 2.0472105115040256,
+      "learning_rate": 5.245172503261331e-06,
+      "loss": 0.3646,
+      "step": 8731
+    },
+    {
+      "epoch": 1.3361897475133895,
+      "grad_norm": 2.223173368949995,
+      "learning_rate": 5.242992529594028e-06,
+      "loss": 0.3406,
+      "step": 8732
+    },
+    {
+      "epoch": 1.3363427697016068,
+      "grad_norm": 2.12524462674687,
+      "learning_rate": 5.240812848073172e-06,
+      "loss": 0.3169,
+      "step": 8733
+    },
+    {
+      "epoch": 1.3364957918898241,
+      "grad_norm": 2.285717102884861,
+      "learning_rate": 5.238633458832607e-06,
+      "loss": 0.3333,
+      "step": 8734
+    },
+    {
+      "epoch": 1.3366488140780413,
+      "grad_norm": 2.1594353097887504,
+      "learning_rate": 5.2364543620061935e-06,
+      "loss": 0.3743,
+      "step": 8735
+    },
+    {
+      "epoch": 1.3368018362662586,
+      "grad_norm": 2.1247874494512238,
+      "learning_rate": 5.234275557727756e-06,
+      "loss": 0.407,
+      "step": 8736
+    },
+    {
+      "epoch": 1.336954858454476,
+      "grad_norm": 2.2331924409935633,
+      "learning_rate": 5.232097046131093e-06,
+      "loss": 0.3355,
+      "step": 8737
+    },
+    {
+      "epoch": 1.3371078806426933,
+      "grad_norm": 2.035105727000153,
+      "learning_rate": 5.2299188273500045e-06,
+      "loss": 0.3884,
+      "step": 8738
+    },
+    {
+      "epoch": 1.3372609028309106,
+      "grad_norm": 2.419555581257721,
+      "learning_rate": 5.227740901518268e-06,
+      "loss": 0.3536,
+      "step": 8739
+    },
+    {
+      "epoch": 1.3374139250191277,
+      "grad_norm": 2.3728108043849385,
+      "learning_rate": 5.225563268769622e-06,
+      "loss": 0.3485,
+      "step": 8740
+    },
+    {
+      "epoch": 1.337566947207345,
+      "grad_norm": 1.9635099062925558,
+      "learning_rate": 5.223385929237819e-06,
+      "loss": 0.3223,
+      "step": 8741
+    },
+    {
+      "epoch": 1.3377199693955624,
+      "grad_norm": 2.0871070571533488,
+      "learning_rate": 5.221208883056571e-06,
+      "loss": 0.3663,
+      "step": 8742
+    },
+    {
+      "epoch": 1.3378729915837797,
+      "grad_norm": 2.1517294117700234,
+      "learning_rate": 5.2190321303595805e-06,
+      "loss": 0.3299,
+      "step": 8743
+    },
+    {
+      "epoch": 1.338026013771997,
+      "grad_norm": 2.2349571562423094,
+      "learning_rate": 5.2168556712805295e-06,
+      "loss": 0.3746,
+      "step": 8744
+    },
+    {
+      "epoch": 1.3381790359602141,
+      "grad_norm": 2.1111718913928983,
+      "learning_rate": 5.214679505953084e-06,
+      "loss": 0.355,
+      "step": 8745
+    },
+    {
+      "epoch": 1.3383320581484315,
+      "grad_norm": 2.071140771626689,
+      "learning_rate": 5.212503634510891e-06,
+      "loss": 0.319,
+      "step": 8746
+    },
+    {
+      "epoch": 1.3384850803366488,
+      "grad_norm": 2.1976321183634453,
+      "learning_rate": 5.210328057087579e-06,
+      "loss": 0.327,
+      "step": 8747
+    },
+    {
+      "epoch": 1.3386381025248661,
+      "grad_norm": 2.3923131141546423,
+      "learning_rate": 5.208152773816757e-06,
+      "loss": 0.4028,
+      "step": 8748
+    },
+    {
+      "epoch": 1.3387911247130835,
+      "grad_norm": 2.283782792526242,
+      "learning_rate": 5.205977784832019e-06,
+      "loss": 0.3377,
+      "step": 8749
+    },
+    {
+      "epoch": 1.3389441469013006,
+      "grad_norm": 2.101349490871268,
+      "learning_rate": 5.203803090266941e-06,
+      "loss": 0.281,
+      "step": 8750
+    },
+    {
+      "epoch": 1.339097169089518,
+      "grad_norm": 2.0532807907074253,
+      "learning_rate": 5.201628690255076e-06,
+      "loss": 0.3238,
+      "step": 8751
+    },
+    {
+      "epoch": 1.3392501912777353,
+      "grad_norm": 2.362591506710098,
+      "learning_rate": 5.199454584929966e-06,
+      "loss": 0.4185,
+      "step": 8752
+    },
+    {
+      "epoch": 1.3394032134659526,
+      "grad_norm": 1.8928761243937846,
+      "learning_rate": 5.197280774425129e-06,
+      "loss": 0.3364,
+      "step": 8753
+    },
+    {
+      "epoch": 1.33955623565417,
+      "grad_norm": 2.198285759440423,
+      "learning_rate": 5.195107258874068e-06,
+      "loss": 0.3444,
+      "step": 8754
+    },
+    {
+      "epoch": 1.339709257842387,
+      "grad_norm": 2.189265677914583,
+      "learning_rate": 5.1929340384102645e-06,
+      "loss": 0.3567,
+      "step": 8755
+    },
+    {
+      "epoch": 1.3398622800306044,
+      "grad_norm": 1.9479819167709145,
+      "learning_rate": 5.1907611131671935e-06,
+      "loss": 0.3587,
+      "step": 8756
+    },
+    {
+      "epoch": 1.3400153022188217,
+      "grad_norm": 2.002771514837011,
+      "learning_rate": 5.188588483278293e-06,
+      "loss": 0.3138,
+      "step": 8757
+    },
+    {
+      "epoch": 1.340168324407039,
+      "grad_norm": 2.510943872707632,
+      "learning_rate": 5.186416148876991e-06,
+      "loss": 0.3849,
+      "step": 8758
+    },
+    {
+      "epoch": 1.3403213465952564,
+      "grad_norm": 1.9571166172475,
+      "learning_rate": 5.184244110096715e-06,
+      "loss": 0.3232,
+      "step": 8759
+    },
+    {
+      "epoch": 1.3404743687834735,
+      "grad_norm": 1.9050058458205525,
+      "learning_rate": 5.182072367070841e-06,
+      "loss": 0.3442,
+      "step": 8760
+    },
+    {
+      "epoch": 1.3406273909716908,
+      "grad_norm": 2.4189439891658333,
+      "learning_rate": 5.1799009199327465e-06,
+      "loss": 0.4003,
+      "step": 8761
+    },
+    {
+      "epoch": 1.3407804131599081,
+      "grad_norm": 2.0763029109194515,
+      "learning_rate": 5.177729768815802e-06,
+      "loss": 0.3211,
+      "step": 8762
+    },
+    {
+      "epoch": 1.3409334353481255,
+      "grad_norm": 2.116101575691697,
+      "learning_rate": 5.175558913853329e-06,
+      "loss": 0.3508,
+      "step": 8763
+    },
+    {
+      "epoch": 1.3410864575363428,
+      "grad_norm": 1.7887941597743418,
+      "learning_rate": 5.17338835517866e-06,
+      "loss": 0.3165,
+      "step": 8764
+    },
+    {
+      "epoch": 1.34123947972456,
+      "grad_norm": 2.2457710693606088,
+      "learning_rate": 5.171218092925099e-06,
+      "loss": 0.3803,
+      "step": 8765
+    },
+    {
+      "epoch": 1.3413925019127775,
+      "grad_norm": 2.0683172511693626,
+      "learning_rate": 5.169048127225915e-06,
+      "loss": 0.3428,
+      "step": 8766
+    },
+    {
+      "epoch": 1.3415455241009946,
+      "grad_norm": 2.1635182390813905,
+      "learning_rate": 5.166878458214391e-06,
+      "loss": 0.3602,
+      "step": 8767
+    },
+    {
+      "epoch": 1.341698546289212,
+      "grad_norm": 2.050582046768065,
+      "learning_rate": 5.16470908602377e-06,
+      "loss": 0.3622,
+      "step": 8768
+    },
+    {
+      "epoch": 1.3418515684774293,
+      "grad_norm": 2.026326185696356,
+      "learning_rate": 5.162540010787273e-06,
+      "loss": 0.3138,
+      "step": 8769
+    },
+    {
+      "epoch": 1.3420045906656466,
+      "grad_norm": 2.0367330156221497,
+      "learning_rate": 5.160371232638122e-06,
+      "loss": 0.303,
+      "step": 8770
+    },
+    {
+      "epoch": 1.342157612853864,
+      "grad_norm": 2.1169309129042215,
+      "learning_rate": 5.1582027517095065e-06,
+      "loss": 0.3154,
+      "step": 8771
+    },
+    {
+      "epoch": 1.342310635042081,
+      "grad_norm": 2.2530862240372493,
+      "learning_rate": 5.156034568134601e-06,
+      "loss": 0.3594,
+      "step": 8772
+    },
+    {
+      "epoch": 1.3424636572302984,
+      "grad_norm": 2.040678051046319,
+      "learning_rate": 5.153866682046562e-06,
+      "loss": 0.3179,
+      "step": 8773
+    },
+    {
+      "epoch": 1.3426166794185157,
+      "grad_norm": 1.9484942018819191,
+      "learning_rate": 5.151699093578528e-06,
+      "loss": 0.348,
+      "step": 8774
+    },
+    {
+      "epoch": 1.342769701606733,
+      "grad_norm": 2.49836830449986,
+      "learning_rate": 5.149531802863621e-06,
+      "loss": 0.4379,
+      "step": 8775
+    },
+    {
+      "epoch": 1.3429227237949504,
+      "grad_norm": 2.272462907141933,
+      "learning_rate": 5.14736481003494e-06,
+      "loss": 0.3124,
+      "step": 8776
+    },
+    {
+      "epoch": 1.3430757459831675,
+      "grad_norm": 2.0838929898866057,
+      "learning_rate": 5.1451981152255695e-06,
+      "loss": 0.3627,
+      "step": 8777
+    },
+    {
+      "epoch": 1.3432287681713848,
+      "grad_norm": 2.0977369640414794,
+      "learning_rate": 5.143031718568575e-06,
+      "loss": 0.463,
+      "step": 8778
+    },
+    {
+      "epoch": 1.3433817903596021,
+      "grad_norm": 2.074263154199314,
+      "learning_rate": 5.140865620197e-06,
+      "loss": 0.3553,
+      "step": 8779
+    },
+    {
+      "epoch": 1.3435348125478195,
+      "grad_norm": 2.3498251424537955,
+      "learning_rate": 5.138699820243882e-06,
+      "loss": 0.4366,
+      "step": 8780
+    },
+    {
+      "epoch": 1.3436878347360368,
+      "grad_norm": 2.132767275315044,
+      "learning_rate": 5.1365343188422235e-06,
+      "loss": 0.3064,
+      "step": 8781
+    },
+    {
+      "epoch": 1.343840856924254,
+      "grad_norm": 2.0608549554257265,
+      "learning_rate": 5.134369116125015e-06,
+      "loss": 0.3042,
+      "step": 8782
+    },
+    {
+      "epoch": 1.3439938791124713,
+      "grad_norm": 2.0378925886824732,
+      "learning_rate": 5.13220421222524e-06,
+      "loss": 0.3163,
+      "step": 8783
+    },
+    {
+      "epoch": 1.3441469013006886,
+      "grad_norm": 2.2278406207902672,
+      "learning_rate": 5.130039607275844e-06,
+      "loss": 0.3278,
+      "step": 8784
+    },
+    {
+      "epoch": 1.344299923488906,
+      "grad_norm": 2.4507146833070106,
+      "learning_rate": 5.127875301409764e-06,
+      "loss": 0.3432,
+      "step": 8785
+    },
+    {
+      "epoch": 1.3444529456771233,
+      "grad_norm": 2.048076636441864,
+      "learning_rate": 5.125711294759927e-06,
+      "loss": 0.2924,
+      "step": 8786
+    },
+    {
+      "epoch": 1.3446059678653404,
+      "grad_norm": 2.2006190582319496,
+      "learning_rate": 5.123547587459221e-06,
+      "loss": 0.3902,
+      "step": 8787
+    },
+    {
+      "epoch": 1.3447589900535577,
+      "grad_norm": 2.1200415996383155,
+      "learning_rate": 5.121384179640539e-06,
+      "loss": 0.4043,
+      "step": 8788
+    },
+    {
+      "epoch": 1.344912012241775,
+      "grad_norm": 1.9167752593186782,
+      "learning_rate": 5.119221071436744e-06,
+      "loss": 0.3091,
+      "step": 8789
+    },
+    {
+      "epoch": 1.3450650344299924,
+      "grad_norm": 1.7729997894919598,
+      "learning_rate": 5.117058262980668e-06,
+      "loss": 0.2764,
+      "step": 8790
+    },
+    {
+      "epoch": 1.3452180566182097,
+      "grad_norm": 2.277714016209161,
+      "learning_rate": 5.114895754405151e-06,
+      "loss": 0.3934,
+      "step": 8791
+    },
+    {
+      "epoch": 1.3453710788064268,
+      "grad_norm": 2.04919853608827,
+      "learning_rate": 5.112733545842996e-06,
+      "loss": 0.3216,
+      "step": 8792
+    },
+    {
+      "epoch": 1.3455241009946441,
+      "grad_norm": 2.063088383531425,
+      "learning_rate": 5.110571637426992e-06,
+      "loss": 0.3876,
+      "step": 8793
+    },
+    {
+      "epoch": 1.3456771231828615,
+      "grad_norm": 2.3164550683532608,
+      "learning_rate": 5.108410029289912e-06,
+      "loss": 0.368,
+      "step": 8794
+    },
+    {
+      "epoch": 1.3458301453710788,
+      "grad_norm": 1.9412845305314341,
+      "learning_rate": 5.106248721564509e-06,
+      "loss": 0.3319,
+      "step": 8795
+    },
+    {
+      "epoch": 1.3459831675592961,
+      "grad_norm": 2.0398505288619684,
+      "learning_rate": 5.1040877143835154e-06,
+      "loss": 0.3048,
+      "step": 8796
+    },
+    {
+      "epoch": 1.3461361897475133,
+      "grad_norm": 1.9405074387765255,
+      "learning_rate": 5.1019270078796476e-06,
+      "loss": 0.335,
+      "step": 8797
+    },
+    {
+      "epoch": 1.3462892119357308,
+      "grad_norm": 3.0361304228685215,
+      "learning_rate": 5.099766602185605e-06,
+      "loss": 0.3535,
+      "step": 8798
+    },
+    {
+      "epoch": 1.346442234123948,
+      "grad_norm": 2.0142600613872514,
+      "learning_rate": 5.097606497434064e-06,
+      "loss": 0.3425,
+      "step": 8799
+    },
+    {
+      "epoch": 1.3465952563121653,
+      "grad_norm": 2.250187533982632,
+      "learning_rate": 5.095446693757688e-06,
+      "loss": 0.37,
+      "step": 8800
+    },
+    {
+      "epoch": 1.3467482785003826,
+      "grad_norm": 2.2286570301304756,
+      "learning_rate": 5.093287191289116e-06,
+      "loss": 0.4101,
+      "step": 8801
+    },
+    {
+      "epoch": 1.3469013006886,
+      "grad_norm": 2.0538342836101133,
+      "learning_rate": 5.091127990160973e-06,
+      "loss": 0.3505,
+      "step": 8802
+    },
+    {
+      "epoch": 1.3470543228768173,
+      "grad_norm": 2.2048180157371773,
+      "learning_rate": 5.088969090505864e-06,
+      "loss": 0.3738,
+      "step": 8803
+    },
+    {
+      "epoch": 1.3472073450650344,
+      "grad_norm": 1.901363181601386,
+      "learning_rate": 5.086810492456375e-06,
+      "loss": 0.294,
+      "step": 8804
+    },
+    {
+      "epoch": 1.3473603672532517,
+      "grad_norm": 2.07977293833161,
+      "learning_rate": 5.084652196145074e-06,
+      "loss": 0.3492,
+      "step": 8805
+    },
+    {
+      "epoch": 1.347513389441469,
+      "grad_norm": 2.2163815188002967,
+      "learning_rate": 5.082494201704511e-06,
+      "loss": 0.2793,
+      "step": 8806
+    },
+    {
+      "epoch": 1.3476664116296864,
+      "grad_norm": 2.412112333946498,
+      "learning_rate": 5.0803365092672165e-06,
+      "loss": 0.3754,
+      "step": 8807
+    },
+    {
+      "epoch": 1.3478194338179037,
+      "grad_norm": 1.7349318109113925,
+      "learning_rate": 5.078179118965698e-06,
+      "loss": 0.2788,
+      "step": 8808
+    },
+    {
+      "epoch": 1.3479724560061208,
+      "grad_norm": 1.9627355364540608,
+      "learning_rate": 5.076022030932461e-06,
+      "loss": 0.3577,
+      "step": 8809
+    },
+    {
+      "epoch": 1.3481254781943381,
+      "grad_norm": 1.7114120202932948,
+      "learning_rate": 5.0738652452999715e-06,
+      "loss": 0.3282,
+      "step": 8810
+    },
+    {
+      "epoch": 1.3482785003825555,
+      "grad_norm": 2.2111716488185675,
+      "learning_rate": 5.0717087622006844e-06,
+      "loss": 0.2957,
+      "step": 8811
+    },
+    {
+      "epoch": 1.3484315225707728,
+      "grad_norm": 1.9715870922408556,
+      "learning_rate": 5.06955258176705e-06,
+      "loss": 0.3563,
+      "step": 8812
+    },
+    {
+      "epoch": 1.3485845447589901,
+      "grad_norm": 2.060344670247845,
+      "learning_rate": 5.0673967041314755e-06,
+      "loss": 0.2971,
+      "step": 8813
+    },
+    {
+      "epoch": 1.3487375669472073,
+      "grad_norm": 2.13181718352457,
+      "learning_rate": 5.0652411294263615e-06,
+      "loss": 0.3509,
+      "step": 8814
+    },
+    {
+      "epoch": 1.3488905891354246,
+      "grad_norm": 2.3479993718903294,
+      "learning_rate": 5.063085857784102e-06,
+      "loss": 0.3717,
+      "step": 8815
+    },
+    {
+      "epoch": 1.349043611323642,
+      "grad_norm": 2.4033272572666613,
+      "learning_rate": 5.060930889337047e-06,
+      "loss": 0.3488,
+      "step": 8816
+    },
+    {
+      "epoch": 1.3491966335118593,
+      "grad_norm": 2.062225137838282,
+      "learning_rate": 5.05877622421755e-06,
+      "loss": 0.3394,
+      "step": 8817
+    },
+    {
+      "epoch": 1.3493496557000766,
+      "grad_norm": 2.1399741737163778,
+      "learning_rate": 5.05662186255794e-06,
+      "loss": 0.346,
+      "step": 8818
+    },
+    {
+      "epoch": 1.3495026778882937,
+      "grad_norm": 2.044146746709054,
+      "learning_rate": 5.054467804490513e-06,
+      "loss": 0.3237,
+      "step": 8819
+    },
+    {
+      "epoch": 1.349655700076511,
+      "grad_norm": 1.9006878517739,
+      "learning_rate": 5.052314050147567e-06,
+      "loss": 0.3543,
+      "step": 8820
+    },
+    {
+      "epoch": 1.3498087222647284,
+      "grad_norm": 1.928889322164962,
+      "learning_rate": 5.050160599661374e-06,
+      "loss": 0.349,
+      "step": 8821
+    },
+    {
+      "epoch": 1.3499617444529457,
+      "grad_norm": 2.402232435601587,
+      "learning_rate": 5.048007453164178e-06,
+      "loss": 0.3668,
+      "step": 8822
+    },
+    {
+      "epoch": 1.350114766641163,
+      "grad_norm": 2.4871412589958264,
+      "learning_rate": 5.04585461078822e-06,
+      "loss": 0.3781,
+      "step": 8823
+    },
+    {
+      "epoch": 1.3502677888293801,
+      "grad_norm": 2.106796208369143,
+      "learning_rate": 5.043702072665711e-06,
+      "loss": 0.3279,
+      "step": 8824
+    },
+    {
+      "epoch": 1.3504208110175975,
+      "grad_norm": 2.365383560450284,
+      "learning_rate": 5.041549838928845e-06,
+      "loss": 0.4045,
+      "step": 8825
+    },
+    {
+      "epoch": 1.3505738332058148,
+      "grad_norm": 2.451762479109529,
+      "learning_rate": 5.039397909709802e-06,
+      "loss": 0.3966,
+      "step": 8826
+    },
+    {
+      "epoch": 1.3507268553940321,
+      "grad_norm": 1.8946555567238423,
+      "learning_rate": 5.037246285140739e-06,
+      "loss": 0.2784,
+      "step": 8827
+    },
+    {
+      "epoch": 1.3508798775822495,
+      "grad_norm": 2.370556858656519,
+      "learning_rate": 5.035094965353796e-06,
+      "loss": 0.3991,
+      "step": 8828
+    },
+    {
+      "epoch": 1.3510328997704666,
+      "grad_norm": 1.9665823637058073,
+      "learning_rate": 5.032943950481094e-06,
+      "loss": 0.3155,
+      "step": 8829
+    },
+    {
+      "epoch": 1.3511859219586841,
+      "grad_norm": 2.1502421747517393,
+      "learning_rate": 5.0307932406547355e-06,
+      "loss": 0.3399,
+      "step": 8830
+    },
+    {
+      "epoch": 1.3513389441469013,
+      "grad_norm": 1.9684009046771507,
+      "learning_rate": 5.028642836006803e-06,
+      "loss": 0.305,
+      "step": 8831
+    },
+    {
+      "epoch": 1.3514919663351186,
+      "grad_norm": 2.3712114531183466,
+      "learning_rate": 5.0264927366693635e-06,
+      "loss": 0.3648,
+      "step": 8832
+    },
+    {
+      "epoch": 1.351644988523336,
+      "grad_norm": 2.1471009526278237,
+      "learning_rate": 5.0243429427744605e-06,
+      "loss": 0.3872,
+      "step": 8833
+    },
+    {
+      "epoch": 1.3517980107115533,
+      "grad_norm": 1.8739089991135973,
+      "learning_rate": 5.0221934544541225e-06,
+      "loss": 0.2828,
+      "step": 8834
+    },
+    {
+      "epoch": 1.3519510328997706,
+      "grad_norm": 2.4082764083999817,
+      "learning_rate": 5.020044271840358e-06,
+      "loss": 0.3817,
+      "step": 8835
+    },
+    {
+      "epoch": 1.3521040550879877,
+      "grad_norm": 2.155006057515205,
+      "learning_rate": 5.017895395065156e-06,
+      "loss": 0.2847,
+      "step": 8836
+    },
+    {
+      "epoch": 1.352257077276205,
+      "grad_norm": 2.1032880782979957,
+      "learning_rate": 5.0157468242604835e-06,
+      "loss": 0.3277,
+      "step": 8837
+    },
+    {
+      "epoch": 1.3524100994644224,
+      "grad_norm": 2.0750944376718614,
+      "learning_rate": 5.013598559558306e-06,
+      "loss": 0.3113,
+      "step": 8838
+    },
+    {
+      "epoch": 1.3525631216526397,
+      "grad_norm": 2.1837175276655416,
+      "learning_rate": 5.011450601090544e-06,
+      "loss": 0.3165,
+      "step": 8839
+    },
+    {
+      "epoch": 1.352716143840857,
+      "grad_norm": 2.101554636022105,
+      "learning_rate": 5.009302948989111e-06,
+      "loss": 0.3534,
+      "step": 8840
+    },
+    {
+      "epoch": 1.3528691660290741,
+      "grad_norm": 2.259872323501473,
+      "learning_rate": 5.007155603385916e-06,
+      "loss": 0.3817,
+      "step": 8841
+    },
+    {
+      "epoch": 1.3530221882172915,
+      "grad_norm": 2.1371804613305536,
+      "learning_rate": 5.005008564412823e-06,
+      "loss": 0.366,
+      "step": 8842
+    },
+    {
+      "epoch": 1.3531752104055088,
+      "grad_norm": 1.783351213021392,
+      "learning_rate": 5.002861832201691e-06,
+      "loss": 0.2375,
+      "step": 8843
+    },
+    {
+      "epoch": 1.3533282325937261,
+      "grad_norm": 2.4210398863460836,
+      "learning_rate": 5.0007154068843714e-06,
+      "loss": 0.4451,
+      "step": 8844
+    },
+    {
+      "epoch": 1.3534812547819435,
+      "grad_norm": 2.109957799601956,
+      "learning_rate": 4.9985692885926675e-06,
+      "loss": 0.3213,
+      "step": 8845
+    },
+    {
+      "epoch": 1.3536342769701606,
+      "grad_norm": 2.201835379953216,
+      "learning_rate": 4.996423477458393e-06,
+      "loss": 0.3522,
+      "step": 8846
+    },
+    {
+      "epoch": 1.353787299158378,
+      "grad_norm": 2.2332666443797913,
+      "learning_rate": 4.994277973613331e-06,
+      "loss": 0.3499,
+      "step": 8847
+    },
+    {
+      "epoch": 1.3539403213465953,
+      "grad_norm": 2.0828069249606247,
+      "learning_rate": 4.992132777189234e-06,
+      "loss": 0.3706,
+      "step": 8848
+    },
+    {
+      "epoch": 1.3540933435348126,
+      "grad_norm": 2.067829402167768,
+      "learning_rate": 4.9899878883178565e-06,
+      "loss": 0.4055,
+      "step": 8849
+    },
+    {
+      "epoch": 1.35424636572303,
+      "grad_norm": 2.234378046868571,
+      "learning_rate": 4.987843307130923e-06,
+      "loss": 0.318,
+      "step": 8850
+    },
+    {
+      "epoch": 1.354399387911247,
+      "grad_norm": 2.2611687807693874,
+      "learning_rate": 4.98569903376014e-06,
+      "loss": 0.3289,
+      "step": 8851
+    },
+    {
+      "epoch": 1.3545524100994644,
+      "grad_norm": 2.178487669182021,
+      "learning_rate": 4.983555068337194e-06,
+      "loss": 0.3294,
+      "step": 8852
+    },
+    {
+      "epoch": 1.3547054322876817,
+      "grad_norm": 2.223404581120469,
+      "learning_rate": 4.981411410993756e-06,
+      "loss": 0.3945,
+      "step": 8853
+    },
+    {
+      "epoch": 1.354858454475899,
+      "grad_norm": 1.9874431190206212,
+      "learning_rate": 4.979268061861477e-06,
+      "loss": 0.3015,
+      "step": 8854
+    },
+    {
+      "epoch": 1.3550114766641164,
+      "grad_norm": 2.1072934649507506,
+      "learning_rate": 4.977125021071988e-06,
+      "loss": 0.3378,
+      "step": 8855
+    },
+    {
+      "epoch": 1.3551644988523335,
+      "grad_norm": 2.341148964981265,
+      "learning_rate": 4.9749822887569e-06,
+      "loss": 0.3951,
+      "step": 8856
+    },
+    {
+      "epoch": 1.3553175210405508,
+      "grad_norm": 2.208386708307971,
+      "learning_rate": 4.9728398650478075e-06,
+      "loss": 0.3591,
+      "step": 8857
+    },
+    {
+      "epoch": 1.3554705432287681,
+      "grad_norm": 2.1163674385472375,
+      "learning_rate": 4.970697750076282e-06,
+      "loss": 0.3211,
+      "step": 8858
+    },
+    {
+      "epoch": 1.3556235654169855,
+      "grad_norm": 1.960454860325222,
+      "learning_rate": 4.96855594397389e-06,
+      "loss": 0.304,
+      "step": 8859
+    },
+    {
+      "epoch": 1.3557765876052028,
+      "grad_norm": 2.192707172753196,
+      "learning_rate": 4.9664144468721565e-06,
+      "loss": 0.3023,
+      "step": 8860
+    },
+    {
+      "epoch": 1.35592960979342,
+      "grad_norm": 2.0167491244911315,
+      "learning_rate": 4.9642732589025986e-06,
+      "loss": 0.2747,
+      "step": 8861
+    },
+    {
+      "epoch": 1.3560826319816373,
+      "grad_norm": 2.2142639849416836,
+      "learning_rate": 4.962132380196729e-06,
+      "loss": 0.3612,
+      "step": 8862
+    },
+    {
+      "epoch": 1.3562356541698546,
+      "grad_norm": 2.2930940087859524,
+      "learning_rate": 4.959991810886013e-06,
+      "loss": 0.3922,
+      "step": 8863
+    },
+    {
+      "epoch": 1.356388676358072,
+      "grad_norm": 2.0111407859260892,
+      "learning_rate": 4.957851551101915e-06,
+      "loss": 0.2614,
+      "step": 8864
+    },
+    {
+      "epoch": 1.3565416985462893,
+      "grad_norm": 2.2532932543779105,
+      "learning_rate": 4.955711600975883e-06,
+      "loss": 0.3617,
+      "step": 8865
+    },
+    {
+      "epoch": 1.3566947207345064,
+      "grad_norm": 2.3137796151354393,
+      "learning_rate": 4.95357196063933e-06,
+      "loss": 0.3794,
+      "step": 8866
+    },
+    {
+      "epoch": 1.356847742922724,
+      "grad_norm": 2.261134143687513,
+      "learning_rate": 4.9514326302236655e-06,
+      "loss": 0.3776,
+      "step": 8867
+    },
+    {
+      "epoch": 1.357000765110941,
+      "grad_norm": 2.071670886754076,
+      "learning_rate": 4.94929360986028e-06,
+      "loss": 0.3288,
+      "step": 8868
+    },
+    {
+      "epoch": 1.3571537872991584,
+      "grad_norm": 2.1015927260956824,
+      "learning_rate": 4.947154899680523e-06,
+      "loss": 0.3423,
+      "step": 8869
+    },
+    {
+      "epoch": 1.3573068094873757,
+      "grad_norm": 2.140147633865237,
+      "learning_rate": 4.9450164998157545e-06,
+      "loss": 0.339,
+      "step": 8870
+    },
+    {
+      "epoch": 1.357459831675593,
+      "grad_norm": 2.014679294585402,
+      "learning_rate": 4.9428784103973025e-06,
+      "loss": 0.3075,
+      "step": 8871
+    },
+    {
+      "epoch": 1.3576128538638104,
+      "grad_norm": 2.0835625273825706,
+      "learning_rate": 4.940740631556462e-06,
+      "loss": 0.3462,
+      "step": 8872
+    },
+    {
+      "epoch": 1.3577658760520275,
+      "grad_norm": 2.182089133357909,
+      "learning_rate": 4.9386031634245365e-06,
+      "loss": 0.3325,
+      "step": 8873
+    },
+    {
+      "epoch": 1.3579188982402448,
+      "grad_norm": 2.0004273732212234,
+      "learning_rate": 4.936466006132791e-06,
+      "loss": 0.3194,
+      "step": 8874
+    },
+    {
+      "epoch": 1.3580719204284621,
+      "grad_norm": 2.1668334947043952,
+      "learning_rate": 4.934329159812476e-06,
+      "loss": 0.3887,
+      "step": 8875
+    },
+    {
+      "epoch": 1.3582249426166795,
+      "grad_norm": 2.2245033231188946,
+      "learning_rate": 4.932192624594824e-06,
+      "loss": 0.3336,
+      "step": 8876
+    },
+    {
+      "epoch": 1.3583779648048968,
+      "grad_norm": 2.4973226364697925,
+      "learning_rate": 4.9300564006110486e-06,
+      "loss": 0.4022,
+      "step": 8877
+    },
+    {
+      "epoch": 1.358530986993114,
+      "grad_norm": 2.308538977320974,
+      "learning_rate": 4.9279204879923425e-06,
+      "loss": 0.3662,
+      "step": 8878
+    },
+    {
+      "epoch": 1.3586840091813313,
+      "grad_norm": 2.2287780866163462,
+      "learning_rate": 4.925784886869883e-06,
+      "loss": 0.354,
+      "step": 8879
+    },
+    {
+      "epoch": 1.3588370313695486,
+      "grad_norm": 2.26226499220998,
+      "learning_rate": 4.923649597374823e-06,
+      "loss": 0.3948,
+      "step": 8880
+    },
+    {
+      "epoch": 1.358990053557766,
+      "grad_norm": 2.397765277999804,
+      "learning_rate": 4.9215146196383e-06,
+      "loss": 0.3911,
+      "step": 8881
+    },
+    {
+      "epoch": 1.3591430757459833,
+      "grad_norm": 2.095255008366767,
+      "learning_rate": 4.919379953791432e-06,
+      "loss": 0.334,
+      "step": 8882
+    },
+    {
+      "epoch": 1.3592960979342004,
+      "grad_norm": 2.3511037115621978,
+      "learning_rate": 4.917245599965317e-06,
+      "loss": 0.3601,
+      "step": 8883
+    },
+    {
+      "epoch": 1.3594491201224177,
+      "grad_norm": 2.20916949291957,
+      "learning_rate": 4.9151115582910324e-06,
+      "loss": 0.357,
+      "step": 8884
+    },
+    {
+      "epoch": 1.359602142310635,
+      "grad_norm": 2.2283270514852997,
+      "learning_rate": 4.912977828899639e-06,
+      "loss": 0.3641,
+      "step": 8885
+    },
+    {
+      "epoch": 1.3597551644988524,
+      "grad_norm": 1.944639349996175,
+      "learning_rate": 4.910844411922179e-06,
+      "loss": 0.4083,
+      "step": 8886
+    },
+    {
+      "epoch": 1.3599081866870697,
+      "grad_norm": 2.0157061926550712,
+      "learning_rate": 4.908711307489671e-06,
+      "loss": 0.3363,
+      "step": 8887
+    },
+    {
+      "epoch": 1.3600612088752868,
+      "grad_norm": 1.9664016887617008,
+      "learning_rate": 4.90657851573312e-06,
+      "loss": 0.3352,
+      "step": 8888
+    },
+    {
+      "epoch": 1.3602142310635041,
+      "grad_norm": 2.1423036098364703,
+      "learning_rate": 4.904446036783508e-06,
+      "loss": 0.3169,
+      "step": 8889
+    },
+    {
+      "epoch": 1.3603672532517215,
+      "grad_norm": 2.1307687150076404,
+      "learning_rate": 4.902313870771795e-06,
+      "loss": 0.3136,
+      "step": 8890
+    },
+    {
+      "epoch": 1.3605202754399388,
+      "grad_norm": 2.037116383219786,
+      "learning_rate": 4.9001820178289365e-06,
+      "loss": 0.3409,
+      "step": 8891
+    },
+    {
+      "epoch": 1.3606732976281561,
+      "grad_norm": 2.04431583800202,
+      "learning_rate": 4.898050478085846e-06,
+      "loss": 0.3112,
+      "step": 8892
+    },
+    {
+      "epoch": 1.3608263198163733,
+      "grad_norm": 2.1739991034642325,
+      "learning_rate": 4.895919251673432e-06,
+      "loss": 0.339,
+      "step": 8893
+    },
+    {
+      "epoch": 1.3609793420045906,
+      "grad_norm": 1.9316321394254719,
+      "learning_rate": 4.893788338722591e-06,
+      "loss": 0.3547,
+      "step": 8894
+    },
+    {
+      "epoch": 1.361132364192808,
+      "grad_norm": 1.8223272300386113,
+      "learning_rate": 4.891657739364177e-06,
+      "loss": 0.3305,
+      "step": 8895
+    },
+    {
+      "epoch": 1.3612853863810253,
+      "grad_norm": 1.873250474892585,
+      "learning_rate": 4.88952745372905e-06,
+      "loss": 0.3082,
+      "step": 8896
+    },
+    {
+      "epoch": 1.3614384085692426,
+      "grad_norm": 2.205921503536888,
+      "learning_rate": 4.887397481948036e-06,
+      "loss": 0.3402,
+      "step": 8897
+    },
+    {
+      "epoch": 1.3615914307574597,
+      "grad_norm": 2.5016859094591055,
+      "learning_rate": 4.885267824151937e-06,
+      "loss": 0.3385,
+      "step": 8898
+    },
+    {
+      "epoch": 1.3617444529456773,
+      "grad_norm": 2.085347385042572,
+      "learning_rate": 4.883138480471553e-06,
+      "loss": 0.3988,
+      "step": 8899
+    },
+    {
+      "epoch": 1.3618974751338944,
+      "grad_norm": 2.2051244044589557,
+      "learning_rate": 4.881009451037656e-06,
+      "loss": 0.3533,
+      "step": 8900
+    },
+    {
+      "epoch": 1.3620504973221117,
+      "grad_norm": 2.1863458371965683,
+      "learning_rate": 4.878880735980986e-06,
+      "loss": 0.3619,
+      "step": 8901
+    },
+    {
+      "epoch": 1.362203519510329,
+      "grad_norm": 2.271622039133187,
+      "learning_rate": 4.876752335432288e-06,
+      "loss": 0.3355,
+      "step": 8902
+    },
+    {
+      "epoch": 1.3623565416985464,
+      "grad_norm": 2.132277396146395,
+      "learning_rate": 4.874624249522273e-06,
+      "loss": 0.3664,
+      "step": 8903
+    },
+    {
+      "epoch": 1.3625095638867637,
+      "grad_norm": 1.9125739210979713,
+      "learning_rate": 4.872496478381632e-06,
+      "loss": 0.2759,
+      "step": 8904
+    },
+    {
+      "epoch": 1.3626625860749808,
+      "grad_norm": 2.2241444950777916,
+      "learning_rate": 4.870369022141042e-06,
+      "loss": 0.3585,
+      "step": 8905
+    },
+    {
+      "epoch": 1.3628156082631981,
+      "grad_norm": 2.1451735908210527,
+      "learning_rate": 4.868241880931157e-06,
+      "loss": 0.3512,
+      "step": 8906
+    },
+    {
+      "epoch": 1.3629686304514155,
+      "grad_norm": 1.9142230746245907,
+      "learning_rate": 4.866115054882613e-06,
+      "loss": 0.328,
+      "step": 8907
+    },
+    {
+      "epoch": 1.3631216526396328,
+      "grad_norm": 2.368547829079924,
+      "learning_rate": 4.863988544126028e-06,
+      "loss": 0.3551,
+      "step": 8908
+    },
+    {
+      "epoch": 1.3632746748278501,
+      "grad_norm": 2.04742148952542,
+      "learning_rate": 4.861862348791999e-06,
+      "loss": 0.3481,
+      "step": 8909
+    },
+    {
+      "epoch": 1.3634276970160673,
+      "grad_norm": 2.1096845873170342,
+      "learning_rate": 4.859736469011104e-06,
+      "loss": 0.3453,
+      "step": 8910
+    },
+    {
+      "epoch": 1.3635807192042846,
+      "grad_norm": 1.9345016136926325,
+      "learning_rate": 4.8576109049139e-06,
+      "loss": 0.3191,
+      "step": 8911
+    },
+    {
+      "epoch": 1.363733741392502,
+      "grad_norm": 2.064452971674517,
+      "learning_rate": 4.8554856566309286e-06,
+      "loss": 0.3252,
+      "step": 8912
+    },
+    {
+      "epoch": 1.3638867635807193,
+      "grad_norm": 2.164800719738148,
+      "learning_rate": 4.853360724292707e-06,
+      "loss": 0.3702,
+      "step": 8913
+    },
+    {
+      "epoch": 1.3640397857689366,
+      "grad_norm": 2.0398684616518272,
+      "learning_rate": 4.851236108029739e-06,
+      "loss": 0.2936,
+      "step": 8914
+    },
+    {
+      "epoch": 1.3641928079571537,
+      "grad_norm": 2.3919411510047968,
+      "learning_rate": 4.849111807972502e-06,
+      "loss": 0.4252,
+      "step": 8915
+    },
+    {
+      "epoch": 1.364345830145371,
+      "grad_norm": 2.317113758644946,
+      "learning_rate": 4.84698782425146e-06,
+      "loss": 0.3417,
+      "step": 8916
+    },
+    {
+      "epoch": 1.3644988523335884,
+      "grad_norm": 1.813125371501577,
+      "learning_rate": 4.844864156997054e-06,
+      "loss": 0.2696,
+      "step": 8917
+    },
+    {
+      "epoch": 1.3646518745218057,
+      "grad_norm": 2.030357912038102,
+      "learning_rate": 4.842740806339709e-06,
+      "loss": 0.3526,
+      "step": 8918
+    },
+    {
+      "epoch": 1.364804896710023,
+      "grad_norm": 1.9203863871860456,
+      "learning_rate": 4.84061777240982e-06,
+      "loss": 0.3105,
+      "step": 8919
+    },
+    {
+      "epoch": 1.3649579188982401,
+      "grad_norm": 2.4389896556711,
+      "learning_rate": 4.8384950553377865e-06,
+      "loss": 0.4105,
+      "step": 8920
+    },
+    {
+      "epoch": 1.3651109410864575,
+      "grad_norm": 2.156091012678082,
+      "learning_rate": 4.8363726552539595e-06,
+      "loss": 0.357,
+      "step": 8921
+    },
+    {
+      "epoch": 1.3652639632746748,
+      "grad_norm": 2.2964416250356208,
+      "learning_rate": 4.8342505722886835e-06,
+      "loss": 0.4063,
+      "step": 8922
+    },
+    {
+      "epoch": 1.3654169854628921,
+      "grad_norm": 2.103168718904348,
+      "learning_rate": 4.832128806572296e-06,
+      "loss": 0.3337,
+      "step": 8923
+    },
+    {
+      "epoch": 1.3655700076511095,
+      "grad_norm": 1.8160059643896378,
+      "learning_rate": 4.8300073582350924e-06,
+      "loss": 0.3028,
+      "step": 8924
+    },
+    {
+      "epoch": 1.3657230298393266,
+      "grad_norm": 2.2057911487268944,
+      "learning_rate": 4.827886227407358e-06,
+      "loss": 0.3627,
+      "step": 8925
+    },
+    {
+      "epoch": 1.365876052027544,
+      "grad_norm": 2.1394622208954064,
+      "learning_rate": 4.825765414219371e-06,
+      "loss": 0.3153,
+      "step": 8926
+    },
+    {
+      "epoch": 1.3660290742157613,
+      "grad_norm": 1.9790746960660583,
+      "learning_rate": 4.823644918801363e-06,
+      "loss": 0.2966,
+      "step": 8927
+    },
+    {
+      "epoch": 1.3661820964039786,
+      "grad_norm": 2.1585848700459462,
+      "learning_rate": 4.821524741283577e-06,
+      "loss": 0.3889,
+      "step": 8928
+    },
+    {
+      "epoch": 1.366335118592196,
+      "grad_norm": 1.9951219879291964,
+      "learning_rate": 4.819404881796217e-06,
+      "loss": 0.3879,
+      "step": 8929
+    },
+    {
+      "epoch": 1.366488140780413,
+      "grad_norm": 2.0365099355278033,
+      "learning_rate": 4.817285340469462e-06,
+      "loss": 0.3067,
+      "step": 8930
+    },
+    {
+      "epoch": 1.3666411629686306,
+      "grad_norm": 2.386301644489069,
+      "learning_rate": 4.815166117433493e-06,
+      "loss": 0.3994,
+      "step": 8931
+    },
+    {
+      "epoch": 1.3667941851568477,
+      "grad_norm": 2.237602958997356,
+      "learning_rate": 4.813047212818457e-06,
+      "loss": 0.4376,
+      "step": 8932
+    },
+    {
+      "epoch": 1.366947207345065,
+      "grad_norm": 2.2080395359941116,
+      "learning_rate": 4.810928626754482e-06,
+      "loss": 0.3744,
+      "step": 8933
+    },
+    {
+      "epoch": 1.3671002295332824,
+      "grad_norm": 2.213416724594083,
+      "learning_rate": 4.808810359371681e-06,
+      "loss": 0.3219,
+      "step": 8934
+    },
+    {
+      "epoch": 1.3672532517214997,
+      "grad_norm": 2.2757376646140717,
+      "learning_rate": 4.806692410800142e-06,
+      "loss": 0.3351,
+      "step": 8935
+    },
+    {
+      "epoch": 1.367406273909717,
+      "grad_norm": 2.275814186271426,
+      "learning_rate": 4.804574781169941e-06,
+      "loss": 0.3565,
+      "step": 8936
+    },
+    {
+      "epoch": 1.3675592960979341,
+      "grad_norm": 2.00049387430229,
+      "learning_rate": 4.802457470611125e-06,
+      "loss": 0.3045,
+      "step": 8937
+    },
+    {
+      "epoch": 1.3677123182861515,
+      "grad_norm": 2.0273766683214838,
+      "learning_rate": 4.800340479253729e-06,
+      "loss": 0.3273,
+      "step": 8938
+    },
+    {
+      "epoch": 1.3678653404743688,
+      "grad_norm": 2.100353583194502,
+      "learning_rate": 4.798223807227767e-06,
+      "loss": 0.3112,
+      "step": 8939
+    },
+    {
+      "epoch": 1.3680183626625861,
+      "grad_norm": 2.173515984420499,
+      "learning_rate": 4.796107454663225e-06,
+      "loss": 0.3939,
+      "step": 8940
+    },
+    {
+      "epoch": 1.3681713848508035,
+      "grad_norm": 2.2645920139454248,
+      "learning_rate": 4.79399142169009e-06,
+      "loss": 0.3357,
+      "step": 8941
+    },
+    {
+      "epoch": 1.3683244070390206,
+      "grad_norm": 2.1304360217955156,
+      "learning_rate": 4.791875708438304e-06,
+      "loss": 0.3591,
+      "step": 8942
+    },
+    {
+      "epoch": 1.368477429227238,
+      "grad_norm": 2.016668890017917,
+      "learning_rate": 4.789760315037801e-06,
+      "loss": 0.3164,
+      "step": 8943
+    },
+    {
+      "epoch": 1.3686304514154553,
+      "grad_norm": 2.102649703832163,
+      "learning_rate": 4.7876452416185075e-06,
+      "loss": 0.3938,
+      "step": 8944
+    },
+    {
+      "epoch": 1.3687834736036726,
+      "grad_norm": 2.3476389450020188,
+      "learning_rate": 4.785530488310307e-06,
+      "loss": 0.4252,
+      "step": 8945
+    },
+    {
+      "epoch": 1.36893649579189,
+      "grad_norm": 1.8083889949743437,
+      "learning_rate": 4.783416055243074e-06,
+      "loss": 0.2638,
+      "step": 8946
+    },
+    {
+      "epoch": 1.369089517980107,
+      "grad_norm": 1.6592216434472842,
+      "learning_rate": 4.781301942546677e-06,
+      "loss": 0.2525,
+      "step": 8947
+    },
+    {
+      "epoch": 1.3692425401683244,
+      "grad_norm": 2.2613655519561515,
+      "learning_rate": 4.779188150350934e-06,
+      "loss": 0.4171,
+      "step": 8948
+    },
+    {
+      "epoch": 1.3693955623565417,
+      "grad_norm": 2.0986112105450214,
+      "learning_rate": 4.777074678785676e-06,
+      "loss": 0.3432,
+      "step": 8949
+    },
+    {
+      "epoch": 1.369548584544759,
+      "grad_norm": 2.16417414807789,
+      "learning_rate": 4.774961527980697e-06,
+      "loss": 0.3598,
+      "step": 8950
+    },
+    {
+      "epoch": 1.3697016067329764,
+      "grad_norm": 2.5301061654646255,
+      "learning_rate": 4.772848698065764e-06,
+      "loss": 0.4318,
+      "step": 8951
+    },
+    {
+      "epoch": 1.3698546289211935,
+      "grad_norm": 1.9819120670022627,
+      "learning_rate": 4.770736189170645e-06,
+      "loss": 0.3328,
+      "step": 8952
+    },
+    {
+      "epoch": 1.3700076511094108,
+      "grad_norm": 2.079685492030615,
+      "learning_rate": 4.768624001425077e-06,
+      "loss": 0.4102,
+      "step": 8953
+    },
+    {
+      "epoch": 1.3701606732976281,
+      "grad_norm": 2.0838078795974835,
+      "learning_rate": 4.766512134958767e-06,
+      "loss": 0.3124,
+      "step": 8954
+    },
+    {
+      "epoch": 1.3703136954858455,
+      "grad_norm": 1.9182397991758027,
+      "learning_rate": 4.764400589901424e-06,
+      "loss": 0.3356,
+      "step": 8955
+    },
+    {
+      "epoch": 1.3704667176740628,
+      "grad_norm": 2.0187379873078153,
+      "learning_rate": 4.762289366382723e-06,
+      "loss": 0.3335,
+      "step": 8956
+    },
+    {
+      "epoch": 1.37061973986228,
+      "grad_norm": 2.392520798804102,
+      "learning_rate": 4.7601784645323225e-06,
+      "loss": 0.4488,
+      "step": 8957
+    },
+    {
+      "epoch": 1.3707727620504973,
+      "grad_norm": 2.2128964626438536,
+      "learning_rate": 4.75806788447986e-06,
+      "loss": 0.3515,
+      "step": 8958
+    },
+    {
+      "epoch": 1.3709257842387146,
+      "grad_norm": 2.1733235228362147,
+      "learning_rate": 4.755957626354954e-06,
+      "loss": 0.3589,
+      "step": 8959
+    },
+    {
+      "epoch": 1.371078806426932,
+      "grad_norm": 2.0313878062789326,
+      "learning_rate": 4.753847690287207e-06,
+      "loss": 0.2957,
+      "step": 8960
+    },
+    {
+      "epoch": 1.3712318286151493,
+      "grad_norm": 2.1544020326434588,
+      "learning_rate": 4.751738076406196e-06,
+      "loss": 0.345,
+      "step": 8961
+    },
+    {
+      "epoch": 1.3713848508033664,
+      "grad_norm": 2.1167948771347382,
+      "learning_rate": 4.74962878484148e-06,
+      "loss": 0.4013,
+      "step": 8962
+    },
+    {
+      "epoch": 1.371537872991584,
+      "grad_norm": 2.2116358842619195,
+      "learning_rate": 4.747519815722601e-06,
+      "loss": 0.347,
+      "step": 8963
+    },
+    {
+      "epoch": 1.371690895179801,
+      "grad_norm": 2.2257642473680743,
+      "learning_rate": 4.7454111691790785e-06,
+      "loss": 0.4065,
+      "step": 8964
+    },
+    {
+      "epoch": 1.3718439173680184,
+      "grad_norm": 2.373332905259864,
+      "learning_rate": 4.743302845340411e-06,
+      "loss": 0.3881,
+      "step": 8965
+    },
+    {
+      "epoch": 1.3719969395562357,
+      "grad_norm": 1.9971503672491702,
+      "learning_rate": 4.74119484433608e-06,
+      "loss": 0.3295,
+      "step": 8966
+    },
+    {
+      "epoch": 1.372149961744453,
+      "grad_norm": 2.164576400339109,
+      "learning_rate": 4.739087166295546e-06,
+      "loss": 0.368,
+      "step": 8967
+    },
+    {
+      "epoch": 1.3723029839326704,
+      "grad_norm": 2.1304626352278735,
+      "learning_rate": 4.73697981134825e-06,
+      "loss": 0.3502,
+      "step": 8968
+    },
+    {
+      "epoch": 1.3724560061208875,
+      "grad_norm": 1.9351491350127181,
+      "learning_rate": 4.734872779623611e-06,
+      "loss": 0.3074,
+      "step": 8969
+    },
+    {
+      "epoch": 1.3726090283091048,
+      "grad_norm": 1.8804547668574492,
+      "learning_rate": 4.732766071251037e-06,
+      "loss": 0.3111,
+      "step": 8970
+    },
+    {
+      "epoch": 1.3727620504973221,
+      "grad_norm": 2.2549524395110825,
+      "learning_rate": 4.730659686359901e-06,
+      "loss": 0.3449,
+      "step": 8971
+    },
+    {
+      "epoch": 1.3729150726855395,
+      "grad_norm": 2.242806963979162,
+      "learning_rate": 4.728553625079564e-06,
+      "loss": 0.3164,
+      "step": 8972
+    },
+    {
+      "epoch": 1.3730680948737568,
+      "grad_norm": 2.11930174434817,
+      "learning_rate": 4.726447887539378e-06,
+      "loss": 0.3364,
+      "step": 8973
+    },
+    {
+      "epoch": 1.373221117061974,
+      "grad_norm": 2.470435497406207,
+      "learning_rate": 4.724342473868655e-06,
+      "loss": 0.4222,
+      "step": 8974
+    },
+    {
+      "epoch": 1.3733741392501913,
+      "grad_norm": 2.3682646592639736,
+      "learning_rate": 4.722237384196694e-06,
+      "loss": 0.3675,
+      "step": 8975
+    },
+    {
+      "epoch": 1.3735271614384086,
+      "grad_norm": 2.3977878266611667,
+      "learning_rate": 4.72013261865279e-06,
+      "loss": 0.3816,
+      "step": 8976
+    },
+    {
+      "epoch": 1.373680183626626,
+      "grad_norm": 1.8686907034229359,
+      "learning_rate": 4.71802817736619e-06,
+      "loss": 0.2997,
+      "step": 8977
+    },
+    {
+      "epoch": 1.3738332058148433,
+      "grad_norm": 2.14486465113667,
+      "learning_rate": 4.715924060466145e-06,
+      "loss": 0.3721,
+      "step": 8978
+    },
+    {
+      "epoch": 1.3739862280030604,
+      "grad_norm": 2.2985453789295627,
+      "learning_rate": 4.713820268081879e-06,
+      "loss": 0.3437,
+      "step": 8979
+    },
+    {
+      "epoch": 1.3741392501912777,
+      "grad_norm": 2.227704431604356,
+      "learning_rate": 4.711716800342584e-06,
+      "loss": 0.3468,
+      "step": 8980
+    },
+    {
+      "epoch": 1.374292272379495,
+      "grad_norm": 2.204147290592917,
+      "learning_rate": 4.709613657377449e-06,
+      "loss": 0.3204,
+      "step": 8981
+    },
+    {
+      "epoch": 1.3744452945677124,
+      "grad_norm": 2.1549027912943712,
+      "learning_rate": 4.707510839315642e-06,
+      "loss": 0.3232,
+      "step": 8982
+    },
+    {
+      "epoch": 1.3745983167559297,
+      "grad_norm": 1.9921345110290458,
+      "learning_rate": 4.705408346286291e-06,
+      "loss": 0.3144,
+      "step": 8983
+    },
+    {
+      "epoch": 1.3747513389441468,
+      "grad_norm": 2.0591857889441956,
+      "learning_rate": 4.703306178418529e-06,
+      "loss": 0.3305,
+      "step": 8984
+    },
+    {
+      "epoch": 1.3749043611323641,
+      "grad_norm": 2.094890599785986,
+      "learning_rate": 4.701204335841455e-06,
+      "loss": 0.2791,
+      "step": 8985
+    },
+    {
+      "epoch": 1.3750573833205815,
+      "grad_norm": 2.1299628438008753,
+      "learning_rate": 4.699102818684151e-06,
+      "loss": 0.2865,
+      "step": 8986
+    },
+    {
+      "epoch": 1.3752104055087988,
+      "grad_norm": 2.378452431658748,
+      "learning_rate": 4.6970016270756826e-06,
+      "loss": 0.4039,
+      "step": 8987
+    },
+    {
+      "epoch": 1.3753634276970161,
+      "grad_norm": 1.941247460390366,
+      "learning_rate": 4.694900761145088e-06,
+      "loss": 0.2924,
+      "step": 8988
+    },
+    {
+      "epoch": 1.3755164498852332,
+      "grad_norm": 2.0735387694712673,
+      "learning_rate": 4.692800221021392e-06,
+      "loss": 0.283,
+      "step": 8989
+    },
+    {
+      "epoch": 1.3756694720734506,
+      "grad_norm": 2.1748933607293783,
+      "learning_rate": 4.690700006833595e-06,
+      "loss": 0.3652,
+      "step": 8990
+    },
+    {
+      "epoch": 1.375822494261668,
+      "grad_norm": 2.2313085481512647,
+      "learning_rate": 4.688600118710682e-06,
+      "loss": 0.3407,
+      "step": 8991
+    },
+    {
+      "epoch": 1.3759755164498852,
+      "grad_norm": 2.125396345210953,
+      "learning_rate": 4.686500556781614e-06,
+      "loss": 0.3484,
+      "step": 8992
+    },
+    {
+      "epoch": 1.3761285386381026,
+      "grad_norm": 2.442831053015856,
+      "learning_rate": 4.684401321175333e-06,
+      "loss": 0.3715,
+      "step": 8993
+    },
+    {
+      "epoch": 1.3762815608263197,
+      "grad_norm": 2.264403792995592,
+      "learning_rate": 4.6823024120207615e-06,
+      "loss": 0.3353,
+      "step": 8994
+    },
+    {
+      "epoch": 1.376434583014537,
+      "grad_norm": 2.153718862608068,
+      "learning_rate": 4.680203829446802e-06,
+      "loss": 0.3349,
+      "step": 8995
+    },
+    {
+      "epoch": 1.3765876052027544,
+      "grad_norm": 2.2678638376517317,
+      "learning_rate": 4.678105573582337e-06,
+      "loss": 0.3347,
+      "step": 8996
+    },
+    {
+      "epoch": 1.3767406273909717,
+      "grad_norm": 2.143262059852469,
+      "learning_rate": 4.676007644556226e-06,
+      "loss": 0.3476,
+      "step": 8997
+    },
+    {
+      "epoch": 1.376893649579189,
+      "grad_norm": 1.8505114454183178,
+      "learning_rate": 4.673910042497312e-06,
+      "loss": 0.3635,
+      "step": 8998
+    },
+    {
+      "epoch": 1.3770466717674061,
+      "grad_norm": 1.9620249089509953,
+      "learning_rate": 4.671812767534424e-06,
+      "loss": 0.2853,
+      "step": 8999
+    },
+    {
+      "epoch": 1.3771996939556237,
+      "grad_norm": 2.2096240025262563,
+      "learning_rate": 4.669715819796354e-06,
+      "loss": 0.3826,
+      "step": 9000
+    },
+    {
+      "epoch": 1.3773527161438408,
+      "grad_norm": 2.0471221789313048,
+      "learning_rate": 4.667619199411886e-06,
+      "loss": 0.3303,
+      "step": 9001
+    },
+    {
+      "epoch": 1.3775057383320581,
+      "grad_norm": 2.219398951091194,
+      "learning_rate": 4.665522906509789e-06,
+      "loss": 0.3383,
+      "step": 9002
+    },
+    {
+      "epoch": 1.3776587605202755,
+      "grad_norm": 2.1239169778671663,
+      "learning_rate": 4.663426941218796e-06,
+      "loss": 0.3643,
+      "step": 9003
+    },
+    {
+      "epoch": 1.3778117827084928,
+      "grad_norm": 2.2161074087918413,
+      "learning_rate": 4.661331303667627e-06,
+      "loss": 0.3544,
+      "step": 9004
+    },
+    {
+      "epoch": 1.3779648048967101,
+      "grad_norm": 1.8727061163511225,
+      "learning_rate": 4.659235993984995e-06,
+      "loss": 0.2673,
+      "step": 9005
+    },
+    {
+      "epoch": 1.3781178270849272,
+      "grad_norm": 1.9336210558087923,
+      "learning_rate": 4.657141012299567e-06,
+      "loss": 0.3544,
+      "step": 9006
+    },
+    {
+      "epoch": 1.3782708492731446,
+      "grad_norm": 2.126956772127913,
+      "learning_rate": 4.655046358740014e-06,
+      "loss": 0.4919,
+      "step": 9007
+    },
+    {
+      "epoch": 1.378423871461362,
+      "grad_norm": 2.0412566209981478,
+      "learning_rate": 4.652952033434978e-06,
+      "loss": 0.3247,
+      "step": 9008
+    },
+    {
+      "epoch": 1.3785768936495792,
+      "grad_norm": 1.9837314117934364,
+      "learning_rate": 4.650858036513067e-06,
+      "loss": 0.3384,
+      "step": 9009
+    },
+    {
+      "epoch": 1.3787299158377966,
+      "grad_norm": 2.0790714883936134,
+      "learning_rate": 4.648764368102894e-06,
+      "loss": 0.2951,
+      "step": 9010
+    },
+    {
+      "epoch": 1.3788829380260137,
+      "grad_norm": 2.0441032696122616,
+      "learning_rate": 4.646671028333039e-06,
+      "loss": 0.358,
+      "step": 9011
+    },
+    {
+      "epoch": 1.379035960214231,
+      "grad_norm": 2.026421631547,
+      "learning_rate": 4.644578017332052e-06,
+      "loss": 0.4144,
+      "step": 9012
+    },
+    {
+      "epoch": 1.3791889824024484,
+      "grad_norm": 2.2121542332323854,
+      "learning_rate": 4.642485335228483e-06,
+      "loss": 0.3838,
+      "step": 9013
+    },
+    {
+      "epoch": 1.3793420045906657,
+      "grad_norm": 2.218639370811378,
+      "learning_rate": 4.640392982150846e-06,
+      "loss": 0.3431,
+      "step": 9014
+    },
+    {
+      "epoch": 1.379495026778883,
+      "grad_norm": 2.037830779490401,
+      "learning_rate": 4.638300958227645e-06,
+      "loss": 0.3522,
+      "step": 9015
+    },
+    {
+      "epoch": 1.3796480489671001,
+      "grad_norm": 2.410335657240026,
+      "learning_rate": 4.636209263587358e-06,
+      "loss": 0.3976,
+      "step": 9016
+    },
+    {
+      "epoch": 1.3798010711553175,
+      "grad_norm": 2.1241097054830167,
+      "learning_rate": 4.634117898358441e-06,
+      "loss": 0.3065,
+      "step": 9017
+    },
+    {
+      "epoch": 1.3799540933435348,
+      "grad_norm": 2.343825620509157,
+      "learning_rate": 4.632026862669337e-06,
+      "loss": 0.3899,
+      "step": 9018
+    },
+    {
+      "epoch": 1.3801071155317521,
+      "grad_norm": 1.9525031937885948,
+      "learning_rate": 4.629936156648463e-06,
+      "loss": 0.2961,
+      "step": 9019
+    },
+    {
+      "epoch": 1.3802601377199695,
+      "grad_norm": 2.134029971911187,
+      "learning_rate": 4.627845780424217e-06,
+      "loss": 0.3464,
+      "step": 9020
+    },
+    {
+      "epoch": 1.3804131599081866,
+      "grad_norm": 1.9946961858753256,
+      "learning_rate": 4.625755734124977e-06,
+      "loss": 0.3302,
+      "step": 9021
+    },
+    {
+      "epoch": 1.380566182096404,
+      "grad_norm": 1.9751054972586108,
+      "learning_rate": 4.623666017879098e-06,
+      "loss": 0.2879,
+      "step": 9022
+    },
+    {
+      "epoch": 1.3807192042846212,
+      "grad_norm": 2.1883779520669115,
+      "learning_rate": 4.62157663181493e-06,
+      "loss": 0.2725,
+      "step": 9023
+    },
+    {
+      "epoch": 1.3808722264728386,
+      "grad_norm": 1.8160688719204534,
+      "learning_rate": 4.619487576060777e-06,
+      "loss": 0.2921,
+      "step": 9024
+    },
+    {
+      "epoch": 1.381025248661056,
+      "grad_norm": 2.04738295569202,
+      "learning_rate": 4.6173988507449366e-06,
+      "loss": 0.3446,
+      "step": 9025
+    },
+    {
+      "epoch": 1.381178270849273,
+      "grad_norm": 1.970933426371676,
+      "learning_rate": 4.615310455995697e-06,
+      "loss": 0.2965,
+      "step": 9026
+    },
+    {
+      "epoch": 1.3813312930374904,
+      "grad_norm": 2.265268014430741,
+      "learning_rate": 4.613222391941304e-06,
+      "loss": 0.2872,
+      "step": 9027
+    },
+    {
+      "epoch": 1.3814843152257077,
+      "grad_norm": 2.139073356126362,
+      "learning_rate": 4.611134658709992e-06,
+      "loss": 0.3425,
+      "step": 9028
+    },
+    {
+      "epoch": 1.381637337413925,
+      "grad_norm": 2.07292711463194,
+      "learning_rate": 4.609047256429992e-06,
+      "loss": 0.3747,
+      "step": 9029
+    },
+    {
+      "epoch": 1.3817903596021424,
+      "grad_norm": 2.268761758492247,
+      "learning_rate": 4.60696018522948e-06,
+      "loss": 0.38,
+      "step": 9030
+    },
+    {
+      "epoch": 1.3819433817903595,
+      "grad_norm": 2.2366077257417785,
+      "learning_rate": 4.604873445236645e-06,
+      "loss": 0.3249,
+      "step": 9031
+    },
+    {
+      "epoch": 1.382096403978577,
+      "grad_norm": 2.013369317159582,
+      "learning_rate": 4.602787036579643e-06,
+      "loss": 0.3385,
+      "step": 9032
+    },
+    {
+      "epoch": 1.3822494261667941,
+      "grad_norm": 2.0345532889136653,
+      "learning_rate": 4.600700959386592e-06,
+      "loss": 0.2601,
+      "step": 9033
+    },
+    {
+      "epoch": 1.3824024483550115,
+      "grad_norm": 2.1497689903647235,
+      "learning_rate": 4.598615213785624e-06,
+      "loss": 0.3804,
+      "step": 9034
+    },
+    {
+      "epoch": 1.3825554705432288,
+      "grad_norm": 2.249995107110444,
+      "learning_rate": 4.596529799904825e-06,
+      "loss": 0.3559,
+      "step": 9035
+    },
+    {
+      "epoch": 1.3827084927314461,
+      "grad_norm": 1.8239006623904237,
+      "learning_rate": 4.594444717872269e-06,
+      "loss": 0.2621,
+      "step": 9036
+    },
+    {
+      "epoch": 1.3828615149196635,
+      "grad_norm": 2.023991229226579,
+      "learning_rate": 4.592359967816012e-06,
+      "loss": 0.3317,
+      "step": 9037
+    },
+    {
+      "epoch": 1.3830145371078806,
+      "grad_norm": 2.0859580119335326,
+      "learning_rate": 4.590275549864081e-06,
+      "loss": 0.3042,
+      "step": 9038
+    },
+    {
+      "epoch": 1.383167559296098,
+      "grad_norm": 2.376028679244968,
+      "learning_rate": 4.5881914641444936e-06,
+      "loss": 0.3887,
+      "step": 9039
+    },
+    {
+      "epoch": 1.3833205814843152,
+      "grad_norm": 2.0554311647854164,
+      "learning_rate": 4.586107710785238e-06,
+      "loss": 0.3451,
+      "step": 9040
+    },
+    {
+      "epoch": 1.3834736036725326,
+      "grad_norm": 1.7523628481913938,
+      "learning_rate": 4.584024289914288e-06,
+      "loss": 0.3022,
+      "step": 9041
+    },
+    {
+      "epoch": 1.38362662586075,
+      "grad_norm": 1.9390495654961886,
+      "learning_rate": 4.581941201659593e-06,
+      "loss": 0.3678,
+      "step": 9042
+    },
+    {
+      "epoch": 1.383779648048967,
+      "grad_norm": 2.237567627656652,
+      "learning_rate": 4.579858446149086e-06,
+      "loss": 0.4361,
+      "step": 9043
+    },
+    {
+      "epoch": 1.3839326702371844,
+      "grad_norm": 2.3021361434486405,
+      "learning_rate": 4.577776023510674e-06,
+      "loss": 0.3286,
+      "step": 9044
+    },
+    {
+      "epoch": 1.3840856924254017,
+      "grad_norm": 2.3763182088117545,
+      "learning_rate": 4.575693933872248e-06,
+      "loss": 0.3913,
+      "step": 9045
+    },
+    {
+      "epoch": 1.384238714613619,
+      "grad_norm": 1.9626988253783606,
+      "learning_rate": 4.573612177361679e-06,
+      "loss": 0.3182,
+      "step": 9046
+    },
+    {
+      "epoch": 1.3843917368018364,
+      "grad_norm": 2.173280085060481,
+      "learning_rate": 4.571530754106813e-06,
+      "loss": 0.3433,
+      "step": 9047
+    },
+    {
+      "epoch": 1.3845447589900535,
+      "grad_norm": 1.942616106562521,
+      "learning_rate": 4.5694496642354815e-06,
+      "loss": 0.2648,
+      "step": 9048
+    },
+    {
+      "epoch": 1.3846977811782708,
+      "grad_norm": 2.1229756870525462,
+      "learning_rate": 4.567368907875489e-06,
+      "loss": 0.3388,
+      "step": 9049
+    },
+    {
+      "epoch": 1.3848508033664881,
+      "grad_norm": 1.9885131487226595,
+      "learning_rate": 4.565288485154624e-06,
+      "loss": 0.3119,
+      "step": 9050
+    },
+    {
+      "epoch": 1.3850038255547055,
+      "grad_norm": 1.9222230959456768,
+      "learning_rate": 4.563208396200651e-06,
+      "loss": 0.2911,
+      "step": 9051
+    },
+    {
+      "epoch": 1.3851568477429228,
+      "grad_norm": 2.158206506225906,
+      "learning_rate": 4.561128641141327e-06,
+      "loss": 0.3206,
+      "step": 9052
+    },
+    {
+      "epoch": 1.38530986993114,
+      "grad_norm": 2.19345897921204,
+      "learning_rate": 4.559049220104366e-06,
+      "loss": 0.3805,
+      "step": 9053
+    },
+    {
+      "epoch": 1.3854628921193572,
+      "grad_norm": 2.1317407824996044,
+      "learning_rate": 4.5569701332174746e-06,
+      "loss": 0.3375,
+      "step": 9054
+    },
+    {
+      "epoch": 1.3856159143075746,
+      "grad_norm": 2.3366870170548233,
+      "learning_rate": 4.554891380608346e-06,
+      "loss": 0.3741,
+      "step": 9055
+    },
+    {
+      "epoch": 1.385768936495792,
+      "grad_norm": 2.2540854792574363,
+      "learning_rate": 4.552812962404637e-06,
+      "loss": 0.3169,
+      "step": 9056
+    },
+    {
+      "epoch": 1.3859219586840092,
+      "grad_norm": 2.099517950545955,
+      "learning_rate": 4.550734878733989e-06,
+      "loss": 0.3271,
+      "step": 9057
+    },
+    {
+      "epoch": 1.3860749808722264,
+      "grad_norm": 1.9069516040024563,
+      "learning_rate": 4.548657129724038e-06,
+      "loss": 0.2895,
+      "step": 9058
+    },
+    {
+      "epoch": 1.3862280030604437,
+      "grad_norm": 2.161070641348649,
+      "learning_rate": 4.54657971550237e-06,
+      "loss": 0.3233,
+      "step": 9059
+    },
+    {
+      "epoch": 1.386381025248661,
+      "grad_norm": 2.216026925034713,
+      "learning_rate": 4.54450263619658e-06,
+      "loss": 0.3162,
+      "step": 9060
+    },
+    {
+      "epoch": 1.3865340474368784,
+      "grad_norm": 2.1840129464952094,
+      "learning_rate": 4.542425891934229e-06,
+      "loss": 0.3682,
+      "step": 9061
+    },
+    {
+      "epoch": 1.3866870696250957,
+      "grad_norm": 1.884220281115278,
+      "learning_rate": 4.540349482842846e-06,
+      "loss": 0.2823,
+      "step": 9062
+    },
+    {
+      "epoch": 1.3868400918133128,
+      "grad_norm": 1.7360218886526524,
+      "learning_rate": 4.538273409049964e-06,
+      "loss": 0.2798,
+      "step": 9063
+    },
+    {
+      "epoch": 1.3869931140015304,
+      "grad_norm": 2.3949601964182103,
+      "learning_rate": 4.536197670683081e-06,
+      "loss": 0.3681,
+      "step": 9064
+    },
+    {
+      "epoch": 1.3871461361897475,
+      "grad_norm": 2.0561106971906646,
+      "learning_rate": 4.534122267869667e-06,
+      "loss": 0.3101,
+      "step": 9065
+    },
+    {
+      "epoch": 1.3872991583779648,
+      "grad_norm": 2.3957391340285383,
+      "learning_rate": 4.532047200737191e-06,
+      "loss": 0.4326,
+      "step": 9066
+    },
+    {
+      "epoch": 1.3874521805661821,
+      "grad_norm": 1.9010857779535084,
+      "learning_rate": 4.529972469413087e-06,
+      "loss": 0.3132,
+      "step": 9067
+    },
+    {
+      "epoch": 1.3876052027543995,
+      "grad_norm": 2.02976579219007,
+      "learning_rate": 4.527898074024772e-06,
+      "loss": 0.358,
+      "step": 9068
+    },
+    {
+      "epoch": 1.3877582249426168,
+      "grad_norm": 2.29210851094873,
+      "learning_rate": 4.525824014699643e-06,
+      "loss": 0.3679,
+      "step": 9069
+    },
+    {
+      "epoch": 1.387911247130834,
+      "grad_norm": 1.9044989375358745,
+      "learning_rate": 4.523750291565078e-06,
+      "loss": 0.3186,
+      "step": 9070
+    },
+    {
+      "epoch": 1.3880642693190512,
+      "grad_norm": 1.9671211186121362,
+      "learning_rate": 4.52167690474843e-06,
+      "loss": 0.313,
+      "step": 9071
+    },
+    {
+      "epoch": 1.3882172915072686,
+      "grad_norm": 2.1843101523101747,
+      "learning_rate": 4.5196038543770335e-06,
+      "loss": 0.3622,
+      "step": 9072
+    },
+    {
+      "epoch": 1.388370313695486,
+      "grad_norm": 1.8997165102498639,
+      "learning_rate": 4.517531140578205e-06,
+      "loss": 0.3651,
+      "step": 9073
+    },
+    {
+      "epoch": 1.3885233358837032,
+      "grad_norm": 2.11857325406712,
+      "learning_rate": 4.515458763479237e-06,
+      "loss": 0.4028,
+      "step": 9074
+    },
+    {
+      "epoch": 1.3886763580719204,
+      "grad_norm": 2.0344891913610037,
+      "learning_rate": 4.5133867232074e-06,
+      "loss": 0.2762,
+      "step": 9075
+    },
+    {
+      "epoch": 1.3888293802601377,
+      "grad_norm": 1.8699787605804676,
+      "learning_rate": 4.51131501988995e-06,
+      "loss": 0.2799,
+      "step": 9076
+    },
+    {
+      "epoch": 1.388982402448355,
+      "grad_norm": 1.9247216761568173,
+      "learning_rate": 4.509243653654116e-06,
+      "loss": 0.231,
+      "step": 9077
+    },
+    {
+      "epoch": 1.3891354246365724,
+      "grad_norm": 1.8389597441884304,
+      "learning_rate": 4.507172624627109e-06,
+      "loss": 0.2892,
+      "step": 9078
+    },
+    {
+      "epoch": 1.3892884468247897,
+      "grad_norm": 2.1670682348965054,
+      "learning_rate": 4.505101932936119e-06,
+      "loss": 0.3743,
+      "step": 9079
+    },
+    {
+      "epoch": 1.3894414690130068,
+      "grad_norm": 2.04514272674738,
+      "learning_rate": 4.5030315787083126e-06,
+      "loss": 0.3327,
+      "step": 9080
+    },
+    {
+      "epoch": 1.3895944912012241,
+      "grad_norm": 2.3185780179520767,
+      "learning_rate": 4.5009615620708466e-06,
+      "loss": 0.4264,
+      "step": 9081
+    },
+    {
+      "epoch": 1.3897475133894415,
+      "grad_norm": 1.994748159454086,
+      "learning_rate": 4.498891883150842e-06,
+      "loss": 0.3528,
+      "step": 9082
+    },
+    {
+      "epoch": 1.3899005355776588,
+      "grad_norm": 2.2945232098348107,
+      "learning_rate": 4.496822542075403e-06,
+      "loss": 0.3135,
+      "step": 9083
+    },
+    {
+      "epoch": 1.3900535577658761,
+      "grad_norm": 2.19865510079339,
+      "learning_rate": 4.494753538971627e-06,
+      "loss": 0.3625,
+      "step": 9084
+    },
+    {
+      "epoch": 1.3902065799540932,
+      "grad_norm": 1.9457973161006021,
+      "learning_rate": 4.4926848739665695e-06,
+      "loss": 0.3176,
+      "step": 9085
+    },
+    {
+      "epoch": 1.3903596021423106,
+      "grad_norm": 2.2547412719369184,
+      "learning_rate": 4.490616547187275e-06,
+      "loss": 0.3324,
+      "step": 9086
+    },
+    {
+      "epoch": 1.390512624330528,
+      "grad_norm": 2.331565110391067,
+      "learning_rate": 4.488548558760778e-06,
+      "loss": 0.3423,
+      "step": 9087
+    },
+    {
+      "epoch": 1.3906656465187452,
+      "grad_norm": 2.0880204887768783,
+      "learning_rate": 4.486480908814068e-06,
+      "loss": 0.3643,
+      "step": 9088
+    },
+    {
+      "epoch": 1.3908186687069626,
+      "grad_norm": 2.1681981316566596,
+      "learning_rate": 4.4844135974741384e-06,
+      "loss": 0.356,
+      "step": 9089
+    },
+    {
+      "epoch": 1.3909716908951797,
+      "grad_norm": 2.149384143224433,
+      "learning_rate": 4.48234662486795e-06,
+      "loss": 0.3372,
+      "step": 9090
+    },
+    {
+      "epoch": 1.391124713083397,
+      "grad_norm": 2.1414181789883484,
+      "learning_rate": 4.480279991122434e-06,
+      "loss": 0.3571,
+      "step": 9091
+    },
+    {
+      "epoch": 1.3912777352716144,
+      "grad_norm": 1.985829509300672,
+      "learning_rate": 4.478213696364519e-06,
+      "loss": 0.2794,
+      "step": 9092
+    },
+    {
+      "epoch": 1.3914307574598317,
+      "grad_norm": 2.155329738496481,
+      "learning_rate": 4.476147740721108e-06,
+      "loss": 0.3552,
+      "step": 9093
+    },
+    {
+      "epoch": 1.391583779648049,
+      "grad_norm": 2.0895352644773655,
+      "learning_rate": 4.4740821243190655e-06,
+      "loss": 0.3342,
+      "step": 9094
+    },
+    {
+      "epoch": 1.3917368018362661,
+      "grad_norm": 2.031285782158425,
+      "learning_rate": 4.472016847285261e-06,
+      "loss": 0.3187,
+      "step": 9095
+    },
+    {
+      "epoch": 1.3918898240244835,
+      "grad_norm": 2.179929418116259,
+      "learning_rate": 4.469951909746527e-06,
+      "loss": 0.2988,
+      "step": 9096
+    },
+    {
+      "epoch": 1.3920428462127008,
+      "grad_norm": 2.2347385172890992,
+      "learning_rate": 4.467887311829681e-06,
+      "loss": 0.3867,
+      "step": 9097
+    },
+    {
+      "epoch": 1.3921958684009181,
+      "grad_norm": 2.03613306275373,
+      "learning_rate": 4.465823053661515e-06,
+      "loss": 0.4267,
+      "step": 9098
+    },
+    {
+      "epoch": 1.3923488905891355,
+      "grad_norm": 2.362451409078388,
+      "learning_rate": 4.463759135368807e-06,
+      "loss": 0.3879,
+      "step": 9099
+    },
+    {
+      "epoch": 1.3925019127773526,
+      "grad_norm": 2.3193830575924856,
+      "learning_rate": 4.461695557078307e-06,
+      "loss": 0.3888,
+      "step": 9100
+    },
+    {
+      "epoch": 1.3926549349655701,
+      "grad_norm": 2.0010960754941793,
+      "learning_rate": 4.45963231891675e-06,
+      "loss": 0.3607,
+      "step": 9101
+    },
+    {
+      "epoch": 1.3928079571537872,
+      "grad_norm": 1.956606094360901,
+      "learning_rate": 4.457569421010846e-06,
+      "loss": 0.2785,
+      "step": 9102
+    },
+    {
+      "epoch": 1.3929609793420046,
+      "grad_norm": 1.659279994173601,
+      "learning_rate": 4.455506863487285e-06,
+      "loss": 0.2374,
+      "step": 9103
+    },
+    {
+      "epoch": 1.393114001530222,
+      "grad_norm": 2.005492432335445,
+      "learning_rate": 4.453444646472736e-06,
+      "loss": 0.3199,
+      "step": 9104
+    },
+    {
+      "epoch": 1.3932670237184392,
+      "grad_norm": 2.291401472114738,
+      "learning_rate": 4.451382770093855e-06,
+      "loss": 0.3307,
+      "step": 9105
+    },
+    {
+      "epoch": 1.3934200459066566,
+      "grad_norm": 2.2285881389564364,
+      "learning_rate": 4.44932123447726e-06,
+      "loss": 0.3942,
+      "step": 9106
+    },
+    {
+      "epoch": 1.3935730680948737,
+      "grad_norm": 2.5366833904976667,
+      "learning_rate": 4.447260039749562e-06,
+      "loss": 0.4648,
+      "step": 9107
+    },
+    {
+      "epoch": 1.393726090283091,
+      "grad_norm": 2.1735252628773116,
+      "learning_rate": 4.445199186037352e-06,
+      "loss": 0.357,
+      "step": 9108
+    },
+    {
+      "epoch": 1.3938791124713084,
+      "grad_norm": 1.949213191530467,
+      "learning_rate": 4.443138673467183e-06,
+      "loss": 0.2779,
+      "step": 9109
+    },
+    {
+      "epoch": 1.3940321346595257,
+      "grad_norm": 2.0589593708206877,
+      "learning_rate": 4.441078502165613e-06,
+      "loss": 0.3583,
+      "step": 9110
+    },
+    {
+      "epoch": 1.394185156847743,
+      "grad_norm": 2.0886474227263174,
+      "learning_rate": 4.43901867225916e-06,
+      "loss": 0.3245,
+      "step": 9111
+    },
+    {
+      "epoch": 1.3943381790359601,
+      "grad_norm": 2.2564496042284574,
+      "learning_rate": 4.4369591838743175e-06,
+      "loss": 0.4147,
+      "step": 9112
+    },
+    {
+      "epoch": 1.3944912012241775,
+      "grad_norm": 2.224803250753932,
+      "learning_rate": 4.434900037137579e-06,
+      "loss": 0.3826,
+      "step": 9113
+    },
+    {
+      "epoch": 1.3946442234123948,
+      "grad_norm": 1.9444166418014062,
+      "learning_rate": 4.432841232175404e-06,
+      "loss": 0.3276,
+      "step": 9114
+    },
+    {
+      "epoch": 1.3947972456006121,
+      "grad_norm": 1.9814704822594493,
+      "learning_rate": 4.43078276911422e-06,
+      "loss": 0.3291,
+      "step": 9115
+    },
+    {
+      "epoch": 1.3949502677888295,
+      "grad_norm": 2.202246242241708,
+      "learning_rate": 4.428724648080457e-06,
+      "loss": 0.3486,
+      "step": 9116
+    },
+    {
+      "epoch": 1.3951032899770466,
+      "grad_norm": 2.1543385539147315,
+      "learning_rate": 4.426666869200507e-06,
+      "loss": 0.3825,
+      "step": 9117
+    },
+    {
+      "epoch": 1.395256312165264,
+      "grad_norm": 1.9988288994094405,
+      "learning_rate": 4.424609432600749e-06,
+      "loss": 0.3201,
+      "step": 9118
+    },
+    {
+      "epoch": 1.3954093343534812,
+      "grad_norm": 2.339767082170093,
+      "learning_rate": 4.422552338407537e-06,
+      "loss": 0.355,
+      "step": 9119
+    },
+    {
+      "epoch": 1.3955623565416986,
+      "grad_norm": 2.377307692928964,
+      "learning_rate": 4.4204955867472035e-06,
+      "loss": 0.392,
+      "step": 9120
+    },
+    {
+      "epoch": 1.395715378729916,
+      "grad_norm": 1.8196435714068888,
+      "learning_rate": 4.418439177746064e-06,
+      "loss": 0.3027,
+      "step": 9121
+    },
+    {
+      "epoch": 1.395868400918133,
+      "grad_norm": 2.2604499545369934,
+      "learning_rate": 4.41638311153041e-06,
+      "loss": 0.3455,
+      "step": 9122
+    },
+    {
+      "epoch": 1.3960214231063504,
+      "grad_norm": 2.033493361083444,
+      "learning_rate": 4.414327388226511e-06,
+      "loss": 0.3,
+      "step": 9123
+    },
+    {
+      "epoch": 1.3961744452945677,
+      "grad_norm": 2.089188809076427,
+      "learning_rate": 4.412272007960621e-06,
+      "loss": 0.3484,
+      "step": 9124
+    },
+    {
+      "epoch": 1.396327467482785,
+      "grad_norm": 2.0463071223170473,
+      "learning_rate": 4.4102169708589635e-06,
+      "loss": 0.3231,
+      "step": 9125
+    },
+    {
+      "epoch": 1.3964804896710024,
+      "grad_norm": 2.088456068827583,
+      "learning_rate": 4.4081622770477505e-06,
+      "loss": 0.411,
+      "step": 9126
+    },
+    {
+      "epoch": 1.3966335118592195,
+      "grad_norm": 2.1862769519850844,
+      "learning_rate": 4.4061079266531685e-06,
+      "loss": 0.3502,
+      "step": 9127
+    },
+    {
+      "epoch": 1.3967865340474368,
+      "grad_norm": 2.45723775464347,
+      "learning_rate": 4.404053919801381e-06,
+      "loss": 0.3739,
+      "step": 9128
+    },
+    {
+      "epoch": 1.3969395562356541,
+      "grad_norm": 2.25312171783395,
+      "learning_rate": 4.4020002566185336e-06,
+      "loss": 0.3956,
+      "step": 9129
+    },
+    {
+      "epoch": 1.3970925784238715,
+      "grad_norm": 2.303669832441205,
+      "learning_rate": 4.39994693723075e-06,
+      "loss": 0.3633,
+      "step": 9130
+    },
+    {
+      "epoch": 1.3972456006120888,
+      "grad_norm": 2.158839638357098,
+      "learning_rate": 4.3978939617641324e-06,
+      "loss": 0.3417,
+      "step": 9131
+    },
+    {
+      "epoch": 1.397398622800306,
+      "grad_norm": 2.1069280338103487,
+      "learning_rate": 4.395841330344762e-06,
+      "loss": 0.3423,
+      "step": 9132
+    },
+    {
+      "epoch": 1.3975516449885235,
+      "grad_norm": 1.982236354173458,
+      "learning_rate": 4.393789043098697e-06,
+      "loss": 0.346,
+      "step": 9133
+    },
+    {
+      "epoch": 1.3977046671767406,
+      "grad_norm": 2.192432973522562,
+      "learning_rate": 4.391737100151984e-06,
+      "loss": 0.4102,
+      "step": 9134
+    },
+    {
+      "epoch": 1.397857689364958,
+      "grad_norm": 2.2270581821455844,
+      "learning_rate": 4.3896855016306324e-06,
+      "loss": 0.4185,
+      "step": 9135
+    },
+    {
+      "epoch": 1.3980107115531752,
+      "grad_norm": 2.299865181546426,
+      "learning_rate": 4.387634247660638e-06,
+      "loss": 0.4115,
+      "step": 9136
+    },
+    {
+      "epoch": 1.3981637337413926,
+      "grad_norm": 2.474822061012731,
+      "learning_rate": 4.385583338367988e-06,
+      "loss": 0.3599,
+      "step": 9137
+    },
+    {
+      "epoch": 1.39831675592961,
+      "grad_norm": 1.9942991712556746,
+      "learning_rate": 4.38353277387862e-06,
+      "loss": 0.3323,
+      "step": 9138
+    },
+    {
+      "epoch": 1.398469778117827,
+      "grad_norm": 1.9557470552238374,
+      "learning_rate": 4.381482554318481e-06,
+      "loss": 0.2981,
+      "step": 9139
+    },
+    {
+      "epoch": 1.3986228003060444,
+      "grad_norm": 1.9394714605145627,
+      "learning_rate": 4.379432679813482e-06,
+      "loss": 0.3279,
+      "step": 9140
+    },
+    {
+      "epoch": 1.3987758224942617,
+      "grad_norm": 2.227238700846652,
+      "learning_rate": 4.377383150489503e-06,
+      "loss": 0.3151,
+      "step": 9141
+    },
+    {
+      "epoch": 1.398928844682479,
+      "grad_norm": 1.9792871389602236,
+      "learning_rate": 4.375333966472423e-06,
+      "loss": 0.2772,
+      "step": 9142
+    },
+    {
+      "epoch": 1.3990818668706964,
+      "grad_norm": 1.9119147272515082,
+      "learning_rate": 4.373285127888093e-06,
+      "loss": 0.3299,
+      "step": 9143
+    },
+    {
+      "epoch": 1.3992348890589135,
+      "grad_norm": 2.0573878948922553,
+      "learning_rate": 4.3712366348623256e-06,
+      "loss": 0.3293,
+      "step": 9144
+    },
+    {
+      "epoch": 1.3993879112471308,
+      "grad_norm": 2.012802138222529,
+      "learning_rate": 4.369188487520942e-06,
+      "loss": 0.3336,
+      "step": 9145
+    },
+    {
+      "epoch": 1.3995409334353481,
+      "grad_norm": 2.168983085038054,
+      "learning_rate": 4.36714068598972e-06,
+      "loss": 0.4029,
+      "step": 9146
+    },
+    {
+      "epoch": 1.3996939556235655,
+      "grad_norm": 2.1144699722556517,
+      "learning_rate": 4.365093230394425e-06,
+      "loss": 0.337,
+      "step": 9147
+    },
+    {
+      "epoch": 1.3998469778117828,
+      "grad_norm": 2.0987987285199305,
+      "learning_rate": 4.363046120860799e-06,
+      "loss": 0.3245,
+      "step": 9148
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.138558119857808,
+      "learning_rate": 4.360999357514562e-06,
+      "loss": 0.3627,
+      "step": 9149
+    },
+    {
+      "epoch": 1.4001530221882172,
+      "grad_norm": 2.1546083763172135,
+      "learning_rate": 4.358952940481414e-06,
+      "loss": 0.3407,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4003060443764346,
+      "grad_norm": 2.13815010724717,
+      "learning_rate": 4.356906869887034e-06,
+      "loss": 0.3007,
+      "step": 9151
+    },
+    {
+      "epoch": 1.400459066564652,
+      "grad_norm": 2.084680308601072,
+      "learning_rate": 4.354861145857079e-06,
+      "loss": 0.3639,
+      "step": 9152
+    },
+    {
+      "epoch": 1.4006120887528692,
+      "grad_norm": 1.9804950825280374,
+      "learning_rate": 4.352815768517185e-06,
+      "loss": 0.2538,
+      "step": 9153
+    },
+    {
+      "epoch": 1.4007651109410864,
+      "grad_norm": 1.9922835195123851,
+      "learning_rate": 4.350770737992967e-06,
+      "loss": 0.2989,
+      "step": 9154
+    },
+    {
+      "epoch": 1.4009181331293037,
+      "grad_norm": 2.0487239910505006,
+      "learning_rate": 4.3487260544100176e-06,
+      "loss": 0.3304,
+      "step": 9155
+    },
+    {
+      "epoch": 1.401071155317521,
+      "grad_norm": 2.0799090584250557,
+      "learning_rate": 4.346681717893909e-06,
+      "loss": 0.3208,
+      "step": 9156
+    },
+    {
+      "epoch": 1.4012241775057384,
+      "grad_norm": 2.3668579542580246,
+      "learning_rate": 4.3446377285701924e-06,
+      "loss": 0.3714,
+      "step": 9157
+    },
+    {
+      "epoch": 1.4013771996939557,
+      "grad_norm": 2.3259095379886516,
+      "learning_rate": 4.342594086564396e-06,
+      "loss": 0.3701,
+      "step": 9158
+    },
+    {
+      "epoch": 1.4015302218821728,
+      "grad_norm": 2.2272276688925636,
+      "learning_rate": 4.340550792002029e-06,
+      "loss": 0.3585,
+      "step": 9159
+    },
+    {
+      "epoch": 1.4016832440703901,
+      "grad_norm": 2.041804323945471,
+      "learning_rate": 4.338507845008578e-06,
+      "loss": 0.3692,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4018362662586075,
+      "grad_norm": 2.2361335197569723,
+      "learning_rate": 4.3364652457095095e-06,
+      "loss": 0.3295,
+      "step": 9161
+    },
+    {
+      "epoch": 1.4019892884468248,
+      "grad_norm": 2.1316794949736573,
+      "learning_rate": 4.3344229942302605e-06,
+      "loss": 0.3014,
+      "step": 9162
+    },
+    {
+      "epoch": 1.4021423106350421,
+      "grad_norm": 2.06167169272297,
+      "learning_rate": 4.332381090696267e-06,
+      "loss": 0.341,
+      "step": 9163
+    },
+    {
+      "epoch": 1.4022953328232592,
+      "grad_norm": 1.9514353960688686,
+      "learning_rate": 4.3303395352329195e-06,
+      "loss": 0.2991,
+      "step": 9164
+    },
+    {
+      "epoch": 1.4024483550114768,
+      "grad_norm": 2.2184375708560595,
+      "learning_rate": 4.3282983279655965e-06,
+      "loss": 0.2893,
+      "step": 9165
+    },
+    {
+      "epoch": 1.402601377199694,
+      "grad_norm": 2.0421161222765765,
+      "learning_rate": 4.326257469019668e-06,
+      "loss": 0.2997,
+      "step": 9166
+    },
+    {
+      "epoch": 1.4027543993879112,
+      "grad_norm": 1.9957994860733428,
+      "learning_rate": 4.324216958520462e-06,
+      "loss": 0.3234,
+      "step": 9167
+    },
+    {
+      "epoch": 1.4029074215761286,
+      "grad_norm": 2.4207732562865116,
+      "learning_rate": 4.3221767965932915e-06,
+      "loss": 0.344,
+      "step": 9168
+    },
+    {
+      "epoch": 1.403060443764346,
+      "grad_norm": 1.8290840140032734,
+      "learning_rate": 4.320136983363463e-06,
+      "loss": 0.2701,
+      "step": 9169
+    },
+    {
+      "epoch": 1.4032134659525632,
+      "grad_norm": 2.079210865586468,
+      "learning_rate": 4.318097518956236e-06,
+      "loss": 0.3439,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4033664881407804,
+      "grad_norm": 2.3346282697773617,
+      "learning_rate": 4.31605840349687e-06,
+      "loss": 0.3681,
+      "step": 9171
+    },
+    {
+      "epoch": 1.4035195103289977,
+      "grad_norm": 2.3745803397537566,
+      "learning_rate": 4.314019637110598e-06,
+      "loss": 0.4203,
+      "step": 9172
+    },
+    {
+      "epoch": 1.403672532517215,
+      "grad_norm": 2.2195262607467954,
+      "learning_rate": 4.311981219922616e-06,
+      "loss": 0.3696,
+      "step": 9173
+    },
+    {
+      "epoch": 1.4038255547054324,
+      "grad_norm": 2.449637868284868,
+      "learning_rate": 4.309943152058122e-06,
+      "loss": 0.4486,
+      "step": 9174
+    },
+    {
+      "epoch": 1.4039785768936497,
+      "grad_norm": 2.3354398337723348,
+      "learning_rate": 4.30790543364228e-06,
+      "loss": 0.413,
+      "step": 9175
+    },
+    {
+      "epoch": 1.4041315990818668,
+      "grad_norm": 2.036115746289367,
+      "learning_rate": 4.305868064800233e-06,
+      "loss": 0.3874,
+      "step": 9176
+    },
+    {
+      "epoch": 1.4042846212700841,
+      "grad_norm": 2.0707373753851543,
+      "learning_rate": 4.303831045657102e-06,
+      "loss": 0.297,
+      "step": 9177
+    },
+    {
+      "epoch": 1.4044376434583015,
+      "grad_norm": 2.2471500234006294,
+      "learning_rate": 4.301794376337991e-06,
+      "loss": 0.3593,
+      "step": 9178
+    },
+    {
+      "epoch": 1.4045906656465188,
+      "grad_norm": 2.130251359667783,
+      "learning_rate": 4.29975805696798e-06,
+      "loss": 0.3498,
+      "step": 9179
+    },
+    {
+      "epoch": 1.4047436878347361,
+      "grad_norm": 2.135027547460175,
+      "learning_rate": 4.297722087672125e-06,
+      "loss": 0.2999,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4048967100229532,
+      "grad_norm": 2.021138639473653,
+      "learning_rate": 4.295686468575464e-06,
+      "loss": 0.2895,
+      "step": 9181
+    },
+    {
+      "epoch": 1.4050497322111706,
+      "grad_norm": 2.0694117569321455,
+      "learning_rate": 4.2936511998030116e-06,
+      "loss": 0.3878,
+      "step": 9182
+    },
+    {
+      "epoch": 1.405202754399388,
+      "grad_norm": 2.098084448075512,
+      "learning_rate": 4.29161628147976e-06,
+      "loss": 0.3577,
+      "step": 9183
+    },
+    {
+      "epoch": 1.4053557765876052,
+      "grad_norm": 2.300338937806773,
+      "learning_rate": 4.289581713730691e-06,
+      "loss": 0.3517,
+      "step": 9184
+    },
+    {
+      "epoch": 1.4055087987758226,
+      "grad_norm": 2.2472611334731547,
+      "learning_rate": 4.287547496680744e-06,
+      "loss": 0.3293,
+      "step": 9185
+    },
+    {
+      "epoch": 1.4056618209640397,
+      "grad_norm": 1.9924191518899794,
+      "learning_rate": 4.2855136304548495e-06,
+      "loss": 0.3571,
+      "step": 9186
+    },
+    {
+      "epoch": 1.405814843152257,
+      "grad_norm": 2.5741502175394353,
+      "learning_rate": 4.283480115177925e-06,
+      "loss": 0.3318,
+      "step": 9187
+    },
+    {
+      "epoch": 1.4059678653404744,
+      "grad_norm": 2.2106988688252307,
+      "learning_rate": 4.2814469509748465e-06,
+      "loss": 0.3486,
+      "step": 9188
+    },
+    {
+      "epoch": 1.4061208875286917,
+      "grad_norm": 2.037048253663022,
+      "learning_rate": 4.279414137970478e-06,
+      "loss": 0.3917,
+      "step": 9189
+    },
+    {
+      "epoch": 1.406273909716909,
+      "grad_norm": 2.367760531391993,
+      "learning_rate": 4.277381676289673e-06,
+      "loss": 0.4048,
+      "step": 9190
+    },
+    {
+      "epoch": 1.4064269319051261,
+      "grad_norm": 2.3062782712082472,
+      "learning_rate": 4.27534956605724e-06,
+      "loss": 0.3714,
+      "step": 9191
+    },
+    {
+      "epoch": 1.4065799540933435,
+      "grad_norm": 2.023102891670905,
+      "learning_rate": 4.273317807397989e-06,
+      "loss": 0.3212,
+      "step": 9192
+    },
+    {
+      "epoch": 1.4067329762815608,
+      "grad_norm": 2.2230748607200814,
+      "learning_rate": 4.271286400436697e-06,
+      "loss": 0.3564,
+      "step": 9193
+    },
+    {
+      "epoch": 1.4068859984697781,
+      "grad_norm": 1.8733281634689514,
+      "learning_rate": 4.269255345298111e-06,
+      "loss": 0.336,
+      "step": 9194
+    },
+    {
+      "epoch": 1.4070390206579955,
+      "grad_norm": 2.0370688321086527,
+      "learning_rate": 4.267224642106977e-06,
+      "loss": 0.3371,
+      "step": 9195
+    },
+    {
+      "epoch": 1.4071920428462126,
+      "grad_norm": 2.140419683830952,
+      "learning_rate": 4.265194290988008e-06,
+      "loss": 0.3348,
+      "step": 9196
+    },
+    {
+      "epoch": 1.40734506503443,
+      "grad_norm": 2.0255097904919013,
+      "learning_rate": 4.2631642920658845e-06,
+      "loss": 0.3399,
+      "step": 9197
+    },
+    {
+      "epoch": 1.4074980872226472,
+      "grad_norm": 2.6557100376557865,
+      "learning_rate": 4.261134645465288e-06,
+      "loss": 0.3875,
+      "step": 9198
+    },
+    {
+      "epoch": 1.4076511094108646,
+      "grad_norm": 2.3843033837943812,
+      "learning_rate": 4.259105351310864e-06,
+      "loss": 0.3838,
+      "step": 9199
+    },
+    {
+      "epoch": 1.407804131599082,
+      "grad_norm": 2.2177489481232744,
+      "learning_rate": 4.2570764097272385e-06,
+      "loss": 0.3138,
+      "step": 9200
+    },
+    {
+      "epoch": 1.407957153787299,
+      "grad_norm": 2.1430301544228767,
+      "learning_rate": 4.255047820839018e-06,
+      "loss": 0.3316,
+      "step": 9201
+    },
+    {
+      "epoch": 1.4081101759755166,
+      "grad_norm": 2.03948756178882,
+      "learning_rate": 4.253019584770784e-06,
+      "loss": 0.3338,
+      "step": 9202
+    },
+    {
+      "epoch": 1.4082631981637337,
+      "grad_norm": 2.7064411052587674,
+      "learning_rate": 4.2509917016471e-06,
+      "loss": 0.3696,
+      "step": 9203
+    },
+    {
+      "epoch": 1.408416220351951,
+      "grad_norm": 2.157120376788771,
+      "learning_rate": 4.248964171592506e-06,
+      "loss": 0.3714,
+      "step": 9204
+    },
+    {
+      "epoch": 1.4085692425401684,
+      "grad_norm": 1.9402279140569512,
+      "learning_rate": 4.246936994731521e-06,
+      "loss": 0.3287,
+      "step": 9205
+    },
+    {
+      "epoch": 1.4087222647283857,
+      "grad_norm": 2.0005560931211925,
+      "learning_rate": 4.2449101711886405e-06,
+      "loss": 0.265,
+      "step": 9206
+    },
+    {
+      "epoch": 1.408875286916603,
+      "grad_norm": 2.0099989536188745,
+      "learning_rate": 4.242883701088342e-06,
+      "loss": 0.3342,
+      "step": 9207
+    },
+    {
+      "epoch": 1.4090283091048201,
+      "grad_norm": 1.89190018721622,
+      "learning_rate": 4.240857584555075e-06,
+      "loss": 0.3328,
+      "step": 9208
+    },
+    {
+      "epoch": 1.4091813312930375,
+      "grad_norm": 2.094044689048361,
+      "learning_rate": 4.2388318217132755e-06,
+      "loss": 0.3219,
+      "step": 9209
+    },
+    {
+      "epoch": 1.4093343534812548,
+      "grad_norm": 2.1431376339884576,
+      "learning_rate": 4.23680641268735e-06,
+      "loss": 0.3437,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4094873756694721,
+      "grad_norm": 2.1750133312455393,
+      "learning_rate": 4.2347813576016896e-06,
+      "loss": 0.3898,
+      "step": 9211
+    },
+    {
+      "epoch": 1.4096403978576895,
+      "grad_norm": 2.026617738848952,
+      "learning_rate": 4.232756656580655e-06,
+      "loss": 0.3322,
+      "step": 9212
+    },
+    {
+      "epoch": 1.4097934200459066,
+      "grad_norm": 2.2552918578481944,
+      "learning_rate": 4.2307323097486045e-06,
+      "loss": 0.3443,
+      "step": 9213
+    },
+    {
+      "epoch": 1.409946442234124,
+      "grad_norm": 2.237233568584304,
+      "learning_rate": 4.228708317229849e-06,
+      "loss": 0.3377,
+      "step": 9214
+    },
+    {
+      "epoch": 1.4100994644223412,
+      "grad_norm": 2.1424246978573374,
+      "learning_rate": 4.226684679148687e-06,
+      "loss": 0.3334,
+      "step": 9215
+    },
+    {
+      "epoch": 1.4102524866105586,
+      "grad_norm": 2.0460843887698483,
+      "learning_rate": 4.2246613956294135e-06,
+      "loss": 0.3062,
+      "step": 9216
+    },
+    {
+      "epoch": 1.410405508798776,
+      "grad_norm": 2.144472320092002,
+      "learning_rate": 4.222638466796272e-06,
+      "loss": 0.2989,
+      "step": 9217
+    },
+    {
+      "epoch": 1.410558530986993,
+      "grad_norm": 1.9545756311695621,
+      "learning_rate": 4.220615892773501e-06,
+      "loss": 0.2725,
+      "step": 9218
+    },
+    {
+      "epoch": 1.4107115531752104,
+      "grad_norm": 2.2589372985105705,
+      "learning_rate": 4.218593673685324e-06,
+      "loss": 0.3786,
+      "step": 9219
+    },
+    {
+      "epoch": 1.4108645753634277,
+      "grad_norm": 2.010967215877556,
+      "learning_rate": 4.2165718096559196e-06,
+      "loss": 0.2905,
+      "step": 9220
+    },
+    {
+      "epoch": 1.411017597551645,
+      "grad_norm": 2.0807598973462884,
+      "learning_rate": 4.214550300809468e-06,
+      "loss": 0.3706,
+      "step": 9221
+    },
+    {
+      "epoch": 1.4111706197398624,
+      "grad_norm": 2.365584366602693,
+      "learning_rate": 4.212529147270119e-06,
+      "loss": 0.397,
+      "step": 9222
+    },
+    {
+      "epoch": 1.4113236419280795,
+      "grad_norm": 2.510130358885097,
+      "learning_rate": 4.210508349161989e-06,
+      "loss": 0.415,
+      "step": 9223
+    },
+    {
+      "epoch": 1.4114766641162968,
+      "grad_norm": 2.2722053771676265,
+      "learning_rate": 4.208487906609193e-06,
+      "loss": 0.3437,
+      "step": 9224
+    },
+    {
+      "epoch": 1.4116296863045141,
+      "grad_norm": 2.2739261112065448,
+      "learning_rate": 4.2064678197358155e-06,
+      "loss": 0.3325,
+      "step": 9225
+    },
+    {
+      "epoch": 1.4117827084927315,
+      "grad_norm": 2.0286388443546968,
+      "learning_rate": 4.204448088665906e-06,
+      "loss": 0.2738,
+      "step": 9226
+    },
+    {
+      "epoch": 1.4119357306809488,
+      "grad_norm": 2.021410749339718,
+      "learning_rate": 4.2024287135235156e-06,
+      "loss": 0.3109,
+      "step": 9227
+    },
+    {
+      "epoch": 1.412088752869166,
+      "grad_norm": 2.1979983118021713,
+      "learning_rate": 4.200409694432658e-06,
+      "loss": 0.3539,
+      "step": 9228
+    },
+    {
+      "epoch": 1.4122417750573832,
+      "grad_norm": 2.0107755272944243,
+      "learning_rate": 4.198391031517328e-06,
+      "loss": 0.3057,
+      "step": 9229
+    },
+    {
+      "epoch": 1.4123947972456006,
+      "grad_norm": 1.9654965383083711,
+      "learning_rate": 4.196372724901502e-06,
+      "loss": 0.3215,
+      "step": 9230
+    },
+    {
+      "epoch": 1.412547819433818,
+      "grad_norm": 2.0359147985124344,
+      "learning_rate": 4.1943547747091306e-06,
+      "loss": 0.2846,
+      "step": 9231
+    },
+    {
+      "epoch": 1.4127008416220352,
+      "grad_norm": 2.098622048300682,
+      "learning_rate": 4.192337181064145e-06,
+      "loss": 0.3234,
+      "step": 9232
+    },
+    {
+      "epoch": 1.4128538638102524,
+      "grad_norm": 2.4112177721252963,
+      "learning_rate": 4.190319944090452e-06,
+      "loss": 0.4037,
+      "step": 9233
+    },
+    {
+      "epoch": 1.41300688599847,
+      "grad_norm": 2.1986701033338933,
+      "learning_rate": 4.1883030639119385e-06,
+      "loss": 0.3305,
+      "step": 9234
+    },
+    {
+      "epoch": 1.413159908186687,
+      "grad_norm": 2.1304605397689422,
+      "learning_rate": 4.18628654065247e-06,
+      "loss": 0.3071,
+      "step": 9235
+    },
+    {
+      "epoch": 1.4133129303749044,
+      "grad_norm": 2.5068322091767747,
+      "learning_rate": 4.1842703744358885e-06,
+      "loss": 0.4024,
+      "step": 9236
+    },
+    {
+      "epoch": 1.4134659525631217,
+      "grad_norm": 1.8202902088615422,
+      "learning_rate": 4.182254565386015e-06,
+      "loss": 0.2486,
+      "step": 9237
+    },
+    {
+      "epoch": 1.413618974751339,
+      "grad_norm": 1.968323001514379,
+      "learning_rate": 4.180239113626646e-06,
+      "loss": 0.3794,
+      "step": 9238
+    },
+    {
+      "epoch": 1.4137719969395564,
+      "grad_norm": 2.0430665390514577,
+      "learning_rate": 4.1782240192815614e-06,
+      "loss": 0.3269,
+      "step": 9239
+    },
+    {
+      "epoch": 1.4139250191277735,
+      "grad_norm": 2.2109860232560603,
+      "learning_rate": 4.176209282474515e-06,
+      "loss": 0.319,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4140780413159908,
+      "grad_norm": 2.0509503774215148,
+      "learning_rate": 4.174194903329239e-06,
+      "loss": 0.33,
+      "step": 9241
+    },
+    {
+      "epoch": 1.4142310635042081,
+      "grad_norm": 2.07717813707438,
+      "learning_rate": 4.1721808819694445e-06,
+      "loss": 0.2624,
+      "step": 9242
+    },
+    {
+      "epoch": 1.4143840856924255,
+      "grad_norm": 2.294446630754429,
+      "learning_rate": 4.17016721851882e-06,
+      "loss": 0.3071,
+      "step": 9243
+    },
+    {
+      "epoch": 1.4145371078806428,
+      "grad_norm": 2.5392494654134987,
+      "learning_rate": 4.168153913101029e-06,
+      "loss": 0.3624,
+      "step": 9244
+    },
+    {
+      "epoch": 1.41469013006886,
+      "grad_norm": 1.887829520244782,
+      "learning_rate": 4.16614096583973e-06,
+      "loss": 0.3119,
+      "step": 9245
+    },
+    {
+      "epoch": 1.4148431522570772,
+      "grad_norm": 2.5210457622741456,
+      "learning_rate": 4.1641283768585315e-06,
+      "loss": 0.3963,
+      "step": 9246
+    },
+    {
+      "epoch": 1.4149961744452946,
+      "grad_norm": 2.0711867206232166,
+      "learning_rate": 4.162116146281036e-06,
+      "loss": 0.296,
+      "step": 9247
+    },
+    {
+      "epoch": 1.415149196633512,
+      "grad_norm": 2.062152923881803,
+      "learning_rate": 4.160104274230833e-06,
+      "loss": 0.3765,
+      "step": 9248
+    },
+    {
+      "epoch": 1.4153022188217292,
+      "grad_norm": 2.0768357136167945,
+      "learning_rate": 4.1580927608314645e-06,
+      "loss": 0.4489,
+      "step": 9249
+    },
+    {
+      "epoch": 1.4154552410099464,
+      "grad_norm": 2.096843231214301,
+      "learning_rate": 4.156081606206478e-06,
+      "loss": 0.3397,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4156082631981637,
+      "grad_norm": 1.947566872119692,
+      "learning_rate": 4.154070810479385e-06,
+      "loss": 0.316,
+      "step": 9251
+    },
+    {
+      "epoch": 1.415761285386381,
+      "grad_norm": 2.180262409466973,
+      "learning_rate": 4.152060373773665e-06,
+      "loss": 0.3205,
+      "step": 9252
+    },
+    {
+      "epoch": 1.4159143075745984,
+      "grad_norm": 2.19393243995704,
+      "learning_rate": 4.150050296212799e-06,
+      "loss": 0.328,
+      "step": 9253
+    },
+    {
+      "epoch": 1.4160673297628157,
+      "grad_norm": 2.2667999392118077,
+      "learning_rate": 4.148040577920233e-06,
+      "loss": 0.3422,
+      "step": 9254
+    },
+    {
+      "epoch": 1.4162203519510328,
+      "grad_norm": 2.3500757721659347,
+      "learning_rate": 4.146031219019381e-06,
+      "loss": 0.3223,
+      "step": 9255
+    },
+    {
+      "epoch": 1.4163733741392501,
+      "grad_norm": 2.038336086469945,
+      "learning_rate": 4.144022219633656e-06,
+      "loss": 0.3369,
+      "step": 9256
+    },
+    {
+      "epoch": 1.4165263963274675,
+      "grad_norm": 2.47373157191876,
+      "learning_rate": 4.142013579886435e-06,
+      "loss": 0.3522,
+      "step": 9257
+    },
+    {
+      "epoch": 1.4166794185156848,
+      "grad_norm": 2.006527184254753,
+      "learning_rate": 4.140005299901076e-06,
+      "loss": 0.3342,
+      "step": 9258
+    },
+    {
+      "epoch": 1.4168324407039021,
+      "grad_norm": 2.0188062531388566,
+      "learning_rate": 4.137997379800916e-06,
+      "loss": 0.2992,
+      "step": 9259
+    },
+    {
+      "epoch": 1.4169854628921192,
+      "grad_norm": 2.1837388368801305,
+      "learning_rate": 4.135989819709271e-06,
+      "loss": 0.3306,
+      "step": 9260
+    },
+    {
+      "epoch": 1.4171384850803366,
+      "grad_norm": 2.0237473797483907,
+      "learning_rate": 4.1339826197494305e-06,
+      "loss": 0.3127,
+      "step": 9261
+    },
+    {
+      "epoch": 1.417291507268554,
+      "grad_norm": 2.056314469069344,
+      "learning_rate": 4.131975780044665e-06,
+      "loss": 0.3071,
+      "step": 9262
+    },
+    {
+      "epoch": 1.4174445294567712,
+      "grad_norm": 2.196217556752408,
+      "learning_rate": 4.129969300718223e-06,
+      "loss": 0.3395,
+      "step": 9263
+    },
+    {
+      "epoch": 1.4175975516449886,
+      "grad_norm": 2.1032234644707835,
+      "learning_rate": 4.1279631818933295e-06,
+      "loss": 0.2709,
+      "step": 9264
+    },
+    {
+      "epoch": 1.4177505738332057,
+      "grad_norm": 2.4031479684877817,
+      "learning_rate": 4.125957423693186e-06,
+      "loss": 0.42,
+      "step": 9265
+    },
+    {
+      "epoch": 1.4179035960214232,
+      "grad_norm": 2.0786897004206732,
+      "learning_rate": 4.123952026240982e-06,
+      "loss": 0.3923,
+      "step": 9266
+    },
+    {
+      "epoch": 1.4180566182096404,
+      "grad_norm": 2.251632349890319,
+      "learning_rate": 4.121946989659869e-06,
+      "loss": 0.3852,
+      "step": 9267
+    },
+    {
+      "epoch": 1.4182096403978577,
+      "grad_norm": 2.555790821467786,
+      "learning_rate": 4.119942314072982e-06,
+      "loss": 0.411,
+      "step": 9268
+    },
+    {
+      "epoch": 1.418362662586075,
+      "grad_norm": 2.349511897533703,
+      "learning_rate": 4.117937999603448e-06,
+      "loss": 0.3667,
+      "step": 9269
+    },
+    {
+      "epoch": 1.4185156847742924,
+      "grad_norm": 2.4392205660352904,
+      "learning_rate": 4.115934046374348e-06,
+      "loss": 0.316,
+      "step": 9270
+    },
+    {
+      "epoch": 1.4186687069625097,
+      "grad_norm": 2.148913506260671,
+      "learning_rate": 4.1139304545087545e-06,
+      "loss": 0.3347,
+      "step": 9271
+    },
+    {
+      "epoch": 1.4188217291507268,
+      "grad_norm": 2.0878267481393658,
+      "learning_rate": 4.111927224129724e-06,
+      "loss": 0.342,
+      "step": 9272
+    },
+    {
+      "epoch": 1.4189747513389441,
+      "grad_norm": 2.161297083283504,
+      "learning_rate": 4.109924355360271e-06,
+      "loss": 0.3463,
+      "step": 9273
+    },
+    {
+      "epoch": 1.4191277735271615,
+      "grad_norm": 1.976150388866175,
+      "learning_rate": 4.107921848323409e-06,
+      "loss": 0.3227,
+      "step": 9274
+    },
+    {
+      "epoch": 1.4192807957153788,
+      "grad_norm": 2.19615636726198,
+      "learning_rate": 4.1059197031421185e-06,
+      "loss": 0.4099,
+      "step": 9275
+    },
+    {
+      "epoch": 1.4194338179035961,
+      "grad_norm": 1.9991519943828997,
+      "learning_rate": 4.1039179199393495e-06,
+      "loss": 0.2936,
+      "step": 9276
+    },
+    {
+      "epoch": 1.4195868400918132,
+      "grad_norm": 2.1666577229545427,
+      "learning_rate": 4.101916498838052e-06,
+      "loss": 0.3918,
+      "step": 9277
+    },
+    {
+      "epoch": 1.4197398622800306,
+      "grad_norm": 2.0607595100678964,
+      "learning_rate": 4.099915439961135e-06,
+      "loss": 0.2556,
+      "step": 9278
+    },
+    {
+      "epoch": 1.419892884468248,
+      "grad_norm": 2.2451488681433407,
+      "learning_rate": 4.097914743431491e-06,
+      "loss": 0.3761,
+      "step": 9279
+    },
+    {
+      "epoch": 1.4200459066564652,
+      "grad_norm": 2.2674171815366364,
+      "learning_rate": 4.095914409371994e-06,
+      "loss": 0.3863,
+      "step": 9280
+    },
+    {
+      "epoch": 1.4201989288446826,
+      "grad_norm": 2.1163719464876447,
+      "learning_rate": 4.093914437905489e-06,
+      "loss": 0.3283,
+      "step": 9281
+    },
+    {
+      "epoch": 1.4203519510328997,
+      "grad_norm": 2.193899121026817,
+      "learning_rate": 4.091914829154801e-06,
+      "loss": 0.3804,
+      "step": 9282
+    },
+    {
+      "epoch": 1.420504973221117,
+      "grad_norm": 2.1351207017567355,
+      "learning_rate": 4.089915583242738e-06,
+      "loss": 0.3554,
+      "step": 9283
+    },
+    {
+      "epoch": 1.4206579954093344,
+      "grad_norm": 2.151449603371385,
+      "learning_rate": 4.087916700292079e-06,
+      "loss": 0.3385,
+      "step": 9284
+    },
+    {
+      "epoch": 1.4208110175975517,
+      "grad_norm": 2.2234211243154087,
+      "learning_rate": 4.0859181804255845e-06,
+      "loss": 0.3383,
+      "step": 9285
+    },
+    {
+      "epoch": 1.420964039785769,
+      "grad_norm": 2.2507111139185745,
+      "learning_rate": 4.08392002376599e-06,
+      "loss": 0.3811,
+      "step": 9286
+    },
+    {
+      "epoch": 1.4211170619739861,
+      "grad_norm": 1.973399321101313,
+      "learning_rate": 4.08192223043601e-06,
+      "loss": 0.2961,
+      "step": 9287
+    },
+    {
+      "epoch": 1.4212700841622035,
+      "grad_norm": 2.133709019598902,
+      "learning_rate": 4.079924800558338e-06,
+      "loss": 0.3478,
+      "step": 9288
+    },
+    {
+      "epoch": 1.4214231063504208,
+      "grad_norm": 2.158294322387067,
+      "learning_rate": 4.077927734255643e-06,
+      "loss": 0.3751,
+      "step": 9289
+    },
+    {
+      "epoch": 1.4215761285386381,
+      "grad_norm": 1.9920504708122626,
+      "learning_rate": 4.075931031650574e-06,
+      "loss": 0.3415,
+      "step": 9290
+    },
+    {
+      "epoch": 1.4217291507268555,
+      "grad_norm": 2.20512398485947,
+      "learning_rate": 4.073934692865755e-06,
+      "loss": 0.3244,
+      "step": 9291
+    },
+    {
+      "epoch": 1.4218821729150726,
+      "grad_norm": 2.347120560371027,
+      "learning_rate": 4.07193871802379e-06,
+      "loss": 0.3721,
+      "step": 9292
+    },
+    {
+      "epoch": 1.42203519510329,
+      "grad_norm": 2.067692285877767,
+      "learning_rate": 4.069943107247259e-06,
+      "loss": 0.315,
+      "step": 9293
+    },
+    {
+      "epoch": 1.4221882172915072,
+      "grad_norm": 2.0118785839217943,
+      "learning_rate": 4.0679478606587155e-06,
+      "loss": 0.318,
+      "step": 9294
+    },
+    {
+      "epoch": 1.4223412394797246,
+      "grad_norm": 2.1728967714428493,
+      "learning_rate": 4.065952978380708e-06,
+      "loss": 0.3713,
+      "step": 9295
+    },
+    {
+      "epoch": 1.422494261667942,
+      "grad_norm": 2.311326910138758,
+      "learning_rate": 4.063958460535738e-06,
+      "loss": 0.3365,
+      "step": 9296
+    },
+    {
+      "epoch": 1.422647283856159,
+      "grad_norm": 2.1496506605822905,
+      "learning_rate": 4.061964307246299e-06,
+      "loss": 0.3167,
+      "step": 9297
+    },
+    {
+      "epoch": 1.4228003060443766,
+      "grad_norm": 2.143129053849377,
+      "learning_rate": 4.059970518634867e-06,
+      "loss": 0.3739,
+      "step": 9298
+    },
+    {
+      "epoch": 1.4229533282325937,
+      "grad_norm": 2.3506052011866156,
+      "learning_rate": 4.05797709482388e-06,
+      "loss": 0.3054,
+      "step": 9299
+    },
+    {
+      "epoch": 1.423106350420811,
+      "grad_norm": 2.0868900431403006,
+      "learning_rate": 4.055984035935762e-06,
+      "loss": 0.3172,
+      "step": 9300
+    },
+    {
+      "epoch": 1.4232593726090284,
+      "grad_norm": 2.064767658180076,
+      "learning_rate": 4.0539913420929235e-06,
+      "loss": 0.2893,
+      "step": 9301
+    },
+    {
+      "epoch": 1.4234123947972457,
+      "grad_norm": 2.1290755477464978,
+      "learning_rate": 4.051999013417731e-06,
+      "loss": 0.3727,
+      "step": 9302
+    },
+    {
+      "epoch": 1.423565416985463,
+      "grad_norm": 1.8483480553704854,
+      "learning_rate": 4.050007050032552e-06,
+      "loss": 0.2864,
+      "step": 9303
+    },
+    {
+      "epoch": 1.4237184391736801,
+      "grad_norm": 2.0023364387746447,
+      "learning_rate": 4.048015452059719e-06,
+      "loss": 0.3626,
+      "step": 9304
+    },
+    {
+      "epoch": 1.4238714613618975,
+      "grad_norm": 1.8518706736246955,
+      "learning_rate": 4.046024219621535e-06,
+      "loss": 0.2707,
+      "step": 9305
+    },
+    {
+      "epoch": 1.4240244835501148,
+      "grad_norm": 1.9748324071119407,
+      "learning_rate": 4.044033352840299e-06,
+      "loss": 0.3093,
+      "step": 9306
+    },
+    {
+      "epoch": 1.4241775057383321,
+      "grad_norm": 1.9481838626006736,
+      "learning_rate": 4.042042851838278e-06,
+      "loss": 0.2265,
+      "step": 9307
+    },
+    {
+      "epoch": 1.4243305279265495,
+      "grad_norm": 2.210466757271439,
+      "learning_rate": 4.040052716737707e-06,
+      "loss": 0.3629,
+      "step": 9308
+    },
+    {
+      "epoch": 1.4244835501147666,
+      "grad_norm": 2.0610090670428973,
+      "learning_rate": 4.038062947660817e-06,
+      "loss": 0.3107,
+      "step": 9309
+    },
+    {
+      "epoch": 1.424636572302984,
+      "grad_norm": 2.149631892573452,
+      "learning_rate": 4.0360735447298035e-06,
+      "loss": 0.3422,
+      "step": 9310
+    },
+    {
+      "epoch": 1.4247895944912012,
+      "grad_norm": 1.990278427506984,
+      "learning_rate": 4.034084508066846e-06,
+      "loss": 0.3334,
+      "step": 9311
+    },
+    {
+      "epoch": 1.4249426166794186,
+      "grad_norm": 2.2489829280017135,
+      "learning_rate": 4.0320958377940976e-06,
+      "loss": 0.3528,
+      "step": 9312
+    },
+    {
+      "epoch": 1.425095638867636,
+      "grad_norm": 2.3206819559913052,
+      "learning_rate": 4.03010753403369e-06,
+      "loss": 0.3797,
+      "step": 9313
+    },
+    {
+      "epoch": 1.425248661055853,
+      "grad_norm": 2.0766681221229097,
+      "learning_rate": 4.028119596907734e-06,
+      "loss": 0.3592,
+      "step": 9314
+    },
+    {
+      "epoch": 1.4254016832440703,
+      "grad_norm": 2.0848585897926974,
+      "learning_rate": 4.026132026538315e-06,
+      "loss": 0.3219,
+      "step": 9315
+    },
+    {
+      "epoch": 1.4255547054322877,
+      "grad_norm": 2.3234590571082636,
+      "learning_rate": 4.024144823047498e-06,
+      "loss": 0.3569,
+      "step": 9316
+    },
+    {
+      "epoch": 1.425707727620505,
+      "grad_norm": 2.00027416391328,
+      "learning_rate": 4.022157986557325e-06,
+      "loss": 0.3802,
+      "step": 9317
+    },
+    {
+      "epoch": 1.4258607498087223,
+      "grad_norm": 2.0622902572976707,
+      "learning_rate": 4.020171517189816e-06,
+      "loss": 0.3386,
+      "step": 9318
+    },
+    {
+      "epoch": 1.4260137719969395,
+      "grad_norm": 1.8924055155330906,
+      "learning_rate": 4.0181854150669665e-06,
+      "loss": 0.3028,
+      "step": 9319
+    },
+    {
+      "epoch": 1.4261667941851568,
+      "grad_norm": 2.576666626932052,
+      "learning_rate": 4.016199680310753e-06,
+      "loss": 0.3428,
+      "step": 9320
+    },
+    {
+      "epoch": 1.4263198163733741,
+      "grad_norm": 2.248274794714831,
+      "learning_rate": 4.014214313043124e-06,
+      "loss": 0.4452,
+      "step": 9321
+    },
+    {
+      "epoch": 1.4264728385615915,
+      "grad_norm": 1.8804937826661376,
+      "learning_rate": 4.012229313386013e-06,
+      "loss": 0.308,
+      "step": 9322
+    },
+    {
+      "epoch": 1.4266258607498088,
+      "grad_norm": 2.0792187323412,
+      "learning_rate": 4.010244681461319e-06,
+      "loss": 0.3295,
+      "step": 9323
+    },
+    {
+      "epoch": 1.426778882938026,
+      "grad_norm": 2.050108419893287,
+      "learning_rate": 4.008260417390938e-06,
+      "loss": 0.3332,
+      "step": 9324
+    },
+    {
+      "epoch": 1.4269319051262432,
+      "grad_norm": 2.073816356773294,
+      "learning_rate": 4.0062765212967215e-06,
+      "loss": 0.3315,
+      "step": 9325
+    },
+    {
+      "epoch": 1.4270849273144606,
+      "grad_norm": 2.05419996109476,
+      "learning_rate": 4.004292993300508e-06,
+      "loss": 0.3372,
+      "step": 9326
+    },
+    {
+      "epoch": 1.427237949502678,
+      "grad_norm": 2.4100490477444887,
+      "learning_rate": 4.002309833524124e-06,
+      "loss": 0.3862,
+      "step": 9327
+    },
+    {
+      "epoch": 1.4273909716908952,
+      "grad_norm": 1.9825638628012534,
+      "learning_rate": 4.000327042089353e-06,
+      "loss": 0.3253,
+      "step": 9328
+    },
+    {
+      "epoch": 1.4275439938791123,
+      "grad_norm": 2.2351531713408934,
+      "learning_rate": 3.998344619117965e-06,
+      "loss": 0.3989,
+      "step": 9329
+    },
+    {
+      "epoch": 1.4276970160673297,
+      "grad_norm": 2.0848182917817617,
+      "learning_rate": 3.996362564731721e-06,
+      "loss": 0.3643,
+      "step": 9330
+    },
+    {
+      "epoch": 1.427850038255547,
+      "grad_norm": 2.1695084103350344,
+      "learning_rate": 3.99438087905233e-06,
+      "loss": 0.3361,
+      "step": 9331
+    },
+    {
+      "epoch": 1.4280030604437643,
+      "grad_norm": 2.384032138139611,
+      "learning_rate": 3.992399562201507e-06,
+      "loss": 0.3408,
+      "step": 9332
+    },
+    {
+      "epoch": 1.4281560826319817,
+      "grad_norm": 1.984735482035343,
+      "learning_rate": 3.990418614300931e-06,
+      "loss": 0.2962,
+      "step": 9333
+    },
+    {
+      "epoch": 1.4283091048201988,
+      "grad_norm": 1.8606884139982514,
+      "learning_rate": 3.988438035472253e-06,
+      "loss": 0.3065,
+      "step": 9334
+    },
+    {
+      "epoch": 1.4284621270084163,
+      "grad_norm": 2.0350449158318558,
+      "learning_rate": 3.986457825837115e-06,
+      "loss": 0.2656,
+      "step": 9335
+    },
+    {
+      "epoch": 1.4286151491966335,
+      "grad_norm": 2.2668034428389268,
+      "learning_rate": 3.98447798551713e-06,
+      "loss": 0.3188,
+      "step": 9336
+    },
+    {
+      "epoch": 1.4287681713848508,
+      "grad_norm": 2.507306255993113,
+      "learning_rate": 3.982498514633879e-06,
+      "loss": 0.4137,
+      "step": 9337
+    },
+    {
+      "epoch": 1.4289211935730681,
+      "grad_norm": 2.2283287973014194,
+      "learning_rate": 3.980519413308938e-06,
+      "loss": 0.3417,
+      "step": 9338
+    },
+    {
+      "epoch": 1.4290742157612855,
+      "grad_norm": 2.0412354072050025,
+      "learning_rate": 3.9785406816638474e-06,
+      "loss": 0.371,
+      "step": 9339
+    },
+    {
+      "epoch": 1.4292272379495028,
+      "grad_norm": 2.4175597698942464,
+      "learning_rate": 3.97656231982013e-06,
+      "loss": 0.4123,
+      "step": 9340
+    },
+    {
+      "epoch": 1.42938026013772,
+      "grad_norm": 2.1106879107993812,
+      "learning_rate": 3.9745843278992835e-06,
+      "loss": 0.3441,
+      "step": 9341
+    },
+    {
+      "epoch": 1.4295332823259372,
+      "grad_norm": 2.178146614681262,
+      "learning_rate": 3.9726067060227855e-06,
+      "loss": 0.3439,
+      "step": 9342
+    },
+    {
+      "epoch": 1.4296863045141546,
+      "grad_norm": 1.9155241157979555,
+      "learning_rate": 3.97062945431209e-06,
+      "loss": 0.2897,
+      "step": 9343
+    },
+    {
+      "epoch": 1.429839326702372,
+      "grad_norm": 2.1661492829484477,
+      "learning_rate": 3.968652572888626e-06,
+      "loss": 0.3392,
+      "step": 9344
+    },
+    {
+      "epoch": 1.4299923488905892,
+      "grad_norm": 2.1917874006953553,
+      "learning_rate": 3.966676061873802e-06,
+      "loss": 0.3808,
+      "step": 9345
+    },
+    {
+      "epoch": 1.4301453710788063,
+      "grad_norm": 2.1886007253205038,
+      "learning_rate": 3.9646999213890045e-06,
+      "loss": 0.3263,
+      "step": 9346
+    },
+    {
+      "epoch": 1.4302983932670237,
+      "grad_norm": 1.8647429617175988,
+      "learning_rate": 3.96272415155559e-06,
+      "loss": 0.3029,
+      "step": 9347
+    },
+    {
+      "epoch": 1.430451415455241,
+      "grad_norm": 2.145380663913235,
+      "learning_rate": 3.960748752494912e-06,
+      "loss": 0.3374,
+      "step": 9348
+    },
+    {
+      "epoch": 1.4306044376434583,
+      "grad_norm": 1.8874504028970565,
+      "learning_rate": 3.9587737243282745e-06,
+      "loss": 0.3119,
+      "step": 9349
+    },
+    {
+      "epoch": 1.4307574598316757,
+      "grad_norm": 2.0694188207539903,
+      "learning_rate": 3.9567990671769715e-06,
+      "loss": 0.3817,
+      "step": 9350
+    },
+    {
+      "epoch": 1.4309104820198928,
+      "grad_norm": 1.9065080001772796,
+      "learning_rate": 3.954824781162288e-06,
+      "loss": 0.3096,
+      "step": 9351
+    },
+    {
+      "epoch": 1.4310635042081101,
+      "grad_norm": 2.121242007256196,
+      "learning_rate": 3.952850866405455e-06,
+      "loss": 0.3715,
+      "step": 9352
+    },
+    {
+      "epoch": 1.4312165263963275,
+      "grad_norm": 2.0622979357181963,
+      "learning_rate": 3.950877323027711e-06,
+      "loss": 0.3287,
+      "step": 9353
+    },
+    {
+      "epoch": 1.4313695485845448,
+      "grad_norm": 2.4372797945913227,
+      "learning_rate": 3.948904151150258e-06,
+      "loss": 0.4393,
+      "step": 9354
+    },
+    {
+      "epoch": 1.4315225707727621,
+      "grad_norm": 2.201072981321367,
+      "learning_rate": 3.946931350894267e-06,
+      "loss": 0.2913,
+      "step": 9355
+    },
+    {
+      "epoch": 1.4316755929609792,
+      "grad_norm": 2.1034205574833695,
+      "learning_rate": 3.944958922380904e-06,
+      "loss": 0.3524,
+      "step": 9356
+    },
+    {
+      "epoch": 1.4318286151491966,
+      "grad_norm": 2.172868895461568,
+      "learning_rate": 3.942986865731304e-06,
+      "loss": 0.3317,
+      "step": 9357
+    },
+    {
+      "epoch": 1.431981637337414,
+      "grad_norm": 2.0235574851664198,
+      "learning_rate": 3.94101518106657e-06,
+      "loss": 0.3337,
+      "step": 9358
+    },
+    {
+      "epoch": 1.4321346595256312,
+      "grad_norm": 1.9239559729267717,
+      "learning_rate": 3.939043868507801e-06,
+      "loss": 0.3044,
+      "step": 9359
+    },
+    {
+      "epoch": 1.4322876817138486,
+      "grad_norm": 2.116674414107678,
+      "learning_rate": 3.937072928176057e-06,
+      "loss": 0.3301,
+      "step": 9360
+    },
+    {
+      "epoch": 1.4324407039020657,
+      "grad_norm": 2.171452873241745,
+      "learning_rate": 3.935102360192382e-06,
+      "loss": 0.3639,
+      "step": 9361
+    },
+    {
+      "epoch": 1.432593726090283,
+      "grad_norm": 2.1857081150102005,
+      "learning_rate": 3.933132164677799e-06,
+      "loss": 0.3481,
+      "step": 9362
+    },
+    {
+      "epoch": 1.4327467482785003,
+      "grad_norm": 2.169013596341822,
+      "learning_rate": 3.931162341753302e-06,
+      "loss": 0.3509,
+      "step": 9363
+    },
+    {
+      "epoch": 1.4328997704667177,
+      "grad_norm": 1.896338149593098,
+      "learning_rate": 3.9291928915398685e-06,
+      "loss": 0.2785,
+      "step": 9364
+    },
+    {
+      "epoch": 1.433052792654935,
+      "grad_norm": 2.549158690084208,
+      "learning_rate": 3.927223814158447e-06,
+      "loss": 0.3658,
+      "step": 9365
+    },
+    {
+      "epoch": 1.4332058148431521,
+      "grad_norm": 2.0841384735924176,
+      "learning_rate": 3.925255109729969e-06,
+      "loss": 0.3119,
+      "step": 9366
+    },
+    {
+      "epoch": 1.4333588370313697,
+      "grad_norm": 1.9791111620268143,
+      "learning_rate": 3.923286778375339e-06,
+      "loss": 0.3146,
+      "step": 9367
+    },
+    {
+      "epoch": 1.4335118592195868,
+      "grad_norm": 2.2175978956874953,
+      "learning_rate": 3.92131882021544e-06,
+      "loss": 0.3854,
+      "step": 9368
+    },
+    {
+      "epoch": 1.4336648814078041,
+      "grad_norm": 1.9276236760923329,
+      "learning_rate": 3.919351235371133e-06,
+      "loss": 0.2764,
+      "step": 9369
+    },
+    {
+      "epoch": 1.4338179035960215,
+      "grad_norm": 2.4958786979979513,
+      "learning_rate": 3.917384023963254e-06,
+      "loss": 0.3887,
+      "step": 9370
+    },
+    {
+      "epoch": 1.4339709257842388,
+      "grad_norm": 1.8840779023709011,
+      "learning_rate": 3.9154171861126165e-06,
+      "loss": 0.2659,
+      "step": 9371
+    },
+    {
+      "epoch": 1.4341239479724561,
+      "grad_norm": 1.818201570937649,
+      "learning_rate": 3.913450721940013e-06,
+      "loss": 0.2917,
+      "step": 9372
+    },
+    {
+      "epoch": 1.4342769701606732,
+      "grad_norm": 2.309820936836204,
+      "learning_rate": 3.9114846315662114e-06,
+      "loss": 0.3719,
+      "step": 9373
+    },
+    {
+      "epoch": 1.4344299923488906,
+      "grad_norm": 2.023699879434676,
+      "learning_rate": 3.909518915111957e-06,
+      "loss": 0.3303,
+      "step": 9374
+    },
+    {
+      "epoch": 1.434583014537108,
+      "grad_norm": 1.9069980001503113,
+      "learning_rate": 3.9075535726979715e-06,
+      "loss": 0.2974,
+      "step": 9375
+    },
+    {
+      "epoch": 1.4347360367253252,
+      "grad_norm": 2.170571525569376,
+      "learning_rate": 3.905588604444953e-06,
+      "loss": 0.3023,
+      "step": 9376
+    },
+    {
+      "epoch": 1.4348890589135426,
+      "grad_norm": 2.1143488858850663,
+      "learning_rate": 3.903624010473585e-06,
+      "loss": 0.3012,
+      "step": 9377
+    },
+    {
+      "epoch": 1.4350420811017597,
+      "grad_norm": 2.2298559527760418,
+      "learning_rate": 3.901659790904514e-06,
+      "loss": 0.359,
+      "step": 9378
+    },
+    {
+      "epoch": 1.435195103289977,
+      "grad_norm": 1.8232134485763651,
+      "learning_rate": 3.899695945858367e-06,
+      "loss": 0.3267,
+      "step": 9379
+    },
+    {
+      "epoch": 1.4353481254781943,
+      "grad_norm": 2.029838669991213,
+      "learning_rate": 3.897732475455764e-06,
+      "loss": 0.2858,
+      "step": 9380
+    },
+    {
+      "epoch": 1.4355011476664117,
+      "grad_norm": 1.7961842764250062,
+      "learning_rate": 3.895769379817279e-06,
+      "loss": 0.3031,
+      "step": 9381
+    },
+    {
+      "epoch": 1.435654169854629,
+      "grad_norm": 2.169262090657161,
+      "learning_rate": 3.893806659063473e-06,
+      "loss": 0.2784,
+      "step": 9382
+    },
+    {
+      "epoch": 1.4358071920428461,
+      "grad_norm": 1.9856810913149714,
+      "learning_rate": 3.891844313314893e-06,
+      "loss": 0.2901,
+      "step": 9383
+    },
+    {
+      "epoch": 1.4359602142310635,
+      "grad_norm": 2.2450828272982783,
+      "learning_rate": 3.889882342692043e-06,
+      "loss": 0.3681,
+      "step": 9384
+    },
+    {
+      "epoch": 1.4361132364192808,
+      "grad_norm": 2.304650971104986,
+      "learning_rate": 3.8879207473154245e-06,
+      "loss": 0.3499,
+      "step": 9385
+    },
+    {
+      "epoch": 1.4362662586074981,
+      "grad_norm": 2.057041817931834,
+      "learning_rate": 3.885959527305507e-06,
+      "loss": 0.3643,
+      "step": 9386
+    },
+    {
+      "epoch": 1.4364192807957155,
+      "grad_norm": 2.1277655643358195,
+      "learning_rate": 3.883998682782727e-06,
+      "loss": 0.394,
+      "step": 9387
+    },
+    {
+      "epoch": 1.4365723029839326,
+      "grad_norm": 2.073493486256418,
+      "learning_rate": 3.882038213867516e-06,
+      "loss": 0.3343,
+      "step": 9388
+    },
+    {
+      "epoch": 1.43672532517215,
+      "grad_norm": 2.300417499890103,
+      "learning_rate": 3.880078120680273e-06,
+      "loss": 0.4377,
+      "step": 9389
+    },
+    {
+      "epoch": 1.4368783473603672,
+      "grad_norm": 2.1283040533902344,
+      "learning_rate": 3.878118403341373e-06,
+      "loss": 0.3926,
+      "step": 9390
+    },
+    {
+      "epoch": 1.4370313695485846,
+      "grad_norm": 2.0678784247283066,
+      "learning_rate": 3.876159061971172e-06,
+      "loss": 0.3086,
+      "step": 9391
+    },
+    {
+      "epoch": 1.437184391736802,
+      "grad_norm": 2.5933724046485556,
+      "learning_rate": 3.874200096689999e-06,
+      "loss": 0.4141,
+      "step": 9392
+    },
+    {
+      "epoch": 1.437337413925019,
+      "grad_norm": 2.569529420688582,
+      "learning_rate": 3.872241507618162e-06,
+      "loss": 0.3729,
+      "step": 9393
+    },
+    {
+      "epoch": 1.4374904361132363,
+      "grad_norm": 2.424768703143181,
+      "learning_rate": 3.870283294875946e-06,
+      "loss": 0.3334,
+      "step": 9394
+    },
+    {
+      "epoch": 1.4376434583014537,
+      "grad_norm": 2.075534358433235,
+      "learning_rate": 3.868325458583613e-06,
+      "loss": 0.3691,
+      "step": 9395
+    },
+    {
+      "epoch": 1.437796480489671,
+      "grad_norm": 2.035561827260354,
+      "learning_rate": 3.8663679988614004e-06,
+      "loss": 0.2953,
+      "step": 9396
+    },
+    {
+      "epoch": 1.4379495026778883,
+      "grad_norm": 2.5251761093520715,
+      "learning_rate": 3.864410915829523e-06,
+      "loss": 0.3889,
+      "step": 9397
+    },
+    {
+      "epoch": 1.4381025248661055,
+      "grad_norm": 2.4679716294913505,
+      "learning_rate": 3.862454209608175e-06,
+      "loss": 0.317,
+      "step": 9398
+    },
+    {
+      "epoch": 1.438255547054323,
+      "grad_norm": 2.148348350788539,
+      "learning_rate": 3.860497880317523e-06,
+      "loss": 0.3622,
+      "step": 9399
+    },
+    {
+      "epoch": 1.4384085692425401,
+      "grad_norm": 2.1987933936762842,
+      "learning_rate": 3.858541928077716e-06,
+      "loss": 0.3212,
+      "step": 9400
+    },
+    {
+      "epoch": 1.4385615914307575,
+      "grad_norm": 2.0618331098672638,
+      "learning_rate": 3.856586353008873e-06,
+      "loss": 0.3206,
+      "step": 9401
+    },
+    {
+      "epoch": 1.4387146136189748,
+      "grad_norm": 2.37921277084796,
+      "learning_rate": 3.854631155231096e-06,
+      "loss": 0.4022,
+      "step": 9402
+    },
+    {
+      "epoch": 1.4388676358071921,
+      "grad_norm": 2.0436503449931194,
+      "learning_rate": 3.85267633486446e-06,
+      "loss": 0.2922,
+      "step": 9403
+    },
+    {
+      "epoch": 1.4390206579954095,
+      "grad_norm": 1.8930267404115466,
+      "learning_rate": 3.850721892029019e-06,
+      "loss": 0.3687,
+      "step": 9404
+    },
+    {
+      "epoch": 1.4391736801836266,
+      "grad_norm": 2.01203672849637,
+      "learning_rate": 3.8487678268448005e-06,
+      "loss": 0.3022,
+      "step": 9405
+    },
+    {
+      "epoch": 1.439326702371844,
+      "grad_norm": 2.635657231491509,
+      "learning_rate": 3.846814139431819e-06,
+      "loss": 0.3954,
+      "step": 9406
+    },
+    {
+      "epoch": 1.4394797245600612,
+      "grad_norm": 1.7576099379374932,
+      "learning_rate": 3.844860829910048e-06,
+      "loss": 0.2292,
+      "step": 9407
+    },
+    {
+      "epoch": 1.4396327467482786,
+      "grad_norm": 2.0022483302447966,
+      "learning_rate": 3.842907898399452e-06,
+      "loss": 0.287,
+      "step": 9408
+    },
+    {
+      "epoch": 1.439785768936496,
+      "grad_norm": 2.3685326123050348,
+      "learning_rate": 3.840955345019973e-06,
+      "loss": 0.3695,
+      "step": 9409
+    },
+    {
+      "epoch": 1.439938791124713,
+      "grad_norm": 2.131612009218713,
+      "learning_rate": 3.839003169891519e-06,
+      "loss": 0.3578,
+      "step": 9410
+    },
+    {
+      "epoch": 1.4400918133129303,
+      "grad_norm": 2.3468392356985324,
+      "learning_rate": 3.8370513731339775e-06,
+      "loss": 0.3115,
+      "step": 9411
+    },
+    {
+      "epoch": 1.4402448355011477,
+      "grad_norm": 2.223056491521997,
+      "learning_rate": 3.835099954867228e-06,
+      "loss": 0.3995,
+      "step": 9412
+    },
+    {
+      "epoch": 1.440397857689365,
+      "grad_norm": 2.0760931406496095,
+      "learning_rate": 3.833148915211101e-06,
+      "loss": 0.2893,
+      "step": 9413
+    },
+    {
+      "epoch": 1.4405508798775823,
+      "grad_norm": 2.5262822068969313,
+      "learning_rate": 3.831198254285428e-06,
+      "loss": 0.4318,
+      "step": 9414
+    },
+    {
+      "epoch": 1.4407039020657995,
+      "grad_norm": 2.090177035772777,
+      "learning_rate": 3.829247972210005e-06,
+      "loss": 0.3353,
+      "step": 9415
+    },
+    {
+      "epoch": 1.4408569242540168,
+      "grad_norm": 2.2014983845962877,
+      "learning_rate": 3.827298069104598e-06,
+      "loss": 0.3364,
+      "step": 9416
+    },
+    {
+      "epoch": 1.4410099464422341,
+      "grad_norm": 2.12746696908804,
+      "learning_rate": 3.825348545088967e-06,
+      "loss": 0.3543,
+      "step": 9417
+    },
+    {
+      "epoch": 1.4411629686304515,
+      "grad_norm": 2.1360864766502017,
+      "learning_rate": 3.823399400282838e-06,
+      "loss": 0.3685,
+      "step": 9418
+    },
+    {
+      "epoch": 1.4413159908186688,
+      "grad_norm": 1.8667857752382644,
+      "learning_rate": 3.821450634805915e-06,
+      "loss": 0.2794,
+      "step": 9419
+    },
+    {
+      "epoch": 1.441469013006886,
+      "grad_norm": 2.4708104972068607,
+      "learning_rate": 3.819502248777878e-06,
+      "loss": 0.3114,
+      "step": 9420
+    },
+    {
+      "epoch": 1.4416220351951032,
+      "grad_norm": 2.0565915598013644,
+      "learning_rate": 3.8175542423183865e-06,
+      "loss": 0.3356,
+      "step": 9421
+    },
+    {
+      "epoch": 1.4417750573833206,
+      "grad_norm": 1.8670926466183604,
+      "learning_rate": 3.815606615547075e-06,
+      "loss": 0.3092,
+      "step": 9422
+    },
+    {
+      "epoch": 1.441928079571538,
+      "grad_norm": 2.108441622458775,
+      "learning_rate": 3.8136593685835545e-06,
+      "loss": 0.3217,
+      "step": 9423
+    },
+    {
+      "epoch": 1.4420811017597552,
+      "grad_norm": 2.2213785242745914,
+      "learning_rate": 3.811712501547413e-06,
+      "loss": 0.3894,
+      "step": 9424
+    },
+    {
+      "epoch": 1.4422341239479723,
+      "grad_norm": 2.2379539576928034,
+      "learning_rate": 3.809766014558216e-06,
+      "loss": 0.3155,
+      "step": 9425
+    },
+    {
+      "epoch": 1.4423871461361897,
+      "grad_norm": 2.1450936115838046,
+      "learning_rate": 3.8078199077355017e-06,
+      "loss": 0.3084,
+      "step": 9426
+    },
+    {
+      "epoch": 1.442540168324407,
+      "grad_norm": 1.9620617688500854,
+      "learning_rate": 3.8058741811987966e-06,
+      "loss": 0.3054,
+      "step": 9427
+    },
+    {
+      "epoch": 1.4426931905126243,
+      "grad_norm": 2.3851341339259178,
+      "learning_rate": 3.8039288350675862e-06,
+      "loss": 0.3356,
+      "step": 9428
+    },
+    {
+      "epoch": 1.4428462127008417,
+      "grad_norm": 2.002450463199621,
+      "learning_rate": 3.801983869461342e-06,
+      "loss": 0.3458,
+      "step": 9429
+    },
+    {
+      "epoch": 1.4429992348890588,
+      "grad_norm": 2.1660350030336653,
+      "learning_rate": 3.8000392844995227e-06,
+      "loss": 0.3357,
+      "step": 9430
+    },
+    {
+      "epoch": 1.4431522570772761,
+      "grad_norm": 1.869478686298557,
+      "learning_rate": 3.7980950803015417e-06,
+      "loss": 0.3066,
+      "step": 9431
+    },
+    {
+      "epoch": 1.4433052792654935,
+      "grad_norm": 2.3072334674826576,
+      "learning_rate": 3.7961512569867997e-06,
+      "loss": 0.3314,
+      "step": 9432
+    },
+    {
+      "epoch": 1.4434583014537108,
+      "grad_norm": 2.275252761892947,
+      "learning_rate": 3.7942078146746852e-06,
+      "loss": 0.3406,
+      "step": 9433
+    },
+    {
+      "epoch": 1.4436113236419281,
+      "grad_norm": 1.9311399642400455,
+      "learning_rate": 3.79226475348454e-06,
+      "loss": 0.2448,
+      "step": 9434
+    },
+    {
+      "epoch": 1.4437643458301452,
+      "grad_norm": 1.9937141861498646,
+      "learning_rate": 3.7903220735357037e-06,
+      "loss": 0.3383,
+      "step": 9435
+    },
+    {
+      "epoch": 1.4439173680183628,
+      "grad_norm": 2.2548689227421903,
+      "learning_rate": 3.788379774947486e-06,
+      "loss": 0.3373,
+      "step": 9436
+    },
+    {
+      "epoch": 1.44407039020658,
+      "grad_norm": 2.245991213469082,
+      "learning_rate": 3.786437857839158e-06,
+      "loss": 0.3213,
+      "step": 9437
+    },
+    {
+      "epoch": 1.4442234123947972,
+      "grad_norm": 2.1682580136836727,
+      "learning_rate": 3.7844963223299925e-06,
+      "loss": 0.3057,
+      "step": 9438
+    },
+    {
+      "epoch": 1.4443764345830146,
+      "grad_norm": 2.361384023733611,
+      "learning_rate": 3.782555168539227e-06,
+      "loss": 0.2986,
+      "step": 9439
+    },
+    {
+      "epoch": 1.444529456771232,
+      "grad_norm": 2.0796621089626828,
+      "learning_rate": 3.780614396586064e-06,
+      "loss": 0.3345,
+      "step": 9440
+    },
+    {
+      "epoch": 1.4446824789594492,
+      "grad_norm": 2.1994329703056628,
+      "learning_rate": 3.7786740065897055e-06,
+      "loss": 0.3354,
+      "step": 9441
+    },
+    {
+      "epoch": 1.4448355011476663,
+      "grad_norm": 1.8833177217194372,
+      "learning_rate": 3.776733998669314e-06,
+      "loss": 0.2956,
+      "step": 9442
+    },
+    {
+      "epoch": 1.4449885233358837,
+      "grad_norm": 2.3741820074807496,
+      "learning_rate": 3.774794372944033e-06,
+      "loss": 0.3962,
+      "step": 9443
+    },
+    {
+      "epoch": 1.445141545524101,
+      "grad_norm": 2.190022440966505,
+      "learning_rate": 3.772855129532982e-06,
+      "loss": 0.3202,
+      "step": 9444
+    },
+    {
+      "epoch": 1.4452945677123183,
+      "grad_norm": 2.0964969488525247,
+      "learning_rate": 3.7709162685552592e-06,
+      "loss": 0.333,
+      "step": 9445
+    },
+    {
+      "epoch": 1.4454475899005357,
+      "grad_norm": 2.028210724620882,
+      "learning_rate": 3.768977790129936e-06,
+      "loss": 0.2656,
+      "step": 9446
+    },
+    {
+      "epoch": 1.4456006120887528,
+      "grad_norm": 2.1346199022390726,
+      "learning_rate": 3.7670396943760635e-06,
+      "loss": 0.3149,
+      "step": 9447
+    },
+    {
+      "epoch": 1.4457536342769701,
+      "grad_norm": 2.042384274192098,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.351,
+      "step": 9448
+    },
+    {
+      "epoch": 1.4459066564651875,
+      "grad_norm": 2.1653394714366656,
+      "learning_rate": 3.7631646513587473e-06,
+      "loss": 0.3592,
+      "step": 9449
+    },
+    {
+      "epoch": 1.4460596786534048,
+      "grad_norm": 1.9813776866303934,
+      "learning_rate": 3.7612277043332857e-06,
+      "loss": 0.2843,
+      "step": 9450
+    },
+    {
+      "epoch": 1.4462127008416221,
+      "grad_norm": 2.0487930782771206,
+      "learning_rate": 3.759291140455237e-06,
+      "loss": 0.2968,
+      "step": 9451
+    },
+    {
+      "epoch": 1.4463657230298392,
+      "grad_norm": 2.1717832715799537,
+      "learning_rate": 3.7573549598435334e-06,
+      "loss": 0.3326,
+      "step": 9452
+    },
+    {
+      "epoch": 1.4465187452180566,
+      "grad_norm": 2.469544616284677,
+      "learning_rate": 3.755419162617082e-06,
+      "loss": 0.3397,
+      "step": 9453
+    },
+    {
+      "epoch": 1.446671767406274,
+      "grad_norm": 2.3789326927343257,
+      "learning_rate": 3.7534837488947705e-06,
+      "loss": 0.3217,
+      "step": 9454
+    },
+    {
+      "epoch": 1.4468247895944912,
+      "grad_norm": 2.0716322076496927,
+      "learning_rate": 3.7515487187954536e-06,
+      "loss": 0.3123,
+      "step": 9455
+    },
+    {
+      "epoch": 1.4469778117827086,
+      "grad_norm": 2.0332000399276637,
+      "learning_rate": 3.7496140724379815e-06,
+      "loss": 0.3046,
+      "step": 9456
+    },
+    {
+      "epoch": 1.4471308339709257,
+      "grad_norm": 1.999069011734315,
+      "learning_rate": 3.7476798099411583e-06,
+      "loss": 0.3494,
+      "step": 9457
+    },
+    {
+      "epoch": 1.447283856159143,
+      "grad_norm": 2.2347715534841033,
+      "learning_rate": 3.745745931423772e-06,
+      "loss": 0.3115,
+      "step": 9458
+    },
+    {
+      "epoch": 1.4474368783473603,
+      "grad_norm": 2.0316516810035026,
+      "learning_rate": 3.7438124370046025e-06,
+      "loss": 0.3059,
+      "step": 9459
+    },
+    {
+      "epoch": 1.4475899005355777,
+      "grad_norm": 1.8778006090816908,
+      "learning_rate": 3.7418793268023823e-06,
+      "loss": 0.2661,
+      "step": 9460
+    },
+    {
+      "epoch": 1.447742922723795,
+      "grad_norm": 2.087926127627727,
+      "learning_rate": 3.739946600935831e-06,
+      "loss": 0.3777,
+      "step": 9461
+    },
+    {
+      "epoch": 1.4478959449120121,
+      "grad_norm": 1.910762098685788,
+      "learning_rate": 3.738014259523656e-06,
+      "loss": 0.3092,
+      "step": 9462
+    },
+    {
+      "epoch": 1.4480489671002295,
+      "grad_norm": 2.1419651512255604,
+      "learning_rate": 3.736082302684514e-06,
+      "loss": 0.3237,
+      "step": 9463
+    },
+    {
+      "epoch": 1.4482019892884468,
+      "grad_norm": 1.9853049413866304,
+      "learning_rate": 3.734150730537067e-06,
+      "loss": 0.3298,
+      "step": 9464
+    },
+    {
+      "epoch": 1.4483550114766641,
+      "grad_norm": 2.169481095354853,
+      "learning_rate": 3.7322195431999386e-06,
+      "loss": 0.3032,
+      "step": 9465
+    },
+    {
+      "epoch": 1.4485080336648815,
+      "grad_norm": 2.3886882013046042,
+      "learning_rate": 3.730288740791721e-06,
+      "loss": 0.3152,
+      "step": 9466
+    },
+    {
+      "epoch": 1.4486610558530986,
+      "grad_norm": 2.1726594327249535,
+      "learning_rate": 3.728358323431003e-06,
+      "loss": 0.3891,
+      "step": 9467
+    },
+    {
+      "epoch": 1.4488140780413161,
+      "grad_norm": 2.3278280144910983,
+      "learning_rate": 3.726428291236338e-06,
+      "loss": 0.3213,
+      "step": 9468
+    },
+    {
+      "epoch": 1.4489671002295332,
+      "grad_norm": 2.0213566163252144,
+      "learning_rate": 3.7244986443262464e-06,
+      "loss": 0.2588,
+      "step": 9469
+    },
+    {
+      "epoch": 1.4491201224177506,
+      "grad_norm": 2.123312387822712,
+      "learning_rate": 3.7225693828192476e-06,
+      "loss": 0.364,
+      "step": 9470
+    },
+    {
+      "epoch": 1.449273144605968,
+      "grad_norm": 1.95586400386957,
+      "learning_rate": 3.7206405068338203e-06,
+      "loss": 0.3514,
+      "step": 9471
+    },
+    {
+      "epoch": 1.4494261667941852,
+      "grad_norm": 1.8020489683421461,
+      "learning_rate": 3.718712016488425e-06,
+      "loss": 0.2893,
+      "step": 9472
+    },
+    {
+      "epoch": 1.4495791889824026,
+      "grad_norm": 2.1640656820411133,
+      "learning_rate": 3.716783911901496e-06,
+      "loss": 0.3673,
+      "step": 9473
+    },
+    {
+      "epoch": 1.4497322111706197,
+      "grad_norm": 2.0896268775804034,
+      "learning_rate": 3.7148561931914472e-06,
+      "loss": 0.325,
+      "step": 9474
+    },
+    {
+      "epoch": 1.449885233358837,
+      "grad_norm": 2.1222460335419298,
+      "learning_rate": 3.712928860476668e-06,
+      "loss": 0.3339,
+      "step": 9475
+    },
+    {
+      "epoch": 1.4500382555470543,
+      "grad_norm": 2.3419143356584082,
+      "learning_rate": 3.7110019138755227e-06,
+      "loss": 0.3304,
+      "step": 9476
+    },
+    {
+      "epoch": 1.4501912777352717,
+      "grad_norm": 1.9960251059645366,
+      "learning_rate": 3.7090753535063517e-06,
+      "loss": 0.3324,
+      "step": 9477
+    },
+    {
+      "epoch": 1.450344299923489,
+      "grad_norm": 2.0769593423066866,
+      "learning_rate": 3.7071491794874736e-06,
+      "loss": 0.2987,
+      "step": 9478
+    },
+    {
+      "epoch": 1.4504973221117061,
+      "grad_norm": 2.232353678787646,
+      "learning_rate": 3.7052233919371826e-06,
+      "loss": 0.3502,
+      "step": 9479
+    },
+    {
+      "epoch": 1.4506503442999235,
+      "grad_norm": 2.2035358060269954,
+      "learning_rate": 3.703297990973749e-06,
+      "loss": 0.3136,
+      "step": 9480
+    },
+    {
+      "epoch": 1.4508033664881408,
+      "grad_norm": 2.2382851331924285,
+      "learning_rate": 3.7013729767154172e-06,
+      "loss": 0.3332,
+      "step": 9481
+    },
+    {
+      "epoch": 1.4509563886763581,
+      "grad_norm": 2.2981953616018767,
+      "learning_rate": 3.6994483492804124e-06,
+      "loss": 0.3646,
+      "step": 9482
+    },
+    {
+      "epoch": 1.4511094108645755,
+      "grad_norm": 2.1714996604088523,
+      "learning_rate": 3.6975241087869317e-06,
+      "loss": 0.3799,
+      "step": 9483
+    },
+    {
+      "epoch": 1.4512624330527926,
+      "grad_norm": 2.2662738415618726,
+      "learning_rate": 3.6956002553531513e-06,
+      "loss": 0.3537,
+      "step": 9484
+    },
+    {
+      "epoch": 1.45141545524101,
+      "grad_norm": 2.149039370046887,
+      "learning_rate": 3.6936767890972215e-06,
+      "loss": 0.3257,
+      "step": 9485
+    },
+    {
+      "epoch": 1.4515684774292272,
+      "grad_norm": 1.9961110412792133,
+      "learning_rate": 3.6917537101372714e-06,
+      "loss": 0.2954,
+      "step": 9486
+    },
+    {
+      "epoch": 1.4517214996174446,
+      "grad_norm": 2.0931555997712676,
+      "learning_rate": 3.6898310185914e-06,
+      "loss": 0.318,
+      "step": 9487
+    },
+    {
+      "epoch": 1.451874521805662,
+      "grad_norm": 2.483815223080329,
+      "learning_rate": 3.687908714577698e-06,
+      "loss": 0.3547,
+      "step": 9488
+    },
+    {
+      "epoch": 1.452027543993879,
+      "grad_norm": 2.0723286404767487,
+      "learning_rate": 3.6859867982142126e-06,
+      "loss": 0.2933,
+      "step": 9489
+    },
+    {
+      "epoch": 1.4521805661820963,
+      "grad_norm": 2.1016415401998407,
+      "learning_rate": 3.6840652696189737e-06,
+      "loss": 0.3124,
+      "step": 9490
+    },
+    {
+      "epoch": 1.4523335883703137,
+      "grad_norm": 2.363286862435652,
+      "learning_rate": 3.6821441289100025e-06,
+      "loss": 0.3849,
+      "step": 9491
+    },
+    {
+      "epoch": 1.452486610558531,
+      "grad_norm": 1.9554556428601573,
+      "learning_rate": 3.6802233762052687e-06,
+      "loss": 0.2859,
+      "step": 9492
+    },
+    {
+      "epoch": 1.4526396327467483,
+      "grad_norm": 2.2228505660039666,
+      "learning_rate": 3.6783030116227437e-06,
+      "loss": 0.3699,
+      "step": 9493
+    },
+    {
+      "epoch": 1.4527926549349655,
+      "grad_norm": 2.064645992589437,
+      "learning_rate": 3.676383035280364e-06,
+      "loss": 0.3022,
+      "step": 9494
+    },
+    {
+      "epoch": 1.4529456771231828,
+      "grad_norm": 2.0773468735735694,
+      "learning_rate": 3.674463447296035e-06,
+      "loss": 0.2924,
+      "step": 9495
+    },
+    {
+      "epoch": 1.4530986993114001,
+      "grad_norm": 2.433049644238986,
+      "learning_rate": 3.672544247787654e-06,
+      "loss": 0.3803,
+      "step": 9496
+    },
+    {
+      "epoch": 1.4532517214996175,
+      "grad_norm": 1.9816445100523168,
+      "learning_rate": 3.6706254368730877e-06,
+      "loss": 0.3232,
+      "step": 9497
+    },
+    {
+      "epoch": 1.4534047436878348,
+      "grad_norm": 2.6411701926591324,
+      "learning_rate": 3.6687070146701665e-06,
+      "loss": 0.3543,
+      "step": 9498
+    },
+    {
+      "epoch": 1.453557765876052,
+      "grad_norm": 2.2355609159511505,
+      "learning_rate": 3.6667889812967194e-06,
+      "loss": 0.3351,
+      "step": 9499
+    },
+    {
+      "epoch": 1.4537107880642695,
+      "grad_norm": 2.371135827805477,
+      "learning_rate": 3.664871336870537e-06,
+      "loss": 0.3725,
+      "step": 9500
+    },
+    {
+      "epoch": 1.4538638102524866,
+      "grad_norm": 1.9464656212441733,
+      "learning_rate": 3.662954081509388e-06,
+      "loss": 0.3134,
+      "step": 9501
+    },
+    {
+      "epoch": 1.454016832440704,
+      "grad_norm": 2.5268455510244294,
+      "learning_rate": 3.6610372153310202e-06,
+      "loss": 0.355,
+      "step": 9502
+    },
+    {
+      "epoch": 1.4541698546289212,
+      "grad_norm": 2.0024853844314277,
+      "learning_rate": 3.6591207384531557e-06,
+      "loss": 0.2974,
+      "step": 9503
+    },
+    {
+      "epoch": 1.4543228768171386,
+      "grad_norm": 2.352329885285306,
+      "learning_rate": 3.657204650993491e-06,
+      "loss": 0.3534,
+      "step": 9504
+    },
+    {
+      "epoch": 1.454475899005356,
+      "grad_norm": 2.1327778506979604,
+      "learning_rate": 3.6552889530697024e-06,
+      "loss": 0.3452,
+      "step": 9505
+    },
+    {
+      "epoch": 1.454628921193573,
+      "grad_norm": 1.9937031650471355,
+      "learning_rate": 3.6533736447994404e-06,
+      "loss": 0.2642,
+      "step": 9506
+    },
+    {
+      "epoch": 1.4547819433817903,
+      "grad_norm": 2.178908960404664,
+      "learning_rate": 3.651458726300331e-06,
+      "loss": 0.3035,
+      "step": 9507
+    },
+    {
+      "epoch": 1.4549349655700077,
+      "grad_norm": 1.9255218183461211,
+      "learning_rate": 3.649544197689973e-06,
+      "loss": 0.2867,
+      "step": 9508
+    },
+    {
+      "epoch": 1.455087987758225,
+      "grad_norm": 2.154114129270153,
+      "learning_rate": 3.6476300590859547e-06,
+      "loss": 0.2427,
+      "step": 9509
+    },
+    {
+      "epoch": 1.4552410099464423,
+      "grad_norm": 2.2499540356570407,
+      "learning_rate": 3.6457163106058236e-06,
+      "loss": 0.3877,
+      "step": 9510
+    },
+    {
+      "epoch": 1.4553940321346595,
+      "grad_norm": 2.223283595049156,
+      "learning_rate": 3.643802952367107e-06,
+      "loss": 0.3417,
+      "step": 9511
+    },
+    {
+      "epoch": 1.4555470543228768,
+      "grad_norm": 2.92579443127091,
+      "learning_rate": 3.6418899844873246e-06,
+      "loss": 0.3641,
+      "step": 9512
+    },
+    {
+      "epoch": 1.4557000765110941,
+      "grad_norm": 2.2939770131216632,
+      "learning_rate": 3.639977407083948e-06,
+      "loss": 0.2643,
+      "step": 9513
+    },
+    {
+      "epoch": 1.4558530986993115,
+      "grad_norm": 2.201385659868088,
+      "learning_rate": 3.638065220274435e-06,
+      "loss": 0.3657,
+      "step": 9514
+    },
+    {
+      "epoch": 1.4560061208875288,
+      "grad_norm": 2.522618790905064,
+      "learning_rate": 3.636153424176232e-06,
+      "loss": 0.3216,
+      "step": 9515
+    },
+    {
+      "epoch": 1.456159143075746,
+      "grad_norm": 2.041065696565262,
+      "learning_rate": 3.634242018906736e-06,
+      "loss": 0.3,
+      "step": 9516
+    },
+    {
+      "epoch": 1.4563121652639632,
+      "grad_norm": 1.9566790868626256,
+      "learning_rate": 3.632331004583345e-06,
+      "loss": 0.3274,
+      "step": 9517
+    },
+    {
+      "epoch": 1.4564651874521806,
+      "grad_norm": 2.1886566680108515,
+      "learning_rate": 3.6304203813234198e-06,
+      "loss": 0.357,
+      "step": 9518
+    },
+    {
+      "epoch": 1.456618209640398,
+      "grad_norm": 2.1183396413555466,
+      "learning_rate": 3.6285101492442897e-06,
+      "loss": 0.3473,
+      "step": 9519
+    },
+    {
+      "epoch": 1.4567712318286152,
+      "grad_norm": 2.2295734686348543,
+      "learning_rate": 3.6266003084632804e-06,
+      "loss": 0.2975,
+      "step": 9520
+    },
+    {
+      "epoch": 1.4569242540168323,
+      "grad_norm": 2.412345482780333,
+      "learning_rate": 3.6246908590976826e-06,
+      "loss": 0.3353,
+      "step": 9521
+    },
+    {
+      "epoch": 1.4570772762050497,
+      "grad_norm": 2.1028388857901392,
+      "learning_rate": 3.622781801264753e-06,
+      "loss": 0.3837,
+      "step": 9522
+    },
+    {
+      "epoch": 1.457230298393267,
+      "grad_norm": 2.2847187812274217,
+      "learning_rate": 3.6208731350817437e-06,
+      "loss": 0.3641,
+      "step": 9523
+    },
+    {
+      "epoch": 1.4573833205814843,
+      "grad_norm": 2.4648038833329204,
+      "learning_rate": 3.6189648606658712e-06,
+      "loss": 0.3452,
+      "step": 9524
+    },
+    {
+      "epoch": 1.4575363427697017,
+      "grad_norm": 1.9838995278832425,
+      "learning_rate": 3.617056978134329e-06,
+      "loss": 0.2456,
+      "step": 9525
+    },
+    {
+      "epoch": 1.4576893649579188,
+      "grad_norm": 2.623646002262567,
+      "learning_rate": 3.615149487604288e-06,
+      "loss": 0.3425,
+      "step": 9526
+    },
+    {
+      "epoch": 1.4578423871461361,
+      "grad_norm": 1.950155135421621,
+      "learning_rate": 3.6132423891928947e-06,
+      "loss": 0.2618,
+      "step": 9527
+    },
+    {
+      "epoch": 1.4579954093343535,
+      "grad_norm": 2.081492680781751,
+      "learning_rate": 3.611335683017272e-06,
+      "loss": 0.2437,
+      "step": 9528
+    },
+    {
+      "epoch": 1.4581484315225708,
+      "grad_norm": 2.2862166927483623,
+      "learning_rate": 3.609429369194516e-06,
+      "loss": 0.3857,
+      "step": 9529
+    },
+    {
+      "epoch": 1.4583014537107881,
+      "grad_norm": 1.9554078644755999,
+      "learning_rate": 3.6075234478417032e-06,
+      "loss": 0.2799,
+      "step": 9530
+    },
+    {
+      "epoch": 1.4584544758990052,
+      "grad_norm": 2.0393830421015435,
+      "learning_rate": 3.6056179190758833e-06,
+      "loss": 0.3213,
+      "step": 9531
+    },
+    {
+      "epoch": 1.4586074980872226,
+      "grad_norm": 1.8992585160790252,
+      "learning_rate": 3.6037127830140817e-06,
+      "loss": 0.2712,
+      "step": 9532
+    },
+    {
+      "epoch": 1.45876052027544,
+      "grad_norm": 2.1471319143065704,
+      "learning_rate": 3.6018080397733013e-06,
+      "loss": 0.3378,
+      "step": 9533
+    },
+    {
+      "epoch": 1.4589135424636572,
+      "grad_norm": 2.2031908095595756,
+      "learning_rate": 3.5999036894705185e-06,
+      "loss": 0.3217,
+      "step": 9534
+    },
+    {
+      "epoch": 1.4590665646518746,
+      "grad_norm": 2.3220186298678893,
+      "learning_rate": 3.5979997322226866e-06,
+      "loss": 0.3869,
+      "step": 9535
+    },
+    {
+      "epoch": 1.4592195868400917,
+      "grad_norm": 2.5306612080969524,
+      "learning_rate": 3.5960961681467364e-06,
+      "loss": 0.3477,
+      "step": 9536
+    },
+    {
+      "epoch": 1.4593726090283092,
+      "grad_norm": 2.073885257035563,
+      "learning_rate": 3.5941929973595703e-06,
+      "loss": 0.3113,
+      "step": 9537
+    },
+    {
+      "epoch": 1.4595256312165263,
+      "grad_norm": 2.2510208071748097,
+      "learning_rate": 3.5922902199780775e-06,
+      "loss": 0.3027,
+      "step": 9538
+    },
+    {
+      "epoch": 1.4596786534047437,
+      "grad_norm": 2.170543165202409,
+      "learning_rate": 3.590387836119107e-06,
+      "loss": 0.3865,
+      "step": 9539
+    },
+    {
+      "epoch": 1.459831675592961,
+      "grad_norm": 1.9292932434270147,
+      "learning_rate": 3.58848584589949e-06,
+      "loss": 0.2739,
+      "step": 9540
+    },
+    {
+      "epoch": 1.4599846977811783,
+      "grad_norm": 1.8320076614211738,
+      "learning_rate": 3.586584249436046e-06,
+      "loss": 0.2785,
+      "step": 9541
+    },
+    {
+      "epoch": 1.4601377199693957,
+      "grad_norm": 2.1935291445909013,
+      "learning_rate": 3.5846830468455497e-06,
+      "loss": 0.3469,
+      "step": 9542
+    },
+    {
+      "epoch": 1.4602907421576128,
+      "grad_norm": 2.176716608400466,
+      "learning_rate": 3.58278223824476e-06,
+      "loss": 0.272,
+      "step": 9543
+    },
+    {
+      "epoch": 1.4604437643458301,
+      "grad_norm": 1.799210093073552,
+      "learning_rate": 3.580881823750425e-06,
+      "loss": 0.2937,
+      "step": 9544
+    },
+    {
+      "epoch": 1.4605967865340475,
+      "grad_norm": 1.9702394956780114,
+      "learning_rate": 3.578981803479241e-06,
+      "loss": 0.261,
+      "step": 9545
+    },
+    {
+      "epoch": 1.4607498087222648,
+      "grad_norm": 2.2698743825940917,
+      "learning_rate": 3.577082177547908e-06,
+      "loss": 0.3467,
+      "step": 9546
+    },
+    {
+      "epoch": 1.4609028309104821,
+      "grad_norm": 2.4331017290730057,
+      "learning_rate": 3.575182946073089e-06,
+      "loss": 0.3096,
+      "step": 9547
+    },
+    {
+      "epoch": 1.4610558530986992,
+      "grad_norm": 2.2378192392486937,
+      "learning_rate": 3.573284109171412e-06,
+      "loss": 0.2813,
+      "step": 9548
+    },
+    {
+      "epoch": 1.4612088752869166,
+      "grad_norm": 1.949646785682065,
+      "learning_rate": 3.5713856669595025e-06,
+      "loss": 0.2898,
+      "step": 9549
+    },
+    {
+      "epoch": 1.461361897475134,
+      "grad_norm": 2.3546898453130605,
+      "learning_rate": 3.5694876195539506e-06,
+      "loss": 0.3316,
+      "step": 9550
+    },
+    {
+      "epoch": 1.4615149196633512,
+      "grad_norm": 2.0028492455077735,
+      "learning_rate": 3.5675899670713156e-06,
+      "loss": 0.3081,
+      "step": 9551
+    },
+    {
+      "epoch": 1.4616679418515686,
+      "grad_norm": 2.3157113370587377,
+      "learning_rate": 3.565692709628146e-06,
+      "loss": 0.3438,
+      "step": 9552
+    },
+    {
+      "epoch": 1.4618209640397857,
+      "grad_norm": 2.0500920849676345,
+      "learning_rate": 3.5637958473409595e-06,
+      "loss": 0.283,
+      "step": 9553
+    },
+    {
+      "epoch": 1.461973986228003,
+      "grad_norm": 2.2982221861588386,
+      "learning_rate": 3.561899380326248e-06,
+      "loss": 0.3415,
+      "step": 9554
+    },
+    {
+      "epoch": 1.4621270084162203,
+      "grad_norm": 2.1105916046380777,
+      "learning_rate": 3.5600033087004814e-06,
+      "loss": 0.3392,
+      "step": 9555
+    },
+    {
+      "epoch": 1.4622800306044377,
+      "grad_norm": 2.106406268139071,
+      "learning_rate": 3.5581076325801045e-06,
+      "loss": 0.3018,
+      "step": 9556
+    },
+    {
+      "epoch": 1.462433052792655,
+      "grad_norm": 1.8577142920496017,
+      "learning_rate": 3.5562123520815395e-06,
+      "loss": 0.3208,
+      "step": 9557
+    },
+    {
+      "epoch": 1.4625860749808721,
+      "grad_norm": 1.9650333120954178,
+      "learning_rate": 3.554317467321182e-06,
+      "loss": 0.3118,
+      "step": 9558
+    },
+    {
+      "epoch": 1.4627390971690895,
+      "grad_norm": 1.9942111925567405,
+      "learning_rate": 3.552422978415405e-06,
+      "loss": 0.2956,
+      "step": 9559
+    },
+    {
+      "epoch": 1.4628921193573068,
+      "grad_norm": 1.8524074124963654,
+      "learning_rate": 3.5505288854805555e-06,
+      "loss": 0.2709,
+      "step": 9560
+    },
+    {
+      "epoch": 1.4630451415455241,
+      "grad_norm": 2.328829234349851,
+      "learning_rate": 3.548635188632957e-06,
+      "loss": 0.3441,
+      "step": 9561
+    },
+    {
+      "epoch": 1.4631981637337415,
+      "grad_norm": 2.0595048605859057,
+      "learning_rate": 3.5467418879889104e-06,
+      "loss": 0.3298,
+      "step": 9562
+    },
+    {
+      "epoch": 1.4633511859219586,
+      "grad_norm": 2.1752269089426495,
+      "learning_rate": 3.5448489836646895e-06,
+      "loss": 0.351,
+      "step": 9563
+    },
+    {
+      "epoch": 1.463504208110176,
+      "grad_norm": 2.206069599112191,
+      "learning_rate": 3.5429564757765454e-06,
+      "loss": 0.3253,
+      "step": 9564
+    },
+    {
+      "epoch": 1.4636572302983932,
+      "grad_norm": 2.2034438785230352,
+      "learning_rate": 3.541064364440704e-06,
+      "loss": 0.3659,
+      "step": 9565
+    },
+    {
+      "epoch": 1.4638102524866106,
+      "grad_norm": 2.2927363798245346,
+      "learning_rate": 3.5391726497733637e-06,
+      "loss": 0.4403,
+      "step": 9566
+    },
+    {
+      "epoch": 1.463963274674828,
+      "grad_norm": 2.1978108874987576,
+      "learning_rate": 3.537281331890713e-06,
+      "loss": 0.3383,
+      "step": 9567
+    },
+    {
+      "epoch": 1.464116296863045,
+      "grad_norm": 2.44448450236065,
+      "learning_rate": 3.5353904109088945e-06,
+      "loss": 0.304,
+      "step": 9568
+    },
+    {
+      "epoch": 1.4642693190512626,
+      "grad_norm": 2.137051347698189,
+      "learning_rate": 3.533499886944037e-06,
+      "loss": 0.3344,
+      "step": 9569
+    },
+    {
+      "epoch": 1.4644223412394797,
+      "grad_norm": 2.2505696018802697,
+      "learning_rate": 3.531609760112257e-06,
+      "loss": 0.4358,
+      "step": 9570
+    },
+    {
+      "epoch": 1.464575363427697,
+      "grad_norm": 2.051597472463943,
+      "learning_rate": 3.5297200305296208e-06,
+      "loss": 0.2754,
+      "step": 9571
+    },
+    {
+      "epoch": 1.4647283856159143,
+      "grad_norm": 2.1406649536647047,
+      "learning_rate": 3.527830698312187e-06,
+      "loss": 0.3756,
+      "step": 9572
+    },
+    {
+      "epoch": 1.4648814078041317,
+      "grad_norm": 2.0506454652157005,
+      "learning_rate": 3.525941763575995e-06,
+      "loss": 0.3418,
+      "step": 9573
+    },
+    {
+      "epoch": 1.465034429992349,
+      "grad_norm": 2.1326475184422082,
+      "learning_rate": 3.5240532264370396e-06,
+      "loss": 0.2633,
+      "step": 9574
+    },
+    {
+      "epoch": 1.4651874521805661,
+      "grad_norm": 2.1494489636744185,
+      "learning_rate": 3.5221650870113134e-06,
+      "loss": 0.3263,
+      "step": 9575
+    },
+    {
+      "epoch": 1.4653404743687835,
+      "grad_norm": 2.2083023526559318,
+      "learning_rate": 3.5202773454147733e-06,
+      "loss": 0.3563,
+      "step": 9576
+    },
+    {
+      "epoch": 1.4654934965570008,
+      "grad_norm": 2.1455256554573956,
+      "learning_rate": 3.5183900017633445e-06,
+      "loss": 0.3259,
+      "step": 9577
+    },
+    {
+      "epoch": 1.4656465187452181,
+      "grad_norm": 2.0692219151315685,
+      "learning_rate": 3.516503056172944e-06,
+      "loss": 0.3062,
+      "step": 9578
+    },
+    {
+      "epoch": 1.4657995409334355,
+      "grad_norm": 2.435498607808429,
+      "learning_rate": 3.5146165087594586e-06,
+      "loss": 0.3979,
+      "step": 9579
+    },
+    {
+      "epoch": 1.4659525631216526,
+      "grad_norm": 2.251816963015846,
+      "learning_rate": 3.5127303596387386e-06,
+      "loss": 0.3213,
+      "step": 9580
+    },
+    {
+      "epoch": 1.46610558530987,
+      "grad_norm": 2.1677463950149294,
+      "learning_rate": 3.510844608926627e-06,
+      "loss": 0.3407,
+      "step": 9581
+    },
+    {
+      "epoch": 1.4662586074980872,
+      "grad_norm": 2.4337802808014017,
+      "learning_rate": 3.5089592567389352e-06,
+      "loss": 0.3395,
+      "step": 9582
+    },
+    {
+      "epoch": 1.4664116296863046,
+      "grad_norm": 2.178833688485482,
+      "learning_rate": 3.507074303191448e-06,
+      "loss": 0.3467,
+      "step": 9583
+    },
+    {
+      "epoch": 1.466564651874522,
+      "grad_norm": 2.079719791026871,
+      "learning_rate": 3.5051897483999274e-06,
+      "loss": 0.3201,
+      "step": 9584
+    },
+    {
+      "epoch": 1.466717674062739,
+      "grad_norm": 2.318754978550932,
+      "learning_rate": 3.503305592480113e-06,
+      "loss": 0.31,
+      "step": 9585
+    },
+    {
+      "epoch": 1.4668706962509563,
+      "grad_norm": 2.2671430905226684,
+      "learning_rate": 3.501421835547718e-06,
+      "loss": 0.3336,
+      "step": 9586
+    },
+    {
+      "epoch": 1.4670237184391737,
+      "grad_norm": 1.9679858647231845,
+      "learning_rate": 3.4995384777184295e-06,
+      "loss": 0.3267,
+      "step": 9587
+    },
+    {
+      "epoch": 1.467176740627391,
+      "grad_norm": 2.059412945703858,
+      "learning_rate": 3.4976555191079142e-06,
+      "loss": 0.2965,
+      "step": 9588
+    },
+    {
+      "epoch": 1.4673297628156083,
+      "grad_norm": 1.8621298863000937,
+      "learning_rate": 3.4957729598318104e-06,
+      "loss": 0.3184,
+      "step": 9589
+    },
+    {
+      "epoch": 1.4674827850038255,
+      "grad_norm": 2.0029242916594794,
+      "learning_rate": 3.493890800005729e-06,
+      "loss": 0.3288,
+      "step": 9590
+    },
+    {
+      "epoch": 1.4676358071920428,
+      "grad_norm": 2.104395772468994,
+      "learning_rate": 3.4920090397452743e-06,
+      "loss": 0.2823,
+      "step": 9591
+    },
+    {
+      "epoch": 1.4677888293802601,
+      "grad_norm": 1.8928052562757283,
+      "learning_rate": 3.490127679166e-06,
+      "loss": 0.2928,
+      "step": 9592
+    },
+    {
+      "epoch": 1.4679418515684775,
+      "grad_norm": 2.0663466283826906,
+      "learning_rate": 3.4882467183834478e-06,
+      "loss": 0.274,
+      "step": 9593
+    },
+    {
+      "epoch": 1.4680948737566948,
+      "grad_norm": 1.8060456661072481,
+      "learning_rate": 3.486366157513146e-06,
+      "loss": 0.3044,
+      "step": 9594
+    },
+    {
+      "epoch": 1.468247895944912,
+      "grad_norm": 1.9513142978452787,
+      "learning_rate": 3.4844859966705726e-06,
+      "loss": 0.2943,
+      "step": 9595
+    },
+    {
+      "epoch": 1.4684009181331292,
+      "grad_norm": 2.2091242439116665,
+      "learning_rate": 3.4826062359712065e-06,
+      "loss": 0.299,
+      "step": 9596
+    },
+    {
+      "epoch": 1.4685539403213466,
+      "grad_norm": 2.041058385323502,
+      "learning_rate": 3.4807268755304914e-06,
+      "loss": 0.2695,
+      "step": 9597
+    },
+    {
+      "epoch": 1.468706962509564,
+      "grad_norm": 2.1822736478283944,
+      "learning_rate": 3.4788479154638356e-06,
+      "loss": 0.357,
+      "step": 9598
+    },
+    {
+      "epoch": 1.4688599846977812,
+      "grad_norm": 2.1286379228362606,
+      "learning_rate": 3.4769693558866434e-06,
+      "loss": 0.3365,
+      "step": 9599
+    },
+    {
+      "epoch": 1.4690130068859983,
+      "grad_norm": 2.127037295930163,
+      "learning_rate": 3.475091196914284e-06,
+      "loss": 0.3397,
+      "step": 9600
+    },
+    {
+      "epoch": 1.469166029074216,
+      "grad_norm": 2.1368346349315583,
+      "learning_rate": 3.473213438662094e-06,
+      "loss": 0.3509,
+      "step": 9601
+    },
+    {
+      "epoch": 1.469319051262433,
+      "grad_norm": 2.079146890204384,
+      "learning_rate": 3.4713360812454033e-06,
+      "loss": 0.3229,
+      "step": 9602
+    },
+    {
+      "epoch": 1.4694720734506503,
+      "grad_norm": 2.142254720047488,
+      "learning_rate": 3.4694591247795028e-06,
+      "loss": 0.3195,
+      "step": 9603
+    },
+    {
+      "epoch": 1.4696250956388677,
+      "grad_norm": 2.0262041871453778,
+      "learning_rate": 3.4675825693796662e-06,
+      "loss": 0.3646,
+      "step": 9604
+    },
+    {
+      "epoch": 1.469778117827085,
+      "grad_norm": 1.9264466580480444,
+      "learning_rate": 3.4657064151611385e-06,
+      "loss": 0.3186,
+      "step": 9605
+    },
+    {
+      "epoch": 1.4699311400153023,
+      "grad_norm": 2.0276181691278503,
+      "learning_rate": 3.4638306622391426e-06,
+      "loss": 0.3137,
+      "step": 9606
+    },
+    {
+      "epoch": 1.4700841622035195,
+      "grad_norm": 2.534691854587415,
+      "learning_rate": 3.461955310728875e-06,
+      "loss": 0.4236,
+      "step": 9607
+    },
+    {
+      "epoch": 1.4702371843917368,
+      "grad_norm": 1.9148967223518105,
+      "learning_rate": 3.4600803607455093e-06,
+      "loss": 0.3288,
+      "step": 9608
+    },
+    {
+      "epoch": 1.4703902065799541,
+      "grad_norm": 2.2857281715004416,
+      "learning_rate": 3.4582058124041915e-06,
+      "loss": 0.3491,
+      "step": 9609
+    },
+    {
+      "epoch": 1.4705432287681715,
+      "grad_norm": 2.1665463187107314,
+      "learning_rate": 3.4563316658200484e-06,
+      "loss": 0.3467,
+      "step": 9610
+    },
+    {
+      "epoch": 1.4706962509563888,
+      "grad_norm": 2.143549381240846,
+      "learning_rate": 3.4544579211081753e-06,
+      "loss": 0.3849,
+      "step": 9611
+    },
+    {
+      "epoch": 1.470849273144606,
+      "grad_norm": 1.8653919338791964,
+      "learning_rate": 3.4525845783836474e-06,
+      "loss": 0.238,
+      "step": 9612
+    },
+    {
+      "epoch": 1.4710022953328232,
+      "grad_norm": 2.1729156855165352,
+      "learning_rate": 3.4507116377615157e-06,
+      "loss": 0.3141,
+      "step": 9613
+    },
+    {
+      "epoch": 1.4711553175210406,
+      "grad_norm": 2.3265866813135965,
+      "learning_rate": 3.448839099356802e-06,
+      "loss": 0.3338,
+      "step": 9614
+    },
+    {
+      "epoch": 1.471308339709258,
+      "grad_norm": 1.854469823617688,
+      "learning_rate": 3.4469669632845073e-06,
+      "loss": 0.3019,
+      "step": 9615
+    },
+    {
+      "epoch": 1.4714613618974752,
+      "grad_norm": 1.9374657496486478,
+      "learning_rate": 3.4450952296596072e-06,
+      "loss": 0.2676,
+      "step": 9616
+    },
+    {
+      "epoch": 1.4716143840856923,
+      "grad_norm": 1.9448743453259358,
+      "learning_rate": 3.4432238985970523e-06,
+      "loss": 0.2811,
+      "step": 9617
+    },
+    {
+      "epoch": 1.4717674062739097,
+      "grad_norm": 2.1803253590810434,
+      "learning_rate": 3.441352970211769e-06,
+      "loss": 0.3335,
+      "step": 9618
+    },
+    {
+      "epoch": 1.471920428462127,
+      "grad_norm": 2.1774294836502586,
+      "learning_rate": 3.4394824446186527e-06,
+      "loss": 0.3215,
+      "step": 9619
+    },
+    {
+      "epoch": 1.4720734506503443,
+      "grad_norm": 2.1656602648557874,
+      "learning_rate": 3.437612321932591e-06,
+      "loss": 0.3493,
+      "step": 9620
+    },
+    {
+      "epoch": 1.4722264728385617,
+      "grad_norm": 2.111235629798471,
+      "learning_rate": 3.4357426022684257e-06,
+      "loss": 0.2601,
+      "step": 9621
+    },
+    {
+      "epoch": 1.4723794950267788,
+      "grad_norm": 1.912232160522444,
+      "learning_rate": 3.433873285740984e-06,
+      "loss": 0.3647,
+      "step": 9622
+    },
+    {
+      "epoch": 1.4725325172149961,
+      "grad_norm": 2.31825086042197,
+      "learning_rate": 3.4320043724650763e-06,
+      "loss": 0.3651,
+      "step": 9623
+    },
+    {
+      "epoch": 1.4726855394032135,
+      "grad_norm": 1.7970660470580329,
+      "learning_rate": 3.4301358625554713e-06,
+      "loss": 0.2483,
+      "step": 9624
+    },
+    {
+      "epoch": 1.4728385615914308,
+      "grad_norm": 2.2998009808220927,
+      "learning_rate": 3.4282677561269217e-06,
+      "loss": 0.2858,
+      "step": 9625
+    },
+    {
+      "epoch": 1.4729915837796481,
+      "grad_norm": 2.1552214382883514,
+      "learning_rate": 3.4264000532941644e-06,
+      "loss": 0.3032,
+      "step": 9626
+    },
+    {
+      "epoch": 1.4731446059678652,
+      "grad_norm": 2.1964411418119654,
+      "learning_rate": 3.424532754171889e-06,
+      "loss": 0.3218,
+      "step": 9627
+    },
+    {
+      "epoch": 1.4732976281560826,
+      "grad_norm": 2.0569648465887544,
+      "learning_rate": 3.422665858874784e-06,
+      "loss": 0.3243,
+      "step": 9628
+    },
+    {
+      "epoch": 1.4734506503443,
+      "grad_norm": 2.167669952320896,
+      "learning_rate": 3.4207993675175023e-06,
+      "loss": 0.3219,
+      "step": 9629
+    },
+    {
+      "epoch": 1.4736036725325172,
+      "grad_norm": 1.9886377425796664,
+      "learning_rate": 3.4189332802146623e-06,
+      "loss": 0.3845,
+      "step": 9630
+    },
+    {
+      "epoch": 1.4737566947207346,
+      "grad_norm": 2.3674925318781863,
+      "learning_rate": 3.4170675970808797e-06,
+      "loss": 0.3497,
+      "step": 9631
+    },
+    {
+      "epoch": 1.4739097169089517,
+      "grad_norm": 2.0213186273832533,
+      "learning_rate": 3.415202318230727e-06,
+      "loss": 0.3284,
+      "step": 9632
+    },
+    {
+      "epoch": 1.4740627390971692,
+      "grad_norm": 1.9433332846288336,
+      "learning_rate": 3.4133374437787604e-06,
+      "loss": 0.3333,
+      "step": 9633
+    },
+    {
+      "epoch": 1.4742157612853863,
+      "grad_norm": 1.8829935353968115,
+      "learning_rate": 3.411472973839509e-06,
+      "loss": 0.2995,
+      "step": 9634
+    },
+    {
+      "epoch": 1.4743687834736037,
+      "grad_norm": 2.199541576399467,
+      "learning_rate": 3.4096089085274763e-06,
+      "loss": 0.387,
+      "step": 9635
+    },
+    {
+      "epoch": 1.474521805661821,
+      "grad_norm": 2.320246022684784,
+      "learning_rate": 3.4077452479571425e-06,
+      "loss": 0.3588,
+      "step": 9636
+    },
+    {
+      "epoch": 1.4746748278500381,
+      "grad_norm": 2.2086550938995453,
+      "learning_rate": 3.405881992242962e-06,
+      "loss": 0.396,
+      "step": 9637
+    },
+    {
+      "epoch": 1.4748278500382557,
+      "grad_norm": 2.5880420208227575,
+      "learning_rate": 3.404019141499364e-06,
+      "loss": 0.3122,
+      "step": 9638
+    },
+    {
+      "epoch": 1.4749808722264728,
+      "grad_norm": 2.14820259118592,
+      "learning_rate": 3.4021566958407547e-06,
+      "loss": 0.3345,
+      "step": 9639
+    },
+    {
+      "epoch": 1.4751338944146901,
+      "grad_norm": 2.291144556279864,
+      "learning_rate": 3.4002946553815143e-06,
+      "loss": 0.4546,
+      "step": 9640
+    },
+    {
+      "epoch": 1.4752869166029074,
+      "grad_norm": 2.2789618485242107,
+      "learning_rate": 3.398433020235996e-06,
+      "loss": 0.2864,
+      "step": 9641
+    },
+    {
+      "epoch": 1.4754399387911248,
+      "grad_norm": 2.16718251307733,
+      "learning_rate": 3.3965717905185326e-06,
+      "loss": 0.3261,
+      "step": 9642
+    },
+    {
+      "epoch": 1.4755929609793421,
+      "grad_norm": 1.8345178592877107,
+      "learning_rate": 3.3947109663434274e-06,
+      "loss": 0.2828,
+      "step": 9643
+    },
+    {
+      "epoch": 1.4757459831675592,
+      "grad_norm": 2.0816444823106215,
+      "learning_rate": 3.392850547824962e-06,
+      "loss": 0.3154,
+      "step": 9644
+    },
+    {
+      "epoch": 1.4758990053557766,
+      "grad_norm": 2.2321250664297914,
+      "learning_rate": 3.390990535077392e-06,
+      "loss": 0.3723,
+      "step": 9645
+    },
+    {
+      "epoch": 1.476052027543994,
+      "grad_norm": 1.9008291069198198,
+      "learning_rate": 3.3891309282149476e-06,
+      "loss": 0.3035,
+      "step": 9646
+    },
+    {
+      "epoch": 1.4762050497322112,
+      "grad_norm": 2.0403337195080007,
+      "learning_rate": 3.3872717273518352e-06,
+      "loss": 0.3225,
+      "step": 9647
+    },
+    {
+      "epoch": 1.4763580719204286,
+      "grad_norm": 2.3313497369392864,
+      "learning_rate": 3.3854129326022312e-06,
+      "loss": 0.3249,
+      "step": 9648
+    },
+    {
+      "epoch": 1.4765110941086457,
+      "grad_norm": 1.9367820769137207,
+      "learning_rate": 3.383554544080303e-06,
+      "loss": 0.3423,
+      "step": 9649
+    },
+    {
+      "epoch": 1.476664116296863,
+      "grad_norm": 1.9343688897956826,
+      "learning_rate": 3.3816965619001697e-06,
+      "loss": 0.2844,
+      "step": 9650
+    },
+    {
+      "epoch": 1.4768171384850803,
+      "grad_norm": 2.0350599708053183,
+      "learning_rate": 3.3798389861759385e-06,
+      "loss": 0.2567,
+      "step": 9651
+    },
+    {
+      "epoch": 1.4769701606732977,
+      "grad_norm": 2.0884917699535226,
+      "learning_rate": 3.3779818170216994e-06,
+      "loss": 0.3647,
+      "step": 9652
+    },
+    {
+      "epoch": 1.477123182861515,
+      "grad_norm": 2.053173963511413,
+      "learning_rate": 3.3761250545515e-06,
+      "loss": 0.3367,
+      "step": 9653
+    },
+    {
+      "epoch": 1.4772762050497321,
+      "grad_norm": 2.406740971692602,
+      "learning_rate": 3.3742686988793703e-06,
+      "loss": 0.3357,
+      "step": 9654
+    },
+    {
+      "epoch": 1.4774292272379494,
+      "grad_norm": 2.2395311160784273,
+      "learning_rate": 3.372412750119326e-06,
+      "loss": 0.3188,
+      "step": 9655
+    },
+    {
+      "epoch": 1.4775822494261668,
+      "grad_norm": 2.090272851729783,
+      "learning_rate": 3.3705572083853354e-06,
+      "loss": 0.3041,
+      "step": 9656
+    },
+    {
+      "epoch": 1.4777352716143841,
+      "grad_norm": 1.9825871720189625,
+      "learning_rate": 3.3687020737913645e-06,
+      "loss": 0.2794,
+      "step": 9657
+    },
+    {
+      "epoch": 1.4778882938026014,
+      "grad_norm": 1.799646543850758,
+      "learning_rate": 3.366847346451345e-06,
+      "loss": 0.3137,
+      "step": 9658
+    },
+    {
+      "epoch": 1.4780413159908186,
+      "grad_norm": 2.1457698517236365,
+      "learning_rate": 3.364993026479172e-06,
+      "loss": 0.3418,
+      "step": 9659
+    },
+    {
+      "epoch": 1.478194338179036,
+      "grad_norm": 1.881613328637775,
+      "learning_rate": 3.363139113988736e-06,
+      "loss": 0.3176,
+      "step": 9660
+    },
+    {
+      "epoch": 1.4783473603672532,
+      "grad_norm": 2.0964508147709364,
+      "learning_rate": 3.361285609093895e-06,
+      "loss": 0.368,
+      "step": 9661
+    },
+    {
+      "epoch": 1.4785003825554706,
+      "grad_norm": 2.3956986412823214,
+      "learning_rate": 3.359432511908468e-06,
+      "loss": 0.4279,
+      "step": 9662
+    },
+    {
+      "epoch": 1.478653404743688,
+      "grad_norm": 2.177333571774147,
+      "learning_rate": 3.357579822546273e-06,
+      "loss": 0.3332,
+      "step": 9663
+    },
+    {
+      "epoch": 1.478806426931905,
+      "grad_norm": 1.920267606292228,
+      "learning_rate": 3.3557275411210857e-06,
+      "loss": 0.3456,
+      "step": 9664
+    },
+    {
+      "epoch": 1.4789594491201223,
+      "grad_norm": 2.13065516579347,
+      "learning_rate": 3.353875667746662e-06,
+      "loss": 0.2951,
+      "step": 9665
+    },
+    {
+      "epoch": 1.4791124713083397,
+      "grad_norm": 2.1802357900868103,
+      "learning_rate": 3.3520242025367345e-06,
+      "loss": 0.3092,
+      "step": 9666
+    },
+    {
+      "epoch": 1.479265493496557,
+      "grad_norm": 2.2588888010046184,
+      "learning_rate": 3.350173145605007e-06,
+      "loss": 0.3029,
+      "step": 9667
+    },
+    {
+      "epoch": 1.4794185156847743,
+      "grad_norm": 2.149844137638874,
+      "learning_rate": 3.3483224970651618e-06,
+      "loss": 0.3287,
+      "step": 9668
+    },
+    {
+      "epoch": 1.4795715378729914,
+      "grad_norm": 1.9285140452203176,
+      "learning_rate": 3.3464722570308495e-06,
+      "loss": 0.2797,
+      "step": 9669
+    },
+    {
+      "epoch": 1.479724560061209,
+      "grad_norm": 2.174794412352257,
+      "learning_rate": 3.3446224256157113e-06,
+      "loss": 0.3432,
+      "step": 9670
+    },
+    {
+      "epoch": 1.4798775822494261,
+      "grad_norm": 1.9492744650336518,
+      "learning_rate": 3.3427730029333427e-06,
+      "loss": 0.3068,
+      "step": 9671
+    },
+    {
+      "epoch": 1.4800306044376434,
+      "grad_norm": 2.2467264250993737,
+      "learning_rate": 3.3409239890973234e-06,
+      "loss": 0.3348,
+      "step": 9672
+    },
+    {
+      "epoch": 1.4801836266258608,
+      "grad_norm": 2.148025214316827,
+      "learning_rate": 3.3390753842212196e-06,
+      "loss": 0.3306,
+      "step": 9673
+    },
+    {
+      "epoch": 1.4803366488140781,
+      "grad_norm": 2.3871375437801916,
+      "learning_rate": 3.3372271884185503e-06,
+      "loss": 0.3524,
+      "step": 9674
+    },
+    {
+      "epoch": 1.4804896710022954,
+      "grad_norm": 2.281354552877399,
+      "learning_rate": 3.3353794018028216e-06,
+      "loss": 0.339,
+      "step": 9675
+    },
+    {
+      "epoch": 1.4806426931905126,
+      "grad_norm": 2.16698236232759,
+      "learning_rate": 3.3335320244875226e-06,
+      "loss": 0.3163,
+      "step": 9676
+    },
+    {
+      "epoch": 1.48079571537873,
+      "grad_norm": 1.9526029274383037,
+      "learning_rate": 3.3316850565860937e-06,
+      "loss": 0.2826,
+      "step": 9677
+    },
+    {
+      "epoch": 1.4809487375669472,
+      "grad_norm": 2.420557240605018,
+      "learning_rate": 3.3298384982119767e-06,
+      "loss": 0.3786,
+      "step": 9678
+    },
+    {
+      "epoch": 1.4811017597551646,
+      "grad_norm": 2.28465035245862,
+      "learning_rate": 3.327992349478574e-06,
+      "loss": 0.3461,
+      "step": 9679
+    },
+    {
+      "epoch": 1.481254781943382,
+      "grad_norm": 2.0005123469497588,
+      "learning_rate": 3.3261466104992557e-06,
+      "loss": 0.2779,
+      "step": 9680
+    },
+    {
+      "epoch": 1.481407804131599,
+      "grad_norm": 2.264288271376078,
+      "learning_rate": 3.3243012813873854e-06,
+      "loss": 0.3615,
+      "step": 9681
+    },
+    {
+      "epoch": 1.4815608263198163,
+      "grad_norm": 2.1961574735260805,
+      "learning_rate": 3.322456362256292e-06,
+      "loss": 0.3248,
+      "step": 9682
+    },
+    {
+      "epoch": 1.4817138485080337,
+      "grad_norm": 1.987011464688957,
+      "learning_rate": 3.32061185321927e-06,
+      "loss": 0.33,
+      "step": 9683
+    },
+    {
+      "epoch": 1.481866870696251,
+      "grad_norm": 1.9619117786824731,
+      "learning_rate": 3.318767754389607e-06,
+      "loss": 0.3812,
+      "step": 9684
+    },
+    {
+      "epoch": 1.4820198928844683,
+      "grad_norm": 2.075830782082472,
+      "learning_rate": 3.3169240658805533e-06,
+      "loss": 0.2877,
+      "step": 9685
+    },
+    {
+      "epoch": 1.4821729150726854,
+      "grad_norm": 2.0458494055910927,
+      "learning_rate": 3.3150807878053383e-06,
+      "loss": 0.3171,
+      "step": 9686
+    },
+    {
+      "epoch": 1.4823259372609028,
+      "grad_norm": 2.493267414344924,
+      "learning_rate": 3.3132379202771613e-06,
+      "loss": 0.3898,
+      "step": 9687
+    },
+    {
+      "epoch": 1.4824789594491201,
+      "grad_norm": 2.275969640971838,
+      "learning_rate": 3.3113954634092037e-06,
+      "loss": 0.3653,
+      "step": 9688
+    },
+    {
+      "epoch": 1.4826319816373374,
+      "grad_norm": 2.1293537358603514,
+      "learning_rate": 3.309553417314617e-06,
+      "loss": 0.3176,
+      "step": 9689
+    },
+    {
+      "epoch": 1.4827850038255548,
+      "grad_norm": 2.131470108037654,
+      "learning_rate": 3.3077117821065274e-06,
+      "loss": 0.2997,
+      "step": 9690
+    },
+    {
+      "epoch": 1.482938026013772,
+      "grad_norm": 2.2836857699911315,
+      "learning_rate": 3.305870557898039e-06,
+      "loss": 0.3644,
+      "step": 9691
+    },
+    {
+      "epoch": 1.4830910482019892,
+      "grad_norm": 1.8703424111131515,
+      "learning_rate": 3.3040297448022253e-06,
+      "loss": 0.2544,
+      "step": 9692
+    },
+    {
+      "epoch": 1.4832440703902066,
+      "grad_norm": 1.8819671745260147,
+      "learning_rate": 3.3021893429321407e-06,
+      "loss": 0.2973,
+      "step": 9693
+    },
+    {
+      "epoch": 1.483397092578424,
+      "grad_norm": 2.2023836981052343,
+      "learning_rate": 3.300349352400811e-06,
+      "loss": 0.2955,
+      "step": 9694
+    },
+    {
+      "epoch": 1.4835501147666412,
+      "grad_norm": 1.8542863847355404,
+      "learning_rate": 3.2985097733212356e-06,
+      "loss": 0.259,
+      "step": 9695
+    },
+    {
+      "epoch": 1.4837031369548583,
+      "grad_norm": 2.201198857571715,
+      "learning_rate": 3.2966706058063924e-06,
+      "loss": 0.3678,
+      "step": 9696
+    },
+    {
+      "epoch": 1.4838561591430757,
+      "grad_norm": 2.0005225343509174,
+      "learning_rate": 3.294831849969231e-06,
+      "loss": 0.2636,
+      "step": 9697
+    },
+    {
+      "epoch": 1.484009181331293,
+      "grad_norm": 2.0741880232412813,
+      "learning_rate": 3.292993505922676e-06,
+      "loss": 0.3257,
+      "step": 9698
+    },
+    {
+      "epoch": 1.4841622035195103,
+      "grad_norm": 2.1015544039730294,
+      "learning_rate": 3.2911555737796274e-06,
+      "loss": 0.3469,
+      "step": 9699
+    },
+    {
+      "epoch": 1.4843152257077277,
+      "grad_norm": 2.154984766214943,
+      "learning_rate": 3.28931805365296e-06,
+      "loss": 0.3341,
+      "step": 9700
+    },
+    {
+      "epoch": 1.4844682478959448,
+      "grad_norm": 2.00350612128123,
+      "learning_rate": 3.2874809456555214e-06,
+      "loss": 0.3539,
+      "step": 9701
+    },
+    {
+      "epoch": 1.4846212700841623,
+      "grad_norm": 2.2878236503785114,
+      "learning_rate": 3.285644249900143e-06,
+      "loss": 0.3498,
+      "step": 9702
+    },
+    {
+      "epoch": 1.4847742922723794,
+      "grad_norm": 2.0944591858421227,
+      "learning_rate": 3.283807966499615e-06,
+      "loss": 0.311,
+      "step": 9703
+    },
+    {
+      "epoch": 1.4849273144605968,
+      "grad_norm": 2.117959473650116,
+      "learning_rate": 3.281972095566709e-06,
+      "loss": 0.3289,
+      "step": 9704
+    },
+    {
+      "epoch": 1.4850803366488141,
+      "grad_norm": 2.1069078550225284,
+      "learning_rate": 3.2801366372141854e-06,
+      "loss": 0.2824,
+      "step": 9705
+    },
+    {
+      "epoch": 1.4852333588370314,
+      "grad_norm": 1.7445064623021305,
+      "learning_rate": 3.278301591554753e-06,
+      "loss": 0.2507,
+      "step": 9706
+    },
+    {
+      "epoch": 1.4853863810252488,
+      "grad_norm": 2.1744862309060826,
+      "learning_rate": 3.2764669587011176e-06,
+      "loss": 0.2917,
+      "step": 9707
+    },
+    {
+      "epoch": 1.485539403213466,
+      "grad_norm": 2.099115560590431,
+      "learning_rate": 3.2746327387659528e-06,
+      "loss": 0.3,
+      "step": 9708
+    },
+    {
+      "epoch": 1.4856924254016832,
+      "grad_norm": 2.383542556858228,
+      "learning_rate": 3.272798931861895e-06,
+      "loss": 0.3499,
+      "step": 9709
+    },
+    {
+      "epoch": 1.4858454475899006,
+      "grad_norm": 1.958523894698532,
+      "learning_rate": 3.2709655381015747e-06,
+      "loss": 0.3295,
+      "step": 9710
+    },
+    {
+      "epoch": 1.485998469778118,
+      "grad_norm": 2.316274078181258,
+      "learning_rate": 3.2691325575975873e-06,
+      "loss": 0.2929,
+      "step": 9711
+    },
+    {
+      "epoch": 1.4861514919663352,
+      "grad_norm": 2.285956647889531,
+      "learning_rate": 3.267299990462496e-06,
+      "loss": 0.3307,
+      "step": 9712
+    },
+    {
+      "epoch": 1.4863045141545523,
+      "grad_norm": 2.21087747854537,
+      "learning_rate": 3.265467836808852e-06,
+      "loss": 0.3439,
+      "step": 9713
+    },
+    {
+      "epoch": 1.4864575363427697,
+      "grad_norm": 2.512942424866177,
+      "learning_rate": 3.2636360967491753e-06,
+      "loss": 0.3233,
+      "step": 9714
+    },
+    {
+      "epoch": 1.486610558530987,
+      "grad_norm": 2.2420113787509695,
+      "learning_rate": 3.2618047703959587e-06,
+      "loss": 0.3966,
+      "step": 9715
+    },
+    {
+      "epoch": 1.4867635807192043,
+      "grad_norm": 1.827313128956776,
+      "learning_rate": 3.25997385786167e-06,
+      "loss": 0.3345,
+      "step": 9716
+    },
+    {
+      "epoch": 1.4869166029074217,
+      "grad_norm": 2.540856908588475,
+      "learning_rate": 3.258143359258755e-06,
+      "loss": 0.3662,
+      "step": 9717
+    },
+    {
+      "epoch": 1.4870696250956388,
+      "grad_norm": 2.0729164941270626,
+      "learning_rate": 3.2563132746996295e-06,
+      "loss": 0.2955,
+      "step": 9718
+    },
+    {
+      "epoch": 1.4872226472838561,
+      "grad_norm": 2.2431616508070684,
+      "learning_rate": 3.2544836042966887e-06,
+      "loss": 0.2798,
+      "step": 9719
+    },
+    {
+      "epoch": 1.4873756694720734,
+      "grad_norm": 2.388865441696811,
+      "learning_rate": 3.2526543481622972e-06,
+      "loss": 0.3584,
+      "step": 9720
+    },
+    {
+      "epoch": 1.4875286916602908,
+      "grad_norm": 2.3876493187240446,
+      "learning_rate": 3.250825506408798e-06,
+      "loss": 0.3654,
+      "step": 9721
+    },
+    {
+      "epoch": 1.4876817138485081,
+      "grad_norm": 1.8569930412454958,
+      "learning_rate": 3.248997079148508e-06,
+      "loss": 0.3371,
+      "step": 9722
+    },
+    {
+      "epoch": 1.4878347360367252,
+      "grad_norm": 1.9617608209841546,
+      "learning_rate": 3.247169066493717e-06,
+      "loss": 0.3279,
+      "step": 9723
+    },
+    {
+      "epoch": 1.4879877582249426,
+      "grad_norm": 2.1004602635403073,
+      "learning_rate": 3.245341468556691e-06,
+      "loss": 0.3163,
+      "step": 9724
+    },
+    {
+      "epoch": 1.48814078041316,
+      "grad_norm": 1.8281973503764155,
+      "learning_rate": 3.2435142854496695e-06,
+      "loss": 0.3414,
+      "step": 9725
+    },
+    {
+      "epoch": 1.4882938026013772,
+      "grad_norm": 2.0119727311462468,
+      "learning_rate": 3.241687517284867e-06,
+      "loss": 0.3131,
+      "step": 9726
+    },
+    {
+      "epoch": 1.4884468247895946,
+      "grad_norm": 2.0671398705003354,
+      "learning_rate": 3.239861164174474e-06,
+      "loss": 0.3292,
+      "step": 9727
+    },
+    {
+      "epoch": 1.4885998469778117,
+      "grad_norm": 2.183939371421867,
+      "learning_rate": 3.2380352262306516e-06,
+      "loss": 0.3413,
+      "step": 9728
+    },
+    {
+      "epoch": 1.488752869166029,
+      "grad_norm": 2.6176496586308264,
+      "learning_rate": 3.2362097035655395e-06,
+      "loss": 0.3849,
+      "step": 9729
+    },
+    {
+      "epoch": 1.4889058913542463,
+      "grad_norm": 1.984006435842004,
+      "learning_rate": 3.234384596291246e-06,
+      "loss": 0.2825,
+      "step": 9730
+    },
+    {
+      "epoch": 1.4890589135424637,
+      "grad_norm": 2.218983095645015,
+      "learning_rate": 3.2325599045198686e-06,
+      "loss": 0.3254,
+      "step": 9731
+    },
+    {
+      "epoch": 1.489211935730681,
+      "grad_norm": 2.1406761509420753,
+      "learning_rate": 3.230735628363457e-06,
+      "loss": 0.2822,
+      "step": 9732
+    },
+    {
+      "epoch": 1.4893649579188981,
+      "grad_norm": 2.156889368226719,
+      "learning_rate": 3.228911767934051e-06,
+      "loss": 0.3079,
+      "step": 9733
+    },
+    {
+      "epoch": 1.4895179801071157,
+      "grad_norm": 2.2627592273788806,
+      "learning_rate": 3.2270883233436668e-06,
+      "loss": 0.3561,
+      "step": 9734
+    },
+    {
+      "epoch": 1.4896710022953328,
+      "grad_norm": 2.1268555796252,
+      "learning_rate": 3.225265294704278e-06,
+      "loss": 0.2414,
+      "step": 9735
+    },
+    {
+      "epoch": 1.4898240244835501,
+      "grad_norm": 2.5191995631578443,
+      "learning_rate": 3.2234426821278553e-06,
+      "loss": 0.4362,
+      "step": 9736
+    },
+    {
+      "epoch": 1.4899770466717674,
+      "grad_norm": 2.1702180434659444,
+      "learning_rate": 3.221620485726329e-06,
+      "loss": 0.36,
+      "step": 9737
+    },
+    {
+      "epoch": 1.4901300688599848,
+      "grad_norm": 2.2050301008986017,
+      "learning_rate": 3.219798705611601e-06,
+      "loss": 0.3054,
+      "step": 9738
+    },
+    {
+      "epoch": 1.4902830910482021,
+      "grad_norm": 2.2089552013113334,
+      "learning_rate": 3.2179773418955605e-06,
+      "loss": 0.3696,
+      "step": 9739
+    },
+    {
+      "epoch": 1.4904361132364192,
+      "grad_norm": 1.9418105664383982,
+      "learning_rate": 3.2161563946900665e-06,
+      "loss": 0.2662,
+      "step": 9740
+    },
+    {
+      "epoch": 1.4905891354246366,
+      "grad_norm": 2.1010010388460523,
+      "learning_rate": 3.2143358641069412e-06,
+      "loss": 0.294,
+      "step": 9741
+    },
+    {
+      "epoch": 1.490742157612854,
+      "grad_norm": 2.4393911103033235,
+      "learning_rate": 3.2125157502579985e-06,
+      "loss": 0.3728,
+      "step": 9742
+    },
+    {
+      "epoch": 1.4908951798010712,
+      "grad_norm": 2.1391415962137,
+      "learning_rate": 3.2106960532550158e-06,
+      "loss": 0.298,
+      "step": 9743
+    },
+    {
+      "epoch": 1.4910482019892886,
+      "grad_norm": 2.0613432267115925,
+      "learning_rate": 3.2088767732097493e-06,
+      "loss": 0.2871,
+      "step": 9744
+    },
+    {
+      "epoch": 1.4912012241775057,
+      "grad_norm": 1.974687288819572,
+      "learning_rate": 3.2070579102339274e-06,
+      "loss": 0.3126,
+      "step": 9745
+    },
+    {
+      "epoch": 1.491354246365723,
+      "grad_norm": 2.0887558994869884,
+      "learning_rate": 3.2052394644392525e-06,
+      "loss": 0.3445,
+      "step": 9746
+    },
+    {
+      "epoch": 1.4915072685539403,
+      "grad_norm": 2.186477735593589,
+      "learning_rate": 3.2034214359374038e-06,
+      "loss": 0.3372,
+      "step": 9747
+    },
+    {
+      "epoch": 1.4916602907421577,
+      "grad_norm": 2.3663298192862454,
+      "learning_rate": 3.2016038248400338e-06,
+      "loss": 0.3463,
+      "step": 9748
+    },
+    {
+      "epoch": 1.491813312930375,
+      "grad_norm": 1.9550108079123667,
+      "learning_rate": 3.199786631258768e-06,
+      "loss": 0.3061,
+      "step": 9749
+    },
+    {
+      "epoch": 1.4919663351185921,
+      "grad_norm": 2.179613795100359,
+      "learning_rate": 3.1979698553052065e-06,
+      "loss": 0.3472,
+      "step": 9750
+    },
+    {
+      "epoch": 1.4921193573068094,
+      "grad_norm": 2.2746548727858054,
+      "learning_rate": 3.1961534970909237e-06,
+      "loss": 0.4066,
+      "step": 9751
+    },
+    {
+      "epoch": 1.4922723794950268,
+      "grad_norm": 1.939192701965579,
+      "learning_rate": 3.194337556727478e-06,
+      "loss": 0.2677,
+      "step": 9752
+    },
+    {
+      "epoch": 1.4924254016832441,
+      "grad_norm": 2.264172372199918,
+      "learning_rate": 3.192522034326383e-06,
+      "loss": 0.2917,
+      "step": 9753
+    },
+    {
+      "epoch": 1.4925784238714614,
+      "grad_norm": 2.302377291578821,
+      "learning_rate": 3.1907069299991387e-06,
+      "loss": 0.3661,
+      "step": 9754
+    },
+    {
+      "epoch": 1.4927314460596786,
+      "grad_norm": 1.8882191488728923,
+      "learning_rate": 3.1888922438572257e-06,
+      "loss": 0.2901,
+      "step": 9755
+    },
+    {
+      "epoch": 1.492884468247896,
+      "grad_norm": 2.1078551552877935,
+      "learning_rate": 3.1870779760120818e-06,
+      "loss": 0.3591,
+      "step": 9756
+    },
+    {
+      "epoch": 1.4930374904361132,
+      "grad_norm": 2.277226366949766,
+      "learning_rate": 3.1852641265751294e-06,
+      "loss": 0.3308,
+      "step": 9757
+    },
+    {
+      "epoch": 1.4931905126243306,
+      "grad_norm": 2.0104743672396794,
+      "learning_rate": 3.1834506956577716e-06,
+      "loss": 0.2591,
+      "step": 9758
+    },
+    {
+      "epoch": 1.493343534812548,
+      "grad_norm": 2.018241900306442,
+      "learning_rate": 3.181637683371367e-06,
+      "loss": 0.2783,
+      "step": 9759
+    },
+    {
+      "epoch": 1.493496557000765,
+      "grad_norm": 2.220811704079774,
+      "learning_rate": 3.17982508982727e-06,
+      "loss": 0.3125,
+      "step": 9760
+    },
+    {
+      "epoch": 1.4936495791889823,
+      "grad_norm": 2.0207958924684144,
+      "learning_rate": 3.1780129151367965e-06,
+      "loss": 0.3498,
+      "step": 9761
+    },
+    {
+      "epoch": 1.4938026013771997,
+      "grad_norm": 1.9336176704001233,
+      "learning_rate": 3.176201159411232e-06,
+      "loss": 0.2944,
+      "step": 9762
+    },
+    {
+      "epoch": 1.493955623565417,
+      "grad_norm": 2.0518057739899347,
+      "learning_rate": 3.174389822761853e-06,
+      "loss": 0.3147,
+      "step": 9763
+    },
+    {
+      "epoch": 1.4941086457536343,
+      "grad_norm": 2.0524245884405903,
+      "learning_rate": 3.172578905299899e-06,
+      "loss": 0.3182,
+      "step": 9764
+    },
+    {
+      "epoch": 1.4942616679418514,
+      "grad_norm": 2.403753805984202,
+      "learning_rate": 3.1707684071365786e-06,
+      "loss": 0.3841,
+      "step": 9765
+    },
+    {
+      "epoch": 1.4944146901300688,
+      "grad_norm": 2.057508983016502,
+      "learning_rate": 3.168958328383088e-06,
+      "loss": 0.3566,
+      "step": 9766
+    },
+    {
+      "epoch": 1.4945677123182861,
+      "grad_norm": 1.9234820653276592,
+      "learning_rate": 3.1671486691505906e-06,
+      "loss": 0.2878,
+      "step": 9767
+    },
+    {
+      "epoch": 1.4947207345065034,
+      "grad_norm": 2.063565746243901,
+      "learning_rate": 3.1653394295502226e-06,
+      "loss": 0.3536,
+      "step": 9768
+    },
+    {
+      "epoch": 1.4948737566947208,
+      "grad_norm": 2.19824145243771,
+      "learning_rate": 3.1635306096930985e-06,
+      "loss": 0.3115,
+      "step": 9769
+    },
+    {
+      "epoch": 1.495026778882938,
+      "grad_norm": 2.3743711638225085,
+      "learning_rate": 3.1617222096903045e-06,
+      "loss": 0.3565,
+      "step": 9770
+    },
+    {
+      "epoch": 1.4951798010711554,
+      "grad_norm": 2.041310825172398,
+      "learning_rate": 3.1599142296529005e-06,
+      "loss": 0.3498,
+      "step": 9771
+    },
+    {
+      "epoch": 1.4953328232593726,
+      "grad_norm": 2.103580031841215,
+      "learning_rate": 3.1581066696919216e-06,
+      "loss": 0.3418,
+      "step": 9772
+    },
+    {
+      "epoch": 1.49548584544759,
+      "grad_norm": 1.6728540200328423,
+      "learning_rate": 3.1562995299183786e-06,
+      "loss": 0.293,
+      "step": 9773
+    },
+    {
+      "epoch": 1.4956388676358072,
+      "grad_norm": 2.052812844688033,
+      "learning_rate": 3.1544928104432536e-06,
+      "loss": 0.3172,
+      "step": 9774
+    },
+    {
+      "epoch": 1.4957918898240246,
+      "grad_norm": 1.99378419575446,
+      "learning_rate": 3.1526865113775063e-06,
+      "loss": 0.2531,
+      "step": 9775
+    },
+    {
+      "epoch": 1.495944912012242,
+      "grad_norm": 2.206302011302261,
+      "learning_rate": 3.1508806328320653e-06,
+      "loss": 0.3198,
+      "step": 9776
+    },
+    {
+      "epoch": 1.496097934200459,
+      "grad_norm": 2.375311640575206,
+      "learning_rate": 3.1490751749178394e-06,
+      "loss": 0.3092,
+      "step": 9777
+    },
+    {
+      "epoch": 1.4962509563886763,
+      "grad_norm": 2.3242990738348412,
+      "learning_rate": 3.1472701377457082e-06,
+      "loss": 0.3788,
+      "step": 9778
+    },
+    {
+      "epoch": 1.4964039785768937,
+      "grad_norm": 2.068015816110016,
+      "learning_rate": 3.145465521426525e-06,
+      "loss": 0.3278,
+      "step": 9779
+    },
+    {
+      "epoch": 1.496557000765111,
+      "grad_norm": 2.089112948601707,
+      "learning_rate": 3.1436613260711166e-06,
+      "loss": 0.2761,
+      "step": 9780
+    },
+    {
+      "epoch": 1.4967100229533283,
+      "grad_norm": 2.3392197952388565,
+      "learning_rate": 3.141857551790295e-06,
+      "loss": 0.3837,
+      "step": 9781
+    },
+    {
+      "epoch": 1.4968630451415454,
+      "grad_norm": 2.4626352152836493,
+      "learning_rate": 3.1400541986948274e-06,
+      "loss": 0.3356,
+      "step": 9782
+    },
+    {
+      "epoch": 1.4970160673297628,
+      "grad_norm": 1.943327658989901,
+      "learning_rate": 3.1382512668954635e-06,
+      "loss": 0.2664,
+      "step": 9783
+    },
+    {
+      "epoch": 1.4971690895179801,
+      "grad_norm": 2.0015017364893657,
+      "learning_rate": 3.13644875650294e-06,
+      "loss": 0.277,
+      "step": 9784
+    },
+    {
+      "epoch": 1.4973221117061974,
+      "grad_norm": 1.991863536015547,
+      "learning_rate": 3.1346466676279453e-06,
+      "loss": 0.3068,
+      "step": 9785
+    },
+    {
+      "epoch": 1.4974751338944148,
+      "grad_norm": 2.1289913120304527,
+      "learning_rate": 3.1328450003811526e-06,
+      "loss": 0.2519,
+      "step": 9786
+    },
+    {
+      "epoch": 1.497628156082632,
+      "grad_norm": 2.4980243023068014,
+      "learning_rate": 3.131043754873219e-06,
+      "loss": 0.3528,
+      "step": 9787
+    },
+    {
+      "epoch": 1.4977811782708492,
+      "grad_norm": 2.2767558515508273,
+      "learning_rate": 3.1292429312147542e-06,
+      "loss": 0.368,
+      "step": 9788
+    },
+    {
+      "epoch": 1.4979342004590666,
+      "grad_norm": 2.16350370343632,
+      "learning_rate": 3.127442529516362e-06,
+      "loss": 0.4144,
+      "step": 9789
+    },
+    {
+      "epoch": 1.498087222647284,
+      "grad_norm": 2.3882333717742847,
+      "learning_rate": 3.1256425498886132e-06,
+      "loss": 0.3766,
+      "step": 9790
+    },
+    {
+      "epoch": 1.4982402448355012,
+      "grad_norm": 2.1894069009771053,
+      "learning_rate": 3.123842992442042e-06,
+      "loss": 0.356,
+      "step": 9791
+    },
+    {
+      "epoch": 1.4983932670237183,
+      "grad_norm": 2.1408227407508496,
+      "learning_rate": 3.122043857287176e-06,
+      "loss": 0.2882,
+      "step": 9792
+    },
+    {
+      "epoch": 1.4985462892119357,
+      "grad_norm": 1.9419697702027892,
+      "learning_rate": 3.120245144534505e-06,
+      "loss": 0.2727,
+      "step": 9793
+    },
+    {
+      "epoch": 1.498699311400153,
+      "grad_norm": 2.0701350644107066,
+      "learning_rate": 3.118446854294488e-06,
+      "loss": 0.2933,
+      "step": 9794
+    },
+    {
+      "epoch": 1.4988523335883703,
+      "grad_norm": 1.7478356478182149,
+      "learning_rate": 3.116648986677574e-06,
+      "loss": 0.2833,
+      "step": 9795
+    },
+    {
+      "epoch": 1.4990053557765877,
+      "grad_norm": 2.0429077290330424,
+      "learning_rate": 3.114851541794173e-06,
+      "loss": 0.2941,
+      "step": 9796
+    },
+    {
+      "epoch": 1.4991583779648048,
+      "grad_norm": 2.2981563657573347,
+      "learning_rate": 3.1130545197546735e-06,
+      "loss": 0.3765,
+      "step": 9797
+    },
+    {
+      "epoch": 1.4993114001530221,
+      "grad_norm": 2.1883567916548214,
+      "learning_rate": 3.111257920669438e-06,
+      "loss": 0.3353,
+      "step": 9798
+    },
+    {
+      "epoch": 1.4994644223412394,
+      "grad_norm": 2.0688655353997576,
+      "learning_rate": 3.109461744648803e-06,
+      "loss": 0.2819,
+      "step": 9799
+    },
+    {
+      "epoch": 1.4996174445294568,
+      "grad_norm": 2.1705196804072133,
+      "learning_rate": 3.107665991803078e-06,
+      "loss": 0.368,
+      "step": 9800
+    },
+    {
+      "epoch": 1.4997704667176741,
+      "grad_norm": 2.2061746818166714,
+      "learning_rate": 3.1058706622425473e-06,
+      "loss": 0.3808,
+      "step": 9801
+    },
+    {
+      "epoch": 1.4999234889058912,
+      "grad_norm": 2.054827111884937,
+      "learning_rate": 3.1040757560774694e-06,
+      "loss": 0.3585,
+      "step": 9802
+    },
+    {
+      "epoch": 1.5000765110941088,
+      "grad_norm": 2.121121071144255,
+      "learning_rate": 3.1022812734180765e-06,
+      "loss": 0.274,
+      "step": 9803
+    },
+    {
+      "epoch": 1.5002295332823259,
+      "grad_norm": 2.080239488774426,
+      "learning_rate": 3.1004872143745734e-06,
+      "loss": 0.2902,
+      "step": 9804
+    },
+    {
+      "epoch": 1.5003825554705432,
+      "grad_norm": 2.049230990374885,
+      "learning_rate": 3.0986935790571427e-06,
+      "loss": 0.3381,
+      "step": 9805
+    },
+    {
+      "epoch": 1.5005355776587606,
+      "grad_norm": 2.2067647222870383,
+      "learning_rate": 3.0969003675759368e-06,
+      "loss": 0.3135,
+      "step": 9806
+    },
+    {
+      "epoch": 1.5006885998469777,
+      "grad_norm": 1.9066668807836955,
+      "learning_rate": 3.0951075800410847e-06,
+      "loss": 0.4619,
+      "step": 9807
+    },
+    {
+      "epoch": 1.5008416220351952,
+      "grad_norm": 2.2749545171097,
+      "learning_rate": 3.093315216562688e-06,
+      "loss": 0.3464,
+      "step": 9808
+    },
+    {
+      "epoch": 1.5009946442234123,
+      "grad_norm": 2.0411962491402984,
+      "learning_rate": 3.0915232772508196e-06,
+      "loss": 0.2835,
+      "step": 9809
+    },
+    {
+      "epoch": 1.5011476664116297,
+      "grad_norm": 2.191472037931345,
+      "learning_rate": 3.08973176221554e-06,
+      "loss": 0.2974,
+      "step": 9810
+    },
+    {
+      "epoch": 1.501300688599847,
+      "grad_norm": 2.1908312969139843,
+      "learning_rate": 3.0879406715668623e-06,
+      "loss": 0.3244,
+      "step": 9811
+    },
+    {
+      "epoch": 1.501453710788064,
+      "grad_norm": 2.136787196257136,
+      "learning_rate": 3.086150005414784e-06,
+      "loss": 0.3084,
+      "step": 9812
+    },
+    {
+      "epoch": 1.5016067329762817,
+      "grad_norm": 1.894927128405544,
+      "learning_rate": 3.084359763869289e-06,
+      "loss": 0.3001,
+      "step": 9813
+    },
+    {
+      "epoch": 1.5017597551644988,
+      "grad_norm": 2.1745865419199646,
+      "learning_rate": 3.0825699470403114e-06,
+      "loss": 0.3858,
+      "step": 9814
+    },
+    {
+      "epoch": 1.501912777352716,
+      "grad_norm": 1.9608382991991724,
+      "learning_rate": 3.080780555037771e-06,
+      "loss": 0.3078,
+      "step": 9815
+    },
+    {
+      "epoch": 1.5020657995409334,
+      "grad_norm": 2.1040554688231343,
+      "learning_rate": 3.0789915879715715e-06,
+      "loss": 0.3482,
+      "step": 9816
+    },
+    {
+      "epoch": 1.5022188217291508,
+      "grad_norm": 1.9905730959553891,
+      "learning_rate": 3.077203045951567e-06,
+      "loss": 0.277,
+      "step": 9817
+    },
+    {
+      "epoch": 1.502371843917368,
+      "grad_norm": 2.1890814408914747,
+      "learning_rate": 3.075414929087609e-06,
+      "loss": 0.2968,
+      "step": 9818
+    },
+    {
+      "epoch": 1.5025248661055852,
+      "grad_norm": 2.495997552639193,
+      "learning_rate": 3.0736272374895115e-06,
+      "loss": 0.2815,
+      "step": 9819
+    },
+    {
+      "epoch": 1.5026778882938026,
+      "grad_norm": 2.0086629122191604,
+      "learning_rate": 3.0718399712670566e-06,
+      "loss": 0.2857,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5028309104820199,
+      "grad_norm": 2.2914428417288217,
+      "learning_rate": 3.0700531305300153e-06,
+      "loss": 0.3293,
+      "step": 9821
+    },
+    {
+      "epoch": 1.5029839326702372,
+      "grad_norm": 2.1358875597800973,
+      "learning_rate": 3.0682667153881228e-06,
+      "loss": 0.273,
+      "step": 9822
+    },
+    {
+      "epoch": 1.5031369548584546,
+      "grad_norm": 1.958831411297443,
+      "learning_rate": 3.0664807259510842e-06,
+      "loss": 0.3609,
+      "step": 9823
+    },
+    {
+      "epoch": 1.5032899770466717,
+      "grad_norm": 2.0320226227226743,
+      "learning_rate": 3.0646951623285904e-06,
+      "loss": 0.2767,
+      "step": 9824
+    },
+    {
+      "epoch": 1.5034429992348892,
+      "grad_norm": 2.2398279587072087,
+      "learning_rate": 3.062910024630298e-06,
+      "loss": 0.3316,
+      "step": 9825
+    },
+    {
+      "epoch": 1.5035960214231063,
+      "grad_norm": 2.0793764036363735,
+      "learning_rate": 3.061125312965838e-06,
+      "loss": 0.2947,
+      "step": 9826
+    },
+    {
+      "epoch": 1.5037490436113237,
+      "grad_norm": 2.350717021248169,
+      "learning_rate": 3.059341027444819e-06,
+      "loss": 0.3965,
+      "step": 9827
+    },
+    {
+      "epoch": 1.503902065799541,
+      "grad_norm": 2.0080218668074705,
+      "learning_rate": 3.057557168176819e-06,
+      "loss": 0.4766,
+      "step": 9828
+    },
+    {
+      "epoch": 1.504055087987758,
+      "grad_norm": 2.192980069727335,
+      "learning_rate": 3.0557737352713925e-06,
+      "loss": 0.3138,
+      "step": 9829
+    },
+    {
+      "epoch": 1.5042081101759757,
+      "grad_norm": 2.277627119535517,
+      "learning_rate": 3.0539907288380664e-06,
+      "loss": 0.271,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5043611323641928,
+      "grad_norm": 2.060097566540218,
+      "learning_rate": 3.0522081489863433e-06,
+      "loss": 0.2744,
+      "step": 9831
+    },
+    {
+      "epoch": 1.50451415455241,
+      "grad_norm": 2.1953025152712855,
+      "learning_rate": 3.0504259958256966e-06,
+      "loss": 0.2951,
+      "step": 9832
+    },
+    {
+      "epoch": 1.5046671767406274,
+      "grad_norm": 2.0592054511449187,
+      "learning_rate": 3.0486442694655737e-06,
+      "loss": 0.3091,
+      "step": 9833
+    },
+    {
+      "epoch": 1.5048201989288446,
+      "grad_norm": 2.3291429300803124,
+      "learning_rate": 3.0468629700154072e-06,
+      "loss": 0.3384,
+      "step": 9834
+    },
+    {
+      "epoch": 1.504973221117062,
+      "grad_norm": 2.1766038192419943,
+      "learning_rate": 3.045082097584583e-06,
+      "loss": 0.3497,
+      "step": 9835
+    },
+    {
+      "epoch": 1.5051262433052792,
+      "grad_norm": 2.269668303049232,
+      "learning_rate": 3.043301652282471e-06,
+      "loss": 0.3362,
+      "step": 9836
+    },
+    {
+      "epoch": 1.5052792654934966,
+      "grad_norm": 2.488874865174472,
+      "learning_rate": 3.041521634218426e-06,
+      "loss": 0.3862,
+      "step": 9837
+    },
+    {
+      "epoch": 1.5054322876817139,
+      "grad_norm": 1.9458045410505225,
+      "learning_rate": 3.0397420435017565e-06,
+      "loss": 0.2328,
+      "step": 9838
+    },
+    {
+      "epoch": 1.505585309869931,
+      "grad_norm": 2.125888170369278,
+      "learning_rate": 3.0379628802417525e-06,
+      "loss": 0.2961,
+      "step": 9839
+    },
+    {
+      "epoch": 1.5057383320581486,
+      "grad_norm": 2.210814431359663,
+      "learning_rate": 3.0361841445476914e-06,
+      "loss": 0.3397,
+      "step": 9840
+    },
+    {
+      "epoch": 1.5058913542463657,
+      "grad_norm": 2.1884877256421937,
+      "learning_rate": 3.0344058365287977e-06,
+      "loss": 0.3376,
+      "step": 9841
+    },
+    {
+      "epoch": 1.506044376434583,
+      "grad_norm": 2.419550996918986,
+      "learning_rate": 3.032627956294294e-06,
+      "loss": 0.3823,
+      "step": 9842
+    },
+    {
+      "epoch": 1.5061973986228003,
+      "grad_norm": 2.299565953799042,
+      "learning_rate": 3.0308505039533675e-06,
+      "loss": 0.3659,
+      "step": 9843
+    },
+    {
+      "epoch": 1.5063504208110174,
+      "grad_norm": 2.213477786477071,
+      "learning_rate": 3.0290734796151687e-06,
+      "loss": 0.3586,
+      "step": 9844
+    },
+    {
+      "epoch": 1.506503442999235,
+      "grad_norm": 2.1247224716904185,
+      "learning_rate": 3.0272968833888407e-06,
+      "loss": 0.3304,
+      "step": 9845
+    },
+    {
+      "epoch": 1.506656465187452,
+      "grad_norm": 2.1254476906607307,
+      "learning_rate": 3.0255207153834886e-06,
+      "loss": 0.2704,
+      "step": 9846
+    },
+    {
+      "epoch": 1.5068094873756694,
+      "grad_norm": 2.436520978872859,
+      "learning_rate": 3.023744975708194e-06,
+      "loss": 0.3755,
+      "step": 9847
+    },
+    {
+      "epoch": 1.5069625095638868,
+      "grad_norm": 2.2069707007141095,
+      "learning_rate": 3.021969664472012e-06,
+      "loss": 0.2866,
+      "step": 9848
+    },
+    {
+      "epoch": 1.507115531752104,
+      "grad_norm": 1.9306592406380063,
+      "learning_rate": 3.0201947817839704e-06,
+      "loss": 0.3086,
+      "step": 9849
+    },
+    {
+      "epoch": 1.5072685539403214,
+      "grad_norm": 2.26216940279369,
+      "learning_rate": 3.0184203277530723e-06,
+      "loss": 0.38,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5074215761285386,
+      "grad_norm": 2.0832013755845,
+      "learning_rate": 3.0166463024882943e-06,
+      "loss": 0.3438,
+      "step": 9851
+    },
+    {
+      "epoch": 1.5075745983167559,
+      "grad_norm": 2.2765780863751166,
+      "learning_rate": 3.014872706098586e-06,
+      "loss": 0.3067,
+      "step": 9852
+    },
+    {
+      "epoch": 1.5077276205049732,
+      "grad_norm": 2.5699472822824223,
+      "learning_rate": 3.0130995386928706e-06,
+      "loss": 0.3503,
+      "step": 9853
+    },
+    {
+      "epoch": 1.5078806426931906,
+      "grad_norm": 2.005607019002461,
+      "learning_rate": 3.0113268003800456e-06,
+      "loss": 0.2863,
+      "step": 9854
+    },
+    {
+      "epoch": 1.5080336648814079,
+      "grad_norm": 1.9684914601471286,
+      "learning_rate": 3.0095544912689822e-06,
+      "loss": 0.2837,
+      "step": 9855
+    },
+    {
+      "epoch": 1.508186687069625,
+      "grad_norm": 1.7614254445638609,
+      "learning_rate": 3.007782611468524e-06,
+      "loss": 0.3154,
+      "step": 9856
+    },
+    {
+      "epoch": 1.5083397092578426,
+      "grad_norm": 2.0535358999385505,
+      "learning_rate": 3.0060111610874886e-06,
+      "loss": 0.2978,
+      "step": 9857
+    },
+    {
+      "epoch": 1.5084927314460597,
+      "grad_norm": 2.2628065523811456,
+      "learning_rate": 3.0042401402346687e-06,
+      "loss": 0.3182,
+      "step": 9858
+    },
+    {
+      "epoch": 1.508645753634277,
+      "grad_norm": 1.9545007110492567,
+      "learning_rate": 3.0024695490188296e-06,
+      "loss": 0.3188,
+      "step": 9859
+    },
+    {
+      "epoch": 1.5087987758224943,
+      "grad_norm": 2.1848377979596627,
+      "learning_rate": 3.00069938754871e-06,
+      "loss": 0.3695,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5089517980107114,
+      "grad_norm": 2.1635874874215633,
+      "learning_rate": 2.9989296559330215e-06,
+      "loss": 0.3139,
+      "step": 9861
+    },
+    {
+      "epoch": 1.509104820198929,
+      "grad_norm": 2.338481731423221,
+      "learning_rate": 2.9971603542804495e-06,
+      "loss": 0.3425,
+      "step": 9862
+    },
+    {
+      "epoch": 1.509257842387146,
+      "grad_norm": 2.2282324761992554,
+      "learning_rate": 2.995391482699661e-06,
+      "loss": 0.3796,
+      "step": 9863
+    },
+    {
+      "epoch": 1.5094108645753634,
+      "grad_norm": 2.276627350922109,
+      "learning_rate": 2.9936230412992805e-06,
+      "loss": 0.2884,
+      "step": 9864
+    },
+    {
+      "epoch": 1.5095638867635808,
+      "grad_norm": 1.9414960546579652,
+      "learning_rate": 2.9918550301879145e-06,
+      "loss": 0.336,
+      "step": 9865
+    },
+    {
+      "epoch": 1.5097169089517979,
+      "grad_norm": 2.1301779835444377,
+      "learning_rate": 2.990087449474154e-06,
+      "loss": 0.3303,
+      "step": 9866
+    },
+    {
+      "epoch": 1.5098699311400154,
+      "grad_norm": 1.9945124667754917,
+      "learning_rate": 2.9883202992665438e-06,
+      "loss": 0.2838,
+      "step": 9867
+    },
+    {
+      "epoch": 1.5100229533282326,
+      "grad_norm": 2.2782006257588474,
+      "learning_rate": 2.986553579673609e-06,
+      "loss": 0.3139,
+      "step": 9868
+    },
+    {
+      "epoch": 1.5101759755164499,
+      "grad_norm": 1.9001158252077175,
+      "learning_rate": 2.984787290803863e-06,
+      "loss": 0.279,
+      "step": 9869
+    },
+    {
+      "epoch": 1.5103289977046672,
+      "grad_norm": 1.9890668141893428,
+      "learning_rate": 2.983021432765767e-06,
+      "loss": 0.3103,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5104820198928843,
+      "grad_norm": 1.9261758653878254,
+      "learning_rate": 2.9812560056677785e-06,
+      "loss": 0.2946,
+      "step": 9871
+    },
+    {
+      "epoch": 1.5106350420811019,
+      "grad_norm": 2.2334108039049054,
+      "learning_rate": 2.9794910096183183e-06,
+      "loss": 0.3224,
+      "step": 9872
+    },
+    {
+      "epoch": 1.510788064269319,
+      "grad_norm": 2.1237274490949276,
+      "learning_rate": 2.9777264447257748e-06,
+      "loss": 0.3449,
+      "step": 9873
+    },
+    {
+      "epoch": 1.5109410864575363,
+      "grad_norm": 2.1770374643161667,
+      "learning_rate": 2.9759623110985236e-06,
+      "loss": 0.3317,
+      "step": 9874
+    },
+    {
+      "epoch": 1.5110941086457537,
+      "grad_norm": 2.4761340756648473,
+      "learning_rate": 2.974198608844906e-06,
+      "loss": 0.3783,
+      "step": 9875
+    },
+    {
+      "epoch": 1.5112471308339708,
+      "grad_norm": 2.249465172824092,
+      "learning_rate": 2.9724353380732364e-06,
+      "loss": 0.4108,
+      "step": 9876
+    },
+    {
+      "epoch": 1.5114001530221883,
+      "grad_norm": 2.1262704705662143,
+      "learning_rate": 2.9706724988918043e-06,
+      "loss": 0.2787,
+      "step": 9877
+    },
+    {
+      "epoch": 1.5115531752104054,
+      "grad_norm": 2.608599424875475,
+      "learning_rate": 2.968910091408873e-06,
+      "loss": 0.3657,
+      "step": 9878
+    },
+    {
+      "epoch": 1.5117061973986228,
+      "grad_norm": 2.0937507208713497,
+      "learning_rate": 2.9671481157326785e-06,
+      "loss": 0.273,
+      "step": 9879
+    },
+    {
+      "epoch": 1.51185921958684,
+      "grad_norm": 1.946594115741587,
+      "learning_rate": 2.9653865719714326e-06,
+      "loss": 0.2656,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5120122417750572,
+      "grad_norm": 1.962879276420753,
+      "learning_rate": 2.9636254602333147e-06,
+      "loss": 0.2833,
+      "step": 9881
+    },
+    {
+      "epoch": 1.5121652639632748,
+      "grad_norm": 2.6116433052824033,
+      "learning_rate": 2.9618647806264856e-06,
+      "loss": 0.315,
+      "step": 9882
+    },
+    {
+      "epoch": 1.5123182861514919,
+      "grad_norm": 2.350676279516989,
+      "learning_rate": 2.9601045332590727e-06,
+      "loss": 0.3421,
+      "step": 9883
+    },
+    {
+      "epoch": 1.5124713083397092,
+      "grad_norm": 1.989790085557811,
+      "learning_rate": 2.9583447182391813e-06,
+      "loss": 0.3153,
+      "step": 9884
+    },
+    {
+      "epoch": 1.5126243305279266,
+      "grad_norm": 1.7572196885042126,
+      "learning_rate": 2.9565853356748863e-06,
+      "loss": 0.2197,
+      "step": 9885
+    },
+    {
+      "epoch": 1.5127773527161439,
+      "grad_norm": 2.297276054650986,
+      "learning_rate": 2.954826385674241e-06,
+      "loss": 0.3149,
+      "step": 9886
+    },
+    {
+      "epoch": 1.5129303749043612,
+      "grad_norm": 2.0684649483413184,
+      "learning_rate": 2.9530678683452685e-06,
+      "loss": 0.2807,
+      "step": 9887
+    },
+    {
+      "epoch": 1.5130833970925783,
+      "grad_norm": 2.083082465518758,
+      "learning_rate": 2.951309783795965e-06,
+      "loss": 0.3097,
+      "step": 9888
+    },
+    {
+      "epoch": 1.5132364192807957,
+      "grad_norm": 2.232352667798138,
+      "learning_rate": 2.9495521321343035e-06,
+      "loss": 0.2664,
+      "step": 9889
+    },
+    {
+      "epoch": 1.513389441469013,
+      "grad_norm": 2.0746473923615874,
+      "learning_rate": 2.947794913468226e-06,
+      "loss": 0.3053,
+      "step": 9890
+    },
+    {
+      "epoch": 1.5135424636572303,
+      "grad_norm": 2.248018803212016,
+      "learning_rate": 2.9460381279056482e-06,
+      "loss": 0.285,
+      "step": 9891
+    },
+    {
+      "epoch": 1.5136954858454477,
+      "grad_norm": 2.091935846460022,
+      "learning_rate": 2.9442817755544707e-06,
+      "loss": 0.2858,
+      "step": 9892
+    },
+    {
+      "epoch": 1.5138485080336648,
+      "grad_norm": 1.9566609105598456,
+      "learning_rate": 2.942525856522549e-06,
+      "loss": 0.3297,
+      "step": 9893
+    },
+    {
+      "epoch": 1.5140015302218823,
+      "grad_norm": 2.137727077428057,
+      "learning_rate": 2.9407703709177193e-06,
+      "loss": 0.2879,
+      "step": 9894
+    },
+    {
+      "epoch": 1.5141545524100994,
+      "grad_norm": 2.2093864848262124,
+      "learning_rate": 2.939015318847804e-06,
+      "loss": 0.2882,
+      "step": 9895
+    },
+    {
+      "epoch": 1.5143075745983168,
+      "grad_norm": 2.095268692569443,
+      "learning_rate": 2.9372607004205777e-06,
+      "loss": 0.3523,
+      "step": 9896
+    },
+    {
+      "epoch": 1.514460596786534,
+      "grad_norm": 2.4814418845151143,
+      "learning_rate": 2.935506515743798e-06,
+      "loss": 0.3139,
+      "step": 9897
+    },
+    {
+      "epoch": 1.5146136189747512,
+      "grad_norm": 1.987799327564212,
+      "learning_rate": 2.933752764925205e-06,
+      "loss": 0.2749,
+      "step": 9898
+    },
+    {
+      "epoch": 1.5147666411629688,
+      "grad_norm": 2.2456331156220917,
+      "learning_rate": 2.931999448072492e-06,
+      "loss": 0.3214,
+      "step": 9899
+    },
+    {
+      "epoch": 1.5149196633511859,
+      "grad_norm": 2.0249520005224007,
+      "learning_rate": 2.9302465652933476e-06,
+      "loss": 0.2876,
+      "step": 9900
+    },
+    {
+      "epoch": 1.5150726855394032,
+      "grad_norm": 1.9927759375396923,
+      "learning_rate": 2.9284941166954207e-06,
+      "loss": 0.2752,
+      "step": 9901
+    },
+    {
+      "epoch": 1.5152257077276206,
+      "grad_norm": 1.8998708876323396,
+      "learning_rate": 2.9267421023863274e-06,
+      "loss": 0.3092,
+      "step": 9902
+    },
+    {
+      "epoch": 1.5153787299158377,
+      "grad_norm": 2.1212028678752803,
+      "learning_rate": 2.924990522473676e-06,
+      "loss": 0.3021,
+      "step": 9903
+    },
+    {
+      "epoch": 1.5155317521040552,
+      "grad_norm": 1.9764806188380462,
+      "learning_rate": 2.923239377065038e-06,
+      "loss": 0.2792,
+      "step": 9904
+    },
+    {
+      "epoch": 1.5156847742922723,
+      "grad_norm": 1.919797871945186,
+      "learning_rate": 2.9214886662679467e-06,
+      "loss": 0.2539,
+      "step": 9905
+    },
+    {
+      "epoch": 1.5158377964804897,
+      "grad_norm": 1.7930166385110369,
+      "learning_rate": 2.9197383901899314e-06,
+      "loss": 0.27,
+      "step": 9906
+    },
+    {
+      "epoch": 1.515990818668707,
+      "grad_norm": 1.9817109015167513,
+      "learning_rate": 2.9179885489384797e-06,
+      "loss": 0.2365,
+      "step": 9907
+    },
+    {
+      "epoch": 1.516143840856924,
+      "grad_norm": 2.177398279153643,
+      "learning_rate": 2.916239142621057e-06,
+      "loss": 0.2703,
+      "step": 9908
+    },
+    {
+      "epoch": 1.5162968630451417,
+      "grad_norm": 2.1794166380623454,
+      "learning_rate": 2.914490171345099e-06,
+      "loss": 0.3407,
+      "step": 9909
+    },
+    {
+      "epoch": 1.5164498852333588,
+      "grad_norm": 2.40714069176058,
+      "learning_rate": 2.9127416352180195e-06,
+      "loss": 0.3685,
+      "step": 9910
+    },
+    {
+      "epoch": 1.516602907421576,
+      "grad_norm": 1.8834924452301767,
+      "learning_rate": 2.9109935343472007e-06,
+      "loss": 0.3248,
+      "step": 9911
+    },
+    {
+      "epoch": 1.5167559296097934,
+      "grad_norm": 2.1304268074488304,
+      "learning_rate": 2.9092458688399983e-06,
+      "loss": 0.2673,
+      "step": 9912
+    },
+    {
+      "epoch": 1.5169089517980106,
+      "grad_norm": 2.299973146784591,
+      "learning_rate": 2.907498638803753e-06,
+      "loss": 0.3088,
+      "step": 9913
+    },
+    {
+      "epoch": 1.517061973986228,
+      "grad_norm": 2.0890436295813135,
+      "learning_rate": 2.905751844345759e-06,
+      "loss": 0.2982,
+      "step": 9914
+    },
+    {
+      "epoch": 1.5172149961744452,
+      "grad_norm": 2.243008873416047,
+      "learning_rate": 2.9040054855732934e-06,
+      "loss": 0.3665,
+      "step": 9915
+    },
+    {
+      "epoch": 1.5173680183626626,
+      "grad_norm": 2.231300366616912,
+      "learning_rate": 2.902259562593618e-06,
+      "loss": 0.3656,
+      "step": 9916
+    },
+    {
+      "epoch": 1.5175210405508799,
+      "grad_norm": 2.314260763210238,
+      "learning_rate": 2.900514075513945e-06,
+      "loss": 0.3373,
+      "step": 9917
+    },
+    {
+      "epoch": 1.5176740627390972,
+      "grad_norm": 2.2686857788300676,
+      "learning_rate": 2.8987690244414735e-06,
+      "loss": 0.3182,
+      "step": 9918
+    },
+    {
+      "epoch": 1.5178270849273146,
+      "grad_norm": 2.0332689531706256,
+      "learning_rate": 2.8970244094833834e-06,
+      "loss": 0.2877,
+      "step": 9919
+    },
+    {
+      "epoch": 1.5179801071155317,
+      "grad_norm": 2.332225199392639,
+      "learning_rate": 2.895280230746804e-06,
+      "loss": 0.3246,
+      "step": 9920
+    },
+    {
+      "epoch": 1.518133129303749,
+      "grad_norm": 2.213902332927608,
+      "learning_rate": 2.893536488338863e-06,
+      "loss": 0.2677,
+      "step": 9921
+    },
+    {
+      "epoch": 1.5182861514919663,
+      "grad_norm": 2.058077346563243,
+      "learning_rate": 2.89179318236665e-06,
+      "loss": 0.3128,
+      "step": 9922
+    },
+    {
+      "epoch": 1.5184391736801837,
+      "grad_norm": 2.1206458507799684,
+      "learning_rate": 2.890050312937218e-06,
+      "loss": 0.2985,
+      "step": 9923
+    },
+    {
+      "epoch": 1.518592195868401,
+      "grad_norm": 2.5255851317820666,
+      "learning_rate": 2.8883078801576148e-06,
+      "loss": 0.3492,
+      "step": 9924
+    },
+    {
+      "epoch": 1.518745218056618,
+      "grad_norm": 2.239478346754188,
+      "learning_rate": 2.886565884134849e-06,
+      "loss": 0.3186,
+      "step": 9925
+    },
+    {
+      "epoch": 1.5188982402448357,
+      "grad_norm": 2.1589701070470655,
+      "learning_rate": 2.8848243249758934e-06,
+      "loss": 0.2861,
+      "step": 9926
+    },
+    {
+      "epoch": 1.5190512624330528,
+      "grad_norm": 2.038957431232218,
+      "learning_rate": 2.8830832027877134e-06,
+      "loss": 0.3916,
+      "step": 9927
+    },
+    {
+      "epoch": 1.51920428462127,
+      "grad_norm": 2.290410983887635,
+      "learning_rate": 2.8813425176772357e-06,
+      "loss": 0.3217,
+      "step": 9928
+    },
+    {
+      "epoch": 1.5193573068094874,
+      "grad_norm": 2.1377659989683018,
+      "learning_rate": 2.8796022697513627e-06,
+      "loss": 0.3644,
+      "step": 9929
+    },
+    {
+      "epoch": 1.5195103289977046,
+      "grad_norm": 2.1557974796471977,
+      "learning_rate": 2.87786245911697e-06,
+      "loss": 0.3259,
+      "step": 9930
+    },
+    {
+      "epoch": 1.519663351185922,
+      "grad_norm": 2.197537522420924,
+      "learning_rate": 2.876123085880904e-06,
+      "loss": 0.3123,
+      "step": 9931
+    },
+    {
+      "epoch": 1.5198163733741392,
+      "grad_norm": 1.8849797423696186,
+      "learning_rate": 2.874384150149989e-06,
+      "loss": 0.2838,
+      "step": 9932
+    },
+    {
+      "epoch": 1.5199693955623566,
+      "grad_norm": 2.355866028332641,
+      "learning_rate": 2.87264565203102e-06,
+      "loss": 0.3106,
+      "step": 9933
+    },
+    {
+      "epoch": 1.5201224177505739,
+      "grad_norm": 2.4298439369368086,
+      "learning_rate": 2.8709075916307626e-06,
+      "loss": 0.3564,
+      "step": 9934
+    },
+    {
+      "epoch": 1.520275439938791,
+      "grad_norm": 2.138768183921702,
+      "learning_rate": 2.8691699690559594e-06,
+      "loss": 0.3101,
+      "step": 9935
+    },
+    {
+      "epoch": 1.5204284621270086,
+      "grad_norm": 2.1084569658813064,
+      "learning_rate": 2.8674327844133243e-06,
+      "loss": 0.3014,
+      "step": 9936
+    },
+    {
+      "epoch": 1.5205814843152257,
+      "grad_norm": 1.894295714318006,
+      "learning_rate": 2.8656960378095444e-06,
+      "loss": 0.2854,
+      "step": 9937
+    },
+    {
+      "epoch": 1.520734506503443,
+      "grad_norm": 2.328776255293014,
+      "learning_rate": 2.863959729351281e-06,
+      "loss": 0.3072,
+      "step": 9938
+    },
+    {
+      "epoch": 1.5208875286916603,
+      "grad_norm": 2.4823991572727184,
+      "learning_rate": 2.8622238591451668e-06,
+      "loss": 0.3271,
+      "step": 9939
+    },
+    {
+      "epoch": 1.5210405508798774,
+      "grad_norm": 2.054904895081333,
+      "learning_rate": 2.8604884272978083e-06,
+      "loss": 0.2628,
+      "step": 9940
+    },
+    {
+      "epoch": 1.521193573068095,
+      "grad_norm": 2.0507676713357705,
+      "learning_rate": 2.8587534339157854e-06,
+      "loss": 0.3012,
+      "step": 9941
+    },
+    {
+      "epoch": 1.521346595256312,
+      "grad_norm": 2.174337355448324,
+      "learning_rate": 2.8570188791056507e-06,
+      "loss": 0.3171,
+      "step": 9942
+    },
+    {
+      "epoch": 1.5214996174445294,
+      "grad_norm": 1.8669974265756857,
+      "learning_rate": 2.855284762973931e-06,
+      "loss": 0.2611,
+      "step": 9943
+    },
+    {
+      "epoch": 1.5216526396327468,
+      "grad_norm": 2.063747456601812,
+      "learning_rate": 2.8535510856271207e-06,
+      "loss": 0.3354,
+      "step": 9944
+    },
+    {
+      "epoch": 1.5218056618209639,
+      "grad_norm": 2.4098971885429474,
+      "learning_rate": 2.8518178471717008e-06,
+      "loss": 0.3642,
+      "step": 9945
+    },
+    {
+      "epoch": 1.5219586840091814,
+      "grad_norm": 2.3625660024538857,
+      "learning_rate": 2.8500850477141086e-06,
+      "loss": 0.3025,
+      "step": 9946
+    },
+    {
+      "epoch": 1.5221117061973986,
+      "grad_norm": 2.3558171194167774,
+      "learning_rate": 2.8483526873607603e-06,
+      "loss": 0.3086,
+      "step": 9947
+    },
+    {
+      "epoch": 1.5222647283856159,
+      "grad_norm": 2.101587584902661,
+      "learning_rate": 2.8466207662180578e-06,
+      "loss": 0.2378,
+      "step": 9948
+    },
+    {
+      "epoch": 1.5224177505738332,
+      "grad_norm": 2.147038880991046,
+      "learning_rate": 2.8448892843923513e-06,
+      "loss": 0.3382,
+      "step": 9949
+    },
+    {
+      "epoch": 1.5225707727620506,
+      "grad_norm": 2.1708700692575142,
+      "learning_rate": 2.8431582419899894e-06,
+      "loss": 0.3196,
+      "step": 9950
+    },
+    {
+      "epoch": 1.5227237949502679,
+      "grad_norm": 2.2319621602439548,
+      "learning_rate": 2.841427639117279e-06,
+      "loss": 0.2803,
+      "step": 9951
+    },
+    {
+      "epoch": 1.522876817138485,
+      "grad_norm": 1.840557601791769,
+      "learning_rate": 2.839697475880496e-06,
+      "loss": 0.2953,
+      "step": 9952
+    },
+    {
+      "epoch": 1.5230298393267023,
+      "grad_norm": 1.9537839576314175,
+      "learning_rate": 2.837967752385906e-06,
+      "loss": 0.3168,
+      "step": 9953
+    },
+    {
+      "epoch": 1.5231828615149197,
+      "grad_norm": 1.9900787991732978,
+      "learning_rate": 2.836238468739737e-06,
+      "loss": 0.3043,
+      "step": 9954
+    },
+    {
+      "epoch": 1.523335883703137,
+      "grad_norm": 2.0738800146254412,
+      "learning_rate": 2.834509625048182e-06,
+      "loss": 0.2901,
+      "step": 9955
+    },
+    {
+      "epoch": 1.5234889058913543,
+      "grad_norm": 2.270382910441135,
+      "learning_rate": 2.8327812214174265e-06,
+      "loss": 0.301,
+      "step": 9956
+    },
+    {
+      "epoch": 1.5236419280795714,
+      "grad_norm": 1.9818118098041126,
+      "learning_rate": 2.831053257953613e-06,
+      "loss": 0.2977,
+      "step": 9957
+    },
+    {
+      "epoch": 1.523794950267789,
+      "grad_norm": 2.233450766833048,
+      "learning_rate": 2.8293257347628655e-06,
+      "loss": 0.3289,
+      "step": 9958
+    },
+    {
+      "epoch": 1.523947972456006,
+      "grad_norm": 2.298257535897543,
+      "learning_rate": 2.8275986519512753e-06,
+      "loss": 0.3549,
+      "step": 9959
+    },
+    {
+      "epoch": 1.5241009946442234,
+      "grad_norm": 2.2911297653055316,
+      "learning_rate": 2.8258720096249116e-06,
+      "loss": 0.3306,
+      "step": 9960
+    },
+    {
+      "epoch": 1.5242540168324408,
+      "grad_norm": 2.0928809609607075,
+      "learning_rate": 2.824145807889812e-06,
+      "loss": 0.3591,
+      "step": 9961
+    },
+    {
+      "epoch": 1.5244070390206579,
+      "grad_norm": 2.085267934794212,
+      "learning_rate": 2.8224200468519914e-06,
+      "loss": 0.2872,
+      "step": 9962
+    },
+    {
+      "epoch": 1.5245600612088754,
+      "grad_norm": 2.441486160019589,
+      "learning_rate": 2.8206947266174346e-06,
+      "loss": 0.3302,
+      "step": 9963
+    },
+    {
+      "epoch": 1.5247130833970926,
+      "grad_norm": 2.1910546788096297,
+      "learning_rate": 2.818969847292099e-06,
+      "loss": 0.2417,
+      "step": 9964
+    },
+    {
+      "epoch": 1.5248661055853099,
+      "grad_norm": 2.224279279401526,
+      "learning_rate": 2.817245408981919e-06,
+      "loss": 0.2423,
+      "step": 9965
+    },
+    {
+      "epoch": 1.5250191277735272,
+      "grad_norm": 2.1629445091645914,
+      "learning_rate": 2.8155214117927964e-06,
+      "loss": 0.3308,
+      "step": 9966
+    },
+    {
+      "epoch": 1.5251721499617443,
+      "grad_norm": 1.8555635506479633,
+      "learning_rate": 2.81379785583061e-06,
+      "loss": 0.2878,
+      "step": 9967
+    },
+    {
+      "epoch": 1.5253251721499619,
+      "grad_norm": 1.9671814004381067,
+      "learning_rate": 2.8120747412012096e-06,
+      "loss": 0.285,
+      "step": 9968
+    },
+    {
+      "epoch": 1.525478194338179,
+      "grad_norm": 2.16675648292726,
+      "learning_rate": 2.810352068010419e-06,
+      "loss": 0.3092,
+      "step": 9969
+    },
+    {
+      "epoch": 1.5256312165263963,
+      "grad_norm": 2.2905687791113416,
+      "learning_rate": 2.808629836364034e-06,
+      "loss": 0.3053,
+      "step": 9970
+    },
+    {
+      "epoch": 1.5257842387146137,
+      "grad_norm": 1.9946227786642385,
+      "learning_rate": 2.806908046367823e-06,
+      "loss": 0.2843,
+      "step": 9971
+    },
+    {
+      "epoch": 1.5259372609028308,
+      "grad_norm": 2.174909387644281,
+      "learning_rate": 2.8051866981275298e-06,
+      "loss": 0.3053,
+      "step": 9972
+    },
+    {
+      "epoch": 1.5260902830910483,
+      "grad_norm": 2.359615508272599,
+      "learning_rate": 2.803465791748864e-06,
+      "loss": 0.281,
+      "step": 9973
+    },
+    {
+      "epoch": 1.5262433052792654,
+      "grad_norm": 2.2931829206578898,
+      "learning_rate": 2.8017453273375227e-06,
+      "loss": 0.3119,
+      "step": 9974
+    },
+    {
+      "epoch": 1.5263963274674828,
+      "grad_norm": 2.2215604585576694,
+      "learning_rate": 2.8000253049991577e-06,
+      "loss": 0.3205,
+      "step": 9975
+    },
+    {
+      "epoch": 1.5265493496557,
+      "grad_norm": 2.576249260369154,
+      "learning_rate": 2.798305724839402e-06,
+      "loss": 0.351,
+      "step": 9976
+    },
+    {
+      "epoch": 1.5267023718439172,
+      "grad_norm": 2.59817822012346,
+      "learning_rate": 2.796586586963871e-06,
+      "loss": 0.3171,
+      "step": 9977
+    },
+    {
+      "epoch": 1.5268553940321348,
+      "grad_norm": 2.8692211062905124,
+      "learning_rate": 2.7948678914781347e-06,
+      "loss": 0.3653,
+      "step": 9978
+    },
+    {
+      "epoch": 1.5270084162203519,
+      "grad_norm": 2.1262724881376536,
+      "learning_rate": 2.7931496384877443e-06,
+      "loss": 0.2916,
+      "step": 9979
+    },
+    {
+      "epoch": 1.5271614384085692,
+      "grad_norm": 1.8299487915344057,
+      "learning_rate": 2.7914318280982346e-06,
+      "loss": 0.2591,
+      "step": 9980
+    },
+    {
+      "epoch": 1.5273144605967865,
+      "grad_norm": 2.3353736019494318,
+      "learning_rate": 2.7897144604150907e-06,
+      "loss": 0.3522,
+      "step": 9981
+    },
+    {
+      "epoch": 1.5274674827850037,
+      "grad_norm": 2.333490426475123,
+      "learning_rate": 2.7879975355437904e-06,
+      "loss": 0.3201,
+      "step": 9982
+    },
+    {
+      "epoch": 1.5276205049732212,
+      "grad_norm": 2.195731537268738,
+      "learning_rate": 2.786281053589779e-06,
+      "loss": 0.3418,
+      "step": 9983
+    },
+    {
+      "epoch": 1.5277735271614383,
+      "grad_norm": 2.755496101595831,
+      "learning_rate": 2.7845650146584626e-06,
+      "loss": 0.3516,
+      "step": 9984
+    },
+    {
+      "epoch": 1.5279265493496557,
+      "grad_norm": 2.3936080171100325,
+      "learning_rate": 2.782849418855238e-06,
+      "loss": 0.3365,
+      "step": 9985
+    },
+    {
+      "epoch": 1.528079571537873,
+      "grad_norm": 2.258254891109735,
+      "learning_rate": 2.7811342662854636e-06,
+      "loss": 0.3154,
+      "step": 9986
+    },
+    {
+      "epoch": 1.5282325937260903,
+      "grad_norm": 2.3318371299600336,
+      "learning_rate": 2.7794195570544745e-06,
+      "loss": 0.2555,
+      "step": 9987
+    },
+    {
+      "epoch": 1.5283856159143077,
+      "grad_norm": 2.022711659963486,
+      "learning_rate": 2.7777052912675785e-06,
+      "loss": 0.325,
+      "step": 9988
+    },
+    {
+      "epoch": 1.5285386381025248,
+      "grad_norm": 2.2626567628611745,
+      "learning_rate": 2.7759914690300536e-06,
+      "loss": 0.3974,
+      "step": 9989
+    },
+    {
+      "epoch": 1.528691660290742,
+      "grad_norm": 2.236395178465297,
+      "learning_rate": 2.7742780904471536e-06,
+      "loss": 0.3225,
+      "step": 9990
+    },
+    {
+      "epoch": 1.5288446824789594,
+      "grad_norm": 2.183933947211389,
+      "learning_rate": 2.772565155624103e-06,
+      "loss": 0.3282,
+      "step": 9991
+    },
+    {
+      "epoch": 1.5289977046671768,
+      "grad_norm": 2.3450557515589407,
+      "learning_rate": 2.7708526646660993e-06,
+      "loss": 0.3569,
+      "step": 9992
+    },
+    {
+      "epoch": 1.529150726855394,
+      "grad_norm": 1.7888436151823426,
+      "learning_rate": 2.769140617678315e-06,
+      "loss": 0.2846,
+      "step": 9993
+    },
+    {
+      "epoch": 1.5293037490436112,
+      "grad_norm": 2.0381056570029945,
+      "learning_rate": 2.767429014765889e-06,
+      "loss": 0.3703,
+      "step": 9994
+    },
+    {
+      "epoch": 1.5294567712318288,
+      "grad_norm": 2.2819477215176645,
+      "learning_rate": 2.7657178560339483e-06,
+      "loss": 0.2884,
+      "step": 9995
+    },
+    {
+      "epoch": 1.5296097934200459,
+      "grad_norm": 2.3538836527548694,
+      "learning_rate": 2.7640071415875703e-06,
+      "loss": 0.2908,
+      "step": 9996
+    },
+    {
+      "epoch": 1.5297628156082632,
+      "grad_norm": 2.017525350368832,
+      "learning_rate": 2.762296871531819e-06,
+      "loss": 0.369,
+      "step": 9997
+    },
+    {
+      "epoch": 1.5299158377964805,
+      "grad_norm": 2.146410794652233,
+      "learning_rate": 2.7605870459717367e-06,
+      "loss": 0.3486,
+      "step": 9998
+    },
+    {
+      "epoch": 1.5300688599846977,
+      "grad_norm": 2.170593065314875,
+      "learning_rate": 2.7588776650123215e-06,
+      "loss": 0.312,
+      "step": 9999
+    },
+    {
+      "epoch": 1.5302218821729152,
+      "grad_norm": 2.277094271624863,
+      "learning_rate": 2.7571687287585524e-06,
+      "loss": 0.3413,
+      "step": 10000
+    },
+    {
+      "epoch": 1.5303749043611323,
+      "grad_norm": 2.468686944978311,
+      "learning_rate": 2.7554602373153938e-06,
+      "loss": 0.3502,
+      "step": 10001
+    },
+    {
+      "epoch": 1.5305279265493497,
+      "grad_norm": 2.378155880261049,
+      "learning_rate": 2.7537521907877553e-06,
+      "loss": 0.3338,
+      "step": 10002
+    },
+    {
+      "epoch": 1.530680948737567,
+      "grad_norm": 2.177448943408823,
+      "learning_rate": 2.7520445892805457e-06,
+      "loss": 0.3183,
+      "step": 10003
+    },
+    {
+      "epoch": 1.530833970925784,
+      "grad_norm": 2.3647380308511594,
+      "learning_rate": 2.7503374328986355e-06,
+      "loss": 0.359,
+      "step": 10004
+    },
+    {
+      "epoch": 1.5309869931140017,
+      "grad_norm": 2.137209652082463,
+      "learning_rate": 2.748630721746858e-06,
+      "loss": 0.324,
+      "step": 10005
+    },
+    {
+      "epoch": 1.5311400153022188,
+      "grad_norm": 2.421214324585907,
+      "learning_rate": 2.7469244559300377e-06,
+      "loss": 0.3415,
+      "step": 10006
+    },
+    {
+      "epoch": 1.531293037490436,
+      "grad_norm": 2.1578732014508675,
+      "learning_rate": 2.7452186355529642e-06,
+      "loss": 0.3266,
+      "step": 10007
+    },
+    {
+      "epoch": 1.5314460596786534,
+      "grad_norm": 2.110705447832477,
+      "learning_rate": 2.7435132607203896e-06,
+      "loss": 0.3343,
+      "step": 10008
+    },
+    {
+      "epoch": 1.5315990818668705,
+      "grad_norm": 2.1867109539775305,
+      "learning_rate": 2.7418083315370557e-06,
+      "loss": 0.2972,
+      "step": 10009
+    },
+    {
+      "epoch": 1.531752104055088,
+      "grad_norm": 2.1025768332337798,
+      "learning_rate": 2.740103848107667e-06,
+      "loss": 0.3167,
+      "step": 10010
+    },
+    {
+      "epoch": 1.5319051262433052,
+      "grad_norm": 1.956002678795125,
+      "learning_rate": 2.7383998105369014e-06,
+      "loss": 0.3239,
+      "step": 10011
+    },
+    {
+      "epoch": 1.5320581484315225,
+      "grad_norm": 2.234365363174236,
+      "learning_rate": 2.736696218929411e-06,
+      "loss": 0.3692,
+      "step": 10012
+    },
+    {
+      "epoch": 1.5322111706197399,
+      "grad_norm": 2.360613287989332,
+      "learning_rate": 2.73499307338982e-06,
+      "loss": 0.3553,
+      "step": 10013
+    },
+    {
+      "epoch": 1.532364192807957,
+      "grad_norm": 2.0106390616987166,
+      "learning_rate": 2.7332903740227256e-06,
+      "loss": 0.2996,
+      "step": 10014
+    },
+    {
+      "epoch": 1.5325172149961745,
+      "grad_norm": 1.772493142680713,
+      "learning_rate": 2.7315881209326967e-06,
+      "loss": 0.2977,
+      "step": 10015
+    },
+    {
+      "epoch": 1.5326702371843917,
+      "grad_norm": 2.069526077229527,
+      "learning_rate": 2.729886314224275e-06,
+      "loss": 0.3248,
+      "step": 10016
+    },
+    {
+      "epoch": 1.532823259372609,
+      "grad_norm": 2.1173816114427004,
+      "learning_rate": 2.7281849540019755e-06,
+      "loss": 0.2778,
+      "step": 10017
+    },
+    {
+      "epoch": 1.5329762815608263,
+      "grad_norm": 2.4353246203851686,
+      "learning_rate": 2.726484040370286e-06,
+      "loss": 0.3159,
+      "step": 10018
+    },
+    {
+      "epoch": 1.5331293037490437,
+      "grad_norm": 2.080059375980662,
+      "learning_rate": 2.724783573433666e-06,
+      "loss": 0.2722,
+      "step": 10019
+    },
+    {
+      "epoch": 1.533282325937261,
+      "grad_norm": 2.231678350736275,
+      "learning_rate": 2.7230835532965473e-06,
+      "loss": 0.2778,
+      "step": 10020
+    },
+    {
+      "epoch": 1.533435348125478,
+      "grad_norm": 2.077749797828933,
+      "learning_rate": 2.7213839800633346e-06,
+      "loss": 0.2966,
+      "step": 10021
+    },
+    {
+      "epoch": 1.5335883703136954,
+      "grad_norm": 2.3270159778273998,
+      "learning_rate": 2.719684853838406e-06,
+      "loss": 0.3374,
+      "step": 10022
+    },
+    {
+      "epoch": 1.5337413925019128,
+      "grad_norm": 2.210011146928926,
+      "learning_rate": 2.717986174726108e-06,
+      "loss": 0.3091,
+      "step": 10023
+    },
+    {
+      "epoch": 1.53389441469013,
+      "grad_norm": 2.2217583618277064,
+      "learning_rate": 2.716287942830773e-06,
+      "loss": 0.3315,
+      "step": 10024
+    },
+    {
+      "epoch": 1.5340474368783474,
+      "grad_norm": 2.041482697435787,
+      "learning_rate": 2.714590158256687e-06,
+      "loss": 0.2848,
+      "step": 10025
+    },
+    {
+      "epoch": 1.5342004590665645,
+      "grad_norm": 2.1043763796800947,
+      "learning_rate": 2.7128928211081153e-06,
+      "loss": 0.2653,
+      "step": 10026
+    },
+    {
+      "epoch": 1.534353481254782,
+      "grad_norm": 2.154308009610621,
+      "learning_rate": 2.71119593148931e-06,
+      "loss": 0.3251,
+      "step": 10027
+    },
+    {
+      "epoch": 1.5345065034429992,
+      "grad_norm": 2.1032935774833383,
+      "learning_rate": 2.7094994895044736e-06,
+      "loss": 0.3703,
+      "step": 10028
+    },
+    {
+      "epoch": 1.5346595256312165,
+      "grad_norm": 2.290845349239085,
+      "learning_rate": 2.7078034952577905e-06,
+      "loss": 0.3318,
+      "step": 10029
+    },
+    {
+      "epoch": 1.5348125478194339,
+      "grad_norm": 2.1385725936478908,
+      "learning_rate": 2.706107948853428e-06,
+      "loss": 0.3057,
+      "step": 10030
+    },
+    {
+      "epoch": 1.534965570007651,
+      "grad_norm": 2.129033724441874,
+      "learning_rate": 2.704412850395505e-06,
+      "loss": 0.2826,
+      "step": 10031
+    },
+    {
+      "epoch": 1.5351185921958685,
+      "grad_norm": 2.1770308821410462,
+      "learning_rate": 2.7027181999881326e-06,
+      "loss": 0.4557,
+      "step": 10032
+    },
+    {
+      "epoch": 1.5352716143840857,
+      "grad_norm": 1.7528694429046794,
+      "learning_rate": 2.701023997735385e-06,
+      "loss": 0.2957,
+      "step": 10033
+    },
+    {
+      "epoch": 1.535424636572303,
+      "grad_norm": 2.167855205306468,
+      "learning_rate": 2.6993302437413006e-06,
+      "loss": 0.3485,
+      "step": 10034
+    },
+    {
+      "epoch": 1.5355776587605203,
+      "grad_norm": 2.3501051359832354,
+      "learning_rate": 2.697636938109911e-06,
+      "loss": 0.3644,
+      "step": 10035
+    },
+    {
+      "epoch": 1.5357306809487374,
+      "grad_norm": 1.9241981996516704,
+      "learning_rate": 2.6959440809452055e-06,
+      "loss": 0.2791,
+      "step": 10036
+    },
+    {
+      "epoch": 1.535883703136955,
+      "grad_norm": 1.9695651915181258,
+      "learning_rate": 2.6942516723511424e-06,
+      "loss": 0.3818,
+      "step": 10037
+    },
+    {
+      "epoch": 1.536036725325172,
+      "grad_norm": 1.8640844680454627,
+      "learning_rate": 2.6925597124316673e-06,
+      "loss": 0.29,
+      "step": 10038
+    },
+    {
+      "epoch": 1.5361897475133894,
+      "grad_norm": 2.6503153964138466,
+      "learning_rate": 2.6908682012906874e-06,
+      "loss": 0.2854,
+      "step": 10039
+    },
+    {
+      "epoch": 1.5363427697016068,
+      "grad_norm": 2.1630247633251076,
+      "learning_rate": 2.689177139032084e-06,
+      "loss": 0.2942,
+      "step": 10040
+    },
+    {
+      "epoch": 1.5364957918898239,
+      "grad_norm": 2.031261244127211,
+      "learning_rate": 2.6874865257597126e-06,
+      "loss": 0.2848,
+      "step": 10041
+    },
+    {
+      "epoch": 1.5366488140780414,
+      "grad_norm": 1.9718912748795485,
+      "learning_rate": 2.685796361577402e-06,
+      "loss": 0.2793,
+      "step": 10042
+    },
+    {
+      "epoch": 1.5368018362662585,
+      "grad_norm": 2.306037088480719,
+      "learning_rate": 2.684106646588949e-06,
+      "loss": 0.32,
+      "step": 10043
+    },
+    {
+      "epoch": 1.5369548584544759,
+      "grad_norm": 2.270018705118447,
+      "learning_rate": 2.682417380898126e-06,
+      "loss": 0.397,
+      "step": 10044
+    },
+    {
+      "epoch": 1.5371078806426932,
+      "grad_norm": 1.9460779455142152,
+      "learning_rate": 2.680728564608679e-06,
+      "loss": 0.3066,
+      "step": 10045
+    },
+    {
+      "epoch": 1.5372609028309103,
+      "grad_norm": 2.040083304159909,
+      "learning_rate": 2.679040197824324e-06,
+      "loss": 0.3022,
+      "step": 10046
+    },
+    {
+      "epoch": 1.5374139250191279,
+      "grad_norm": 1.698927479185634,
+      "learning_rate": 2.6773522806487494e-06,
+      "loss": 0.2381,
+      "step": 10047
+    },
+    {
+      "epoch": 1.537566947207345,
+      "grad_norm": 2.159105375673701,
+      "learning_rate": 2.675664813185619e-06,
+      "loss": 0.3544,
+      "step": 10048
+    },
+    {
+      "epoch": 1.5377199693955623,
+      "grad_norm": 2.1797933921292647,
+      "learning_rate": 2.673977795538565e-06,
+      "loss": 0.3073,
+      "step": 10049
+    },
+    {
+      "epoch": 1.5378729915837797,
+      "grad_norm": 2.1427002258465944,
+      "learning_rate": 2.6722912278111936e-06,
+      "loss": 0.3055,
+      "step": 10050
+    },
+    {
+      "epoch": 1.538026013771997,
+      "grad_norm": 2.273707572216091,
+      "learning_rate": 2.670605110107084e-06,
+      "loss": 0.3434,
+      "step": 10051
+    },
+    {
+      "epoch": 1.5381790359602143,
+      "grad_norm": 2.0306393620915353,
+      "learning_rate": 2.668919442529785e-06,
+      "loss": 0.2714,
+      "step": 10052
+    },
+    {
+      "epoch": 1.5383320581484314,
+      "grad_norm": 2.1099140870250404,
+      "learning_rate": 2.6672342251828274e-06,
+      "loss": 0.2864,
+      "step": 10053
+    },
+    {
+      "epoch": 1.5384850803366488,
+      "grad_norm": 2.9079589709237963,
+      "learning_rate": 2.6655494581696994e-06,
+      "loss": 0.4068,
+      "step": 10054
+    },
+    {
+      "epoch": 1.538638102524866,
+      "grad_norm": 1.9689432719376831,
+      "learning_rate": 2.6638651415938697e-06,
+      "loss": 0.3411,
+      "step": 10055
+    },
+    {
+      "epoch": 1.5387911247130834,
+      "grad_norm": 2.114549094429897,
+      "learning_rate": 2.662181275558786e-06,
+      "loss": 0.307,
+      "step": 10056
+    },
+    {
+      "epoch": 1.5389441469013008,
+      "grad_norm": 2.057594249117251,
+      "learning_rate": 2.6604978601678523e-06,
+      "loss": 0.2666,
+      "step": 10057
+    },
+    {
+      "epoch": 1.5390971690895179,
+      "grad_norm": 2.232955568610074,
+      "learning_rate": 2.658814895524455e-06,
+      "loss": 0.3322,
+      "step": 10058
+    },
+    {
+      "epoch": 1.5392501912777354,
+      "grad_norm": 2.2654884542844154,
+      "learning_rate": 2.6571323817319593e-06,
+      "loss": 0.3371,
+      "step": 10059
+    },
+    {
+      "epoch": 1.5394032134659525,
+      "grad_norm": 2.4057706843267748,
+      "learning_rate": 2.6554503188936844e-06,
+      "loss": 0.3654,
+      "step": 10060
+    },
+    {
+      "epoch": 1.5395562356541699,
+      "grad_norm": 2.1902711910222266,
+      "learning_rate": 2.6537687071129404e-06,
+      "loss": 0.3098,
+      "step": 10061
+    },
+    {
+      "epoch": 1.5397092578423872,
+      "grad_norm": 2.073188025318028,
+      "learning_rate": 2.6520875464930006e-06,
+      "loss": 0.2854,
+      "step": 10062
+    },
+    {
+      "epoch": 1.5398622800306043,
+      "grad_norm": 2.0989366632201287,
+      "learning_rate": 2.650406837137104e-06,
+      "loss": 0.3067,
+      "step": 10063
+    },
+    {
+      "epoch": 1.5400153022188219,
+      "grad_norm": 2.2188949698155147,
+      "learning_rate": 2.6487265791484795e-06,
+      "loss": 0.2936,
+      "step": 10064
+    },
+    {
+      "epoch": 1.540168324407039,
+      "grad_norm": 2.263393577409095,
+      "learning_rate": 2.6470467726303163e-06,
+      "loss": 0.3334,
+      "step": 10065
+    },
+    {
+      "epoch": 1.5403213465952563,
+      "grad_norm": 2.1229052913443023,
+      "learning_rate": 2.6453674176857693e-06,
+      "loss": 0.3446,
+      "step": 10066
+    },
+    {
+      "epoch": 1.5404743687834737,
+      "grad_norm": 2.0820187078812995,
+      "learning_rate": 2.6436885144179848e-06,
+      "loss": 0.36,
+      "step": 10067
+    },
+    {
+      "epoch": 1.5406273909716908,
+      "grad_norm": 2.036822752359246,
+      "learning_rate": 2.6420100629300648e-06,
+      "loss": 0.3523,
+      "step": 10068
+    },
+    {
+      "epoch": 1.5407804131599083,
+      "grad_norm": 2.118531167022728,
+      "learning_rate": 2.640332063325093e-06,
+      "loss": 0.2695,
+      "step": 10069
+    },
+    {
+      "epoch": 1.5409334353481254,
+      "grad_norm": 2.1807415957671457,
+      "learning_rate": 2.6386545157061207e-06,
+      "loss": 0.3469,
+      "step": 10070
+    },
+    {
+      "epoch": 1.5410864575363428,
+      "grad_norm": 2.1707900822925947,
+      "learning_rate": 2.636977420176171e-06,
+      "loss": 0.2875,
+      "step": 10071
+    },
+    {
+      "epoch": 1.54123947972456,
+      "grad_norm": 2.4577498652524254,
+      "learning_rate": 2.635300776838243e-06,
+      "loss": 0.3673,
+      "step": 10072
+    },
+    {
+      "epoch": 1.5413925019127772,
+      "grad_norm": 2.0434551137883297,
+      "learning_rate": 2.633624585795305e-06,
+      "loss": 0.3203,
+      "step": 10073
+    },
+    {
+      "epoch": 1.5415455241009948,
+      "grad_norm": 1.7411301508901227,
+      "learning_rate": 2.6319488471502984e-06,
+      "loss": 0.2563,
+      "step": 10074
+    },
+    {
+      "epoch": 1.5416985462892119,
+      "grad_norm": 2.1478712253541206,
+      "learning_rate": 2.630273561006138e-06,
+      "loss": 0.295,
+      "step": 10075
+    },
+    {
+      "epoch": 1.5418515684774292,
+      "grad_norm": 1.90429011588242,
+      "learning_rate": 2.6285987274657045e-06,
+      "loss": 0.2753,
+      "step": 10076
+    },
+    {
+      "epoch": 1.5420045906656465,
+      "grad_norm": 2.1754047440005437,
+      "learning_rate": 2.6269243466318673e-06,
+      "loss": 0.3106,
+      "step": 10077
+    },
+    {
+      "epoch": 1.5421576128538637,
+      "grad_norm": 2.2461452419660524,
+      "learning_rate": 2.625250418607446e-06,
+      "loss": 0.3741,
+      "step": 10078
+    },
+    {
+      "epoch": 1.5423106350420812,
+      "grad_norm": 2.193766353742696,
+      "learning_rate": 2.623576943495244e-06,
+      "loss": 0.3146,
+      "step": 10079
+    },
+    {
+      "epoch": 1.5424636572302983,
+      "grad_norm": 1.7423098728249427,
+      "learning_rate": 2.6219039213980445e-06,
+      "loss": 0.2811,
+      "step": 10080
+    },
+    {
+      "epoch": 1.5426166794185157,
+      "grad_norm": 1.961978767244435,
+      "learning_rate": 2.620231352418585e-06,
+      "loss": 0.5452,
+      "step": 10081
+    },
+    {
+      "epoch": 1.542769701606733,
+      "grad_norm": 2.062082207863998,
+      "learning_rate": 2.618559236659586e-06,
+      "loss": 0.3589,
+      "step": 10082
+    },
+    {
+      "epoch": 1.54292272379495,
+      "grad_norm": 2.0213839010709878,
+      "learning_rate": 2.6168875742237454e-06,
+      "loss": 0.288,
+      "step": 10083
+    },
+    {
+      "epoch": 1.5430757459831677,
+      "grad_norm": 2.0088016852578936,
+      "learning_rate": 2.6152163652137165e-06,
+      "loss": 0.2905,
+      "step": 10084
+    },
+    {
+      "epoch": 1.5432287681713848,
+      "grad_norm": 1.8323811164141244,
+      "learning_rate": 2.613545609732142e-06,
+      "loss": 0.3337,
+      "step": 10085
+    },
+    {
+      "epoch": 1.543381790359602,
+      "grad_norm": 2.0627168321060365,
+      "learning_rate": 2.6118753078816315e-06,
+      "loss": 0.3066,
+      "step": 10086
+    },
+    {
+      "epoch": 1.5435348125478194,
+      "grad_norm": 2.2910285838159834,
+      "learning_rate": 2.610205459764755e-06,
+      "loss": 0.3516,
+      "step": 10087
+    },
+    {
+      "epoch": 1.5436878347360368,
+      "grad_norm": 1.9985040435482564,
+      "learning_rate": 2.608536065484073e-06,
+      "loss": 0.3262,
+      "step": 10088
+    },
+    {
+      "epoch": 1.543840856924254,
+      "grad_norm": 2.2496865624069855,
+      "learning_rate": 2.606867125142107e-06,
+      "loss": 0.3033,
+      "step": 10089
+    },
+    {
+      "epoch": 1.5439938791124712,
+      "grad_norm": 2.241097444447969,
+      "learning_rate": 2.605198638841353e-06,
+      "loss": 0.3357,
+      "step": 10090
+    },
+    {
+      "epoch": 1.5441469013006885,
+      "grad_norm": 2.1992885647227625,
+      "learning_rate": 2.603530606684279e-06,
+      "loss": 0.3333,
+      "step": 10091
+    },
+    {
+      "epoch": 1.5442999234889059,
+      "grad_norm": 1.9012812969047344,
+      "learning_rate": 2.601863028773326e-06,
+      "loss": 0.2586,
+      "step": 10092
+    },
+    {
+      "epoch": 1.5444529456771232,
+      "grad_norm": 1.8123047465528561,
+      "learning_rate": 2.600195905210905e-06,
+      "loss": 0.2823,
+      "step": 10093
+    },
+    {
+      "epoch": 1.5446059678653405,
+      "grad_norm": 2.3152030446910787,
+      "learning_rate": 2.598529236099403e-06,
+      "loss": 0.3443,
+      "step": 10094
+    },
+    {
+      "epoch": 1.5447589900535577,
+      "grad_norm": 2.1743439640613484,
+      "learning_rate": 2.596863021541175e-06,
+      "loss": 0.3019,
+      "step": 10095
+    },
+    {
+      "epoch": 1.5449120122417752,
+      "grad_norm": 1.9172471293093292,
+      "learning_rate": 2.595197261638549e-06,
+      "loss": 0.2936,
+      "step": 10096
+    },
+    {
+      "epoch": 1.5450650344299923,
+      "grad_norm": 1.9253862639636048,
+      "learning_rate": 2.5935319564938275e-06,
+      "loss": 0.2985,
+      "step": 10097
+    },
+    {
+      "epoch": 1.5452180566182097,
+      "grad_norm": 1.9430570051191849,
+      "learning_rate": 2.5918671062092836e-06,
+      "loss": 0.2914,
+      "step": 10098
+    },
+    {
+      "epoch": 1.545371078806427,
+      "grad_norm": 2.2577123270161428,
+      "learning_rate": 2.590202710887161e-06,
+      "loss": 0.3272,
+      "step": 10099
+    },
+    {
+      "epoch": 1.545524100994644,
+      "grad_norm": 2.2632677012722895,
+      "learning_rate": 2.588538770629677e-06,
+      "loss": 0.2962,
+      "step": 10100
+    },
+    {
+      "epoch": 1.5456771231828617,
+      "grad_norm": 1.9504870317490723,
+      "learning_rate": 2.586875285539021e-06,
+      "loss": 0.2772,
+      "step": 10101
+    },
+    {
+      "epoch": 1.5458301453710788,
+      "grad_norm": 1.9342150897613446,
+      "learning_rate": 2.5852122557173542e-06,
+      "loss": 0.3467,
+      "step": 10102
+    },
+    {
+      "epoch": 1.545983167559296,
+      "grad_norm": 1.951851157913786,
+      "learning_rate": 2.5835496812668095e-06,
+      "loss": 0.2735,
+      "step": 10103
+    },
+    {
+      "epoch": 1.5461361897475134,
+      "grad_norm": 2.142026102977652,
+      "learning_rate": 2.581887562289491e-06,
+      "loss": 0.2972,
+      "step": 10104
+    },
+    {
+      "epoch": 1.5462892119357305,
+      "grad_norm": 2.193269441158662,
+      "learning_rate": 2.5802258988874762e-06,
+      "loss": 0.3203,
+      "step": 10105
+    },
+    {
+      "epoch": 1.546442234123948,
+      "grad_norm": 1.8676996400343906,
+      "learning_rate": 2.5785646911628193e-06,
+      "loss": 0.3037,
+      "step": 10106
+    },
+    {
+      "epoch": 1.5465952563121652,
+      "grad_norm": 2.093923674004089,
+      "learning_rate": 2.5769039392175353e-06,
+      "loss": 0.3632,
+      "step": 10107
+    },
+    {
+      "epoch": 1.5467482785003825,
+      "grad_norm": 1.9006140368770028,
+      "learning_rate": 2.5752436431536174e-06,
+      "loss": 0.2753,
+      "step": 10108
+    },
+    {
+      "epoch": 1.5469013006885999,
+      "grad_norm": 2.2120676657834943,
+      "learning_rate": 2.573583803073039e-06,
+      "loss": 0.2726,
+      "step": 10109
+    },
+    {
+      "epoch": 1.547054322876817,
+      "grad_norm": 2.1076883294477424,
+      "learning_rate": 2.571924419077728e-06,
+      "loss": 0.3216,
+      "step": 10110
+    },
+    {
+      "epoch": 1.5472073450650345,
+      "grad_norm": 2.0232240951285716,
+      "learning_rate": 2.5702654912695945e-06,
+      "loss": 0.3267,
+      "step": 10111
+    },
+    {
+      "epoch": 1.5473603672532517,
+      "grad_norm": 2.373217544040392,
+      "learning_rate": 2.568607019750529e-06,
+      "loss": 0.3187,
+      "step": 10112
+    },
+    {
+      "epoch": 1.547513389441469,
+      "grad_norm": 2.182283013798894,
+      "learning_rate": 2.5669490046223713e-06,
+      "loss": 0.3073,
+      "step": 10113
+    },
+    {
+      "epoch": 1.5476664116296863,
+      "grad_norm": 1.9217437998228828,
+      "learning_rate": 2.5652914459869573e-06,
+      "loss": 0.2509,
+      "step": 10114
+    },
+    {
+      "epoch": 1.5478194338179034,
+      "grad_norm": 2.2608685850264014,
+      "learning_rate": 2.563634343946082e-06,
+      "loss": 0.3531,
+      "step": 10115
+    },
+    {
+      "epoch": 1.547972456006121,
+      "grad_norm": 2.32705780663519,
+      "learning_rate": 2.5619776986015077e-06,
+      "loss": 0.3757,
+      "step": 10116
+    },
+    {
+      "epoch": 1.548125478194338,
+      "grad_norm": 2.3217036817700647,
+      "learning_rate": 2.560321510054984e-06,
+      "loss": 0.3321,
+      "step": 10117
+    },
+    {
+      "epoch": 1.5482785003825554,
+      "grad_norm": 2.1098728752447795,
+      "learning_rate": 2.5586657784082236e-06,
+      "loss": 0.3464,
+      "step": 10118
+    },
+    {
+      "epoch": 1.5484315225707728,
+      "grad_norm": 2.1231402121495524,
+      "learning_rate": 2.5570105037629013e-06,
+      "loss": 0.2992,
+      "step": 10119
+    },
+    {
+      "epoch": 1.54858454475899,
+      "grad_norm": 2.303456688738964,
+      "learning_rate": 2.5553556862206852e-06,
+      "loss": 0.3267,
+      "step": 10120
+    },
+    {
+      "epoch": 1.5487375669472074,
+      "grad_norm": 2.12653714669132,
+      "learning_rate": 2.5537013258832e-06,
+      "loss": 0.2904,
+      "step": 10121
+    },
+    {
+      "epoch": 1.5488905891354245,
+      "grad_norm": 2.3782425204672273,
+      "learning_rate": 2.5520474228520454e-06,
+      "loss": 0.3159,
+      "step": 10122
+    },
+    {
+      "epoch": 1.5490436113236419,
+      "grad_norm": 2.0454415788829925,
+      "learning_rate": 2.5503939772287957e-06,
+      "loss": 0.283,
+      "step": 10123
+    },
+    {
+      "epoch": 1.5491966335118592,
+      "grad_norm": 2.184366989942858,
+      "learning_rate": 2.548740989114995e-06,
+      "loss": 0.3246,
+      "step": 10124
+    },
+    {
+      "epoch": 1.5493496557000765,
+      "grad_norm": 2.144316447092154,
+      "learning_rate": 2.5470884586121604e-06,
+      "loss": 0.3074,
+      "step": 10125
+    },
+    {
+      "epoch": 1.5495026778882939,
+      "grad_norm": 2.292841312921272,
+      "learning_rate": 2.5454363858217778e-06,
+      "loss": 0.291,
+      "step": 10126
+    },
+    {
+      "epoch": 1.549655700076511,
+      "grad_norm": 2.352491997320857,
+      "learning_rate": 2.543784770845311e-06,
+      "loss": 0.3496,
+      "step": 10127
+    },
+    {
+      "epoch": 1.5498087222647285,
+      "grad_norm": 2.204005708224258,
+      "learning_rate": 2.542133613784189e-06,
+      "loss": 0.3373,
+      "step": 10128
+    },
+    {
+      "epoch": 1.5499617444529457,
+      "grad_norm": 2.2452574362968023,
+      "learning_rate": 2.540482914739818e-06,
+      "loss": 0.3437,
+      "step": 10129
+    },
+    {
+      "epoch": 1.550114766641163,
+      "grad_norm": 2.279324045464706,
+      "learning_rate": 2.5388326738135726e-06,
+      "loss": 0.3119,
+      "step": 10130
+    },
+    {
+      "epoch": 1.5502677888293803,
+      "grad_norm": 2.256376745605677,
+      "learning_rate": 2.537182891106801e-06,
+      "loss": 0.346,
+      "step": 10131
+    },
+    {
+      "epoch": 1.5504208110175974,
+      "grad_norm": 1.927680952087128,
+      "learning_rate": 2.5355335667208226e-06,
+      "loss": 0.2626,
+      "step": 10132
+    },
+    {
+      "epoch": 1.550573833205815,
+      "grad_norm": 1.925711586506596,
+      "learning_rate": 2.533884700756929e-06,
+      "loss": 0.2571,
+      "step": 10133
+    },
+    {
+      "epoch": 1.550726855394032,
+      "grad_norm": 1.9586609335259824,
+      "learning_rate": 2.5322362933163803e-06,
+      "loss": 0.2807,
+      "step": 10134
+    },
+    {
+      "epoch": 1.5508798775822494,
+      "grad_norm": 2.305373905796804,
+      "learning_rate": 2.5305883445004207e-06,
+      "loss": 0.333,
+      "step": 10135
+    },
+    {
+      "epoch": 1.5510328997704668,
+      "grad_norm": 1.9006313901845546,
+      "learning_rate": 2.5289408544102488e-06,
+      "loss": 0.2386,
+      "step": 10136
+    },
+    {
+      "epoch": 1.5511859219586839,
+      "grad_norm": 1.9406112813516276,
+      "learning_rate": 2.5272938231470433e-06,
+      "loss": 0.3152,
+      "step": 10137
+    },
+    {
+      "epoch": 1.5513389441469014,
+      "grad_norm": 2.021371847556138,
+      "learning_rate": 2.5256472508119633e-06,
+      "loss": 0.2632,
+      "step": 10138
+    },
+    {
+      "epoch": 1.5514919663351185,
+      "grad_norm": 1.9119227896899973,
+      "learning_rate": 2.5240011375061226e-06,
+      "loss": 0.3115,
+      "step": 10139
+    },
+    {
+      "epoch": 1.5516449885233359,
+      "grad_norm": 2.018861968257367,
+      "learning_rate": 2.5223554833306153e-06,
+      "loss": 0.316,
+      "step": 10140
+    },
+    {
+      "epoch": 1.5517980107115532,
+      "grad_norm": 2.133098819530022,
+      "learning_rate": 2.5207102883865163e-06,
+      "loss": 0.2879,
+      "step": 10141
+    },
+    {
+      "epoch": 1.5519510328997703,
+      "grad_norm": 1.928501157699594,
+      "learning_rate": 2.519065552774851e-06,
+      "loss": 0.219,
+      "step": 10142
+    },
+    {
+      "epoch": 1.5521040550879879,
+      "grad_norm": 2.0714889924024713,
+      "learning_rate": 2.5174212765966398e-06,
+      "loss": 0.3353,
+      "step": 10143
+    },
+    {
+      "epoch": 1.552257077276205,
+      "grad_norm": 1.9810537630101772,
+      "learning_rate": 2.5157774599528627e-06,
+      "loss": 0.316,
+      "step": 10144
+    },
+    {
+      "epoch": 1.5524100994644223,
+      "grad_norm": 2.097340227564223,
+      "learning_rate": 2.5141341029444634e-06,
+      "loss": 0.243,
+      "step": 10145
+    },
+    {
+      "epoch": 1.5525631216526397,
+      "grad_norm": 2.322435168108173,
+      "learning_rate": 2.5124912056723784e-06,
+      "loss": 0.3103,
+      "step": 10146
+    },
+    {
+      "epoch": 1.5527161438408568,
+      "grad_norm": 2.3468231639208232,
+      "learning_rate": 2.5108487682375017e-06,
+      "loss": 0.3325,
+      "step": 10147
+    },
+    {
+      "epoch": 1.5528691660290743,
+      "grad_norm": 2.497911269354406,
+      "learning_rate": 2.509206790740694e-06,
+      "loss": 0.3676,
+      "step": 10148
+    },
+    {
+      "epoch": 1.5530221882172914,
+      "grad_norm": 1.8090734411304665,
+      "learning_rate": 2.5075652732828036e-06,
+      "loss": 0.2453,
+      "step": 10149
+    },
+    {
+      "epoch": 1.5531752104055088,
+      "grad_norm": 2.403872379075245,
+      "learning_rate": 2.5059242159646412e-06,
+      "loss": 0.2632,
+      "step": 10150
+    },
+    {
+      "epoch": 1.553328232593726,
+      "grad_norm": 2.430625917359335,
+      "learning_rate": 2.504283618886989e-06,
+      "loss": 0.38,
+      "step": 10151
+    },
+    {
+      "epoch": 1.5534812547819434,
+      "grad_norm": 1.8545837232546376,
+      "learning_rate": 2.502643482150604e-06,
+      "loss": 0.2823,
+      "step": 10152
+    },
+    {
+      "epoch": 1.5536342769701608,
+      "grad_norm": 1.9418639929290726,
+      "learning_rate": 2.5010038058562127e-06,
+      "loss": 0.3061,
+      "step": 10153
+    },
+    {
+      "epoch": 1.5537872991583779,
+      "grad_norm": 1.8986203674410438,
+      "learning_rate": 2.499364590104514e-06,
+      "loss": 0.2776,
+      "step": 10154
+    },
+    {
+      "epoch": 1.5539403213465952,
+      "grad_norm": 2.0270382335653707,
+      "learning_rate": 2.4977258349961774e-06,
+      "loss": 0.2981,
+      "step": 10155
+    },
+    {
+      "epoch": 1.5540933435348125,
+      "grad_norm": 2.402283314200567,
+      "learning_rate": 2.496087540631846e-06,
+      "loss": 0.3866,
+      "step": 10156
+    },
+    {
+      "epoch": 1.5542463657230299,
+      "grad_norm": 2.310733483309164,
+      "learning_rate": 2.4944497071121355e-06,
+      "loss": 0.3839,
+      "step": 10157
+    },
+    {
+      "epoch": 1.5543993879112472,
+      "grad_norm": 1.8695856951851826,
+      "learning_rate": 2.4928123345376276e-06,
+      "loss": 0.249,
+      "step": 10158
+    },
+    {
+      "epoch": 1.5545524100994643,
+      "grad_norm": 2.1169212162195827,
+      "learning_rate": 2.491175423008888e-06,
+      "loss": 0.2898,
+      "step": 10159
+    },
+    {
+      "epoch": 1.5547054322876819,
+      "grad_norm": 2.161399766428231,
+      "learning_rate": 2.4895389726264376e-06,
+      "loss": 0.3391,
+      "step": 10160
+    },
+    {
+      "epoch": 1.554858454475899,
+      "grad_norm": 2.391172981427365,
+      "learning_rate": 2.4879029834907762e-06,
+      "loss": 0.3253,
+      "step": 10161
+    },
+    {
+      "epoch": 1.5550114766641163,
+      "grad_norm": 2.100970641331779,
+      "learning_rate": 2.486267455702387e-06,
+      "loss": 0.3261,
+      "step": 10162
+    },
+    {
+      "epoch": 1.5551644988523337,
+      "grad_norm": 1.9156665737120335,
+      "learning_rate": 2.4846323893616996e-06,
+      "loss": 0.2877,
+      "step": 10163
+    },
+    {
+      "epoch": 1.5553175210405508,
+      "grad_norm": 2.3248612982758057,
+      "learning_rate": 2.4829977845691424e-06,
+      "loss": 0.3276,
+      "step": 10164
+    },
+    {
+      "epoch": 1.5554705432287683,
+      "grad_norm": 2.032228598566149,
+      "learning_rate": 2.4813636414251e-06,
+      "loss": 0.2805,
+      "step": 10165
+    },
+    {
+      "epoch": 1.5556235654169854,
+      "grad_norm": 1.9488852461732409,
+      "learning_rate": 2.479729960029924e-06,
+      "loss": 0.281,
+      "step": 10166
+    },
+    {
+      "epoch": 1.5557765876052028,
+      "grad_norm": 2.137426975125268,
+      "learning_rate": 2.4780967404839528e-06,
+      "loss": 0.3415,
+      "step": 10167
+    },
+    {
+      "epoch": 1.55592960979342,
+      "grad_norm": 1.8447028628909974,
+      "learning_rate": 2.4764639828874905e-06,
+      "loss": 0.2041,
+      "step": 10168
+    },
+    {
+      "epoch": 1.5560826319816372,
+      "grad_norm": 1.8215233612139778,
+      "learning_rate": 2.4748316873408006e-06,
+      "loss": 0.2766,
+      "step": 10169
+    },
+    {
+      "epoch": 1.5562356541698548,
+      "grad_norm": 2.118588620302535,
+      "learning_rate": 2.473199853944138e-06,
+      "loss": 0.3702,
+      "step": 10170
+    },
+    {
+      "epoch": 1.5563886763580719,
+      "grad_norm": 2.316013267038449,
+      "learning_rate": 2.4715684827977183e-06,
+      "loss": 0.3036,
+      "step": 10171
+    },
+    {
+      "epoch": 1.5565416985462892,
+      "grad_norm": 2.30550773332787,
+      "learning_rate": 2.4699375740017296e-06,
+      "loss": 0.3842,
+      "step": 10172
+    },
+    {
+      "epoch": 1.5566947207345065,
+      "grad_norm": 2.0189253349152816,
+      "learning_rate": 2.468307127656331e-06,
+      "loss": 0.2854,
+      "step": 10173
+    },
+    {
+      "epoch": 1.5568477429227237,
+      "grad_norm": 1.8466817397772421,
+      "learning_rate": 2.4666771438616568e-06,
+      "loss": 0.2789,
+      "step": 10174
+    },
+    {
+      "epoch": 1.5570007651109412,
+      "grad_norm": 2.06378417070918,
+      "learning_rate": 2.46504762271781e-06,
+      "loss": 0.2566,
+      "step": 10175
+    },
+    {
+      "epoch": 1.5571537872991583,
+      "grad_norm": 2.0737045550933497,
+      "learning_rate": 2.4634185643248642e-06,
+      "loss": 0.3565,
+      "step": 10176
+    },
+    {
+      "epoch": 1.5573068094873757,
+      "grad_norm": 1.941031063446645,
+      "learning_rate": 2.4617899687828693e-06,
+      "loss": 0.2481,
+      "step": 10177
+    },
+    {
+      "epoch": 1.557459831675593,
+      "grad_norm": 2.225129231890635,
+      "learning_rate": 2.4601618361918413e-06,
+      "loss": 0.3176,
+      "step": 10178
+    },
+    {
+      "epoch": 1.55761285386381,
+      "grad_norm": 2.016757570425307,
+      "learning_rate": 2.45853416665177e-06,
+      "loss": 0.3122,
+      "step": 10179
+    },
+    {
+      "epoch": 1.5577658760520277,
+      "grad_norm": 2.17651730016683,
+      "learning_rate": 2.4569069602626196e-06,
+      "loss": 0.2992,
+      "step": 10180
+    },
+    {
+      "epoch": 1.5579188982402448,
+      "grad_norm": 1.9707363895900585,
+      "learning_rate": 2.4552802171243204e-06,
+      "loss": 0.3621,
+      "step": 10181
+    },
+    {
+      "epoch": 1.558071920428462,
+      "grad_norm": 2.2690988936611562,
+      "learning_rate": 2.4536539373367786e-06,
+      "loss": 0.32,
+      "step": 10182
+    },
+    {
+      "epoch": 1.5582249426166794,
+      "grad_norm": 2.260127613008053,
+      "learning_rate": 2.45202812099987e-06,
+      "loss": 0.313,
+      "step": 10183
+    },
+    {
+      "epoch": 1.5583779648048968,
+      "grad_norm": 2.092136322623526,
+      "learning_rate": 2.4504027682134423e-06,
+      "loss": 0.3156,
+      "step": 10184
+    },
+    {
+      "epoch": 1.558530986993114,
+      "grad_norm": 1.9982530177796707,
+      "learning_rate": 2.4487778790773155e-06,
+      "loss": 0.2735,
+      "step": 10185
+    },
+    {
+      "epoch": 1.5586840091813312,
+      "grad_norm": 2.1708897206182507,
+      "learning_rate": 2.447153453691279e-06,
+      "loss": 0.3133,
+      "step": 10186
+    },
+    {
+      "epoch": 1.5588370313695485,
+      "grad_norm": 2.0960282856628942,
+      "learning_rate": 2.445529492155092e-06,
+      "loss": 0.3201,
+      "step": 10187
+    },
+    {
+      "epoch": 1.5589900535577659,
+      "grad_norm": 2.1436054196035044,
+      "learning_rate": 2.4439059945684997e-06,
+      "loss": 0.3282,
+      "step": 10188
+    },
+    {
+      "epoch": 1.5591430757459832,
+      "grad_norm": 1.9695191815716822,
+      "learning_rate": 2.4422829610311948e-06,
+      "loss": 0.3069,
+      "step": 10189
+    },
+    {
+      "epoch": 1.5592960979342005,
+      "grad_norm": 2.1926718704756807,
+      "learning_rate": 2.4406603916428584e-06,
+      "loss": 0.3344,
+      "step": 10190
+    },
+    {
+      "epoch": 1.5594491201224177,
+      "grad_norm": 2.3383201047133855,
+      "learning_rate": 2.439038286503145e-06,
+      "loss": 0.336,
+      "step": 10191
+    },
+    {
+      "epoch": 1.559602142310635,
+      "grad_norm": 1.8372155716325524,
+      "learning_rate": 2.437416645711662e-06,
+      "loss": 0.2345,
+      "step": 10192
+    },
+    {
+      "epoch": 1.5597551644988523,
+      "grad_norm": 2.495769900944102,
+      "learning_rate": 2.435795469368012e-06,
+      "loss": 0.3478,
+      "step": 10193
+    },
+    {
+      "epoch": 1.5599081866870697,
+      "grad_norm": 2.1221966929936404,
+      "learning_rate": 2.4341747575717554e-06,
+      "loss": 0.3035,
+      "step": 10194
+    },
+    {
+      "epoch": 1.560061208875287,
+      "grad_norm": 2.0135930107293802,
+      "learning_rate": 2.43255451042242e-06,
+      "loss": 0.299,
+      "step": 10195
+    },
+    {
+      "epoch": 1.560214231063504,
+      "grad_norm": 2.1633570926567245,
+      "learning_rate": 2.4309347280195183e-06,
+      "loss": 0.3154,
+      "step": 10196
+    },
+    {
+      "epoch": 1.5603672532517217,
+      "grad_norm": 1.9875556110795365,
+      "learning_rate": 2.4293154104625282e-06,
+      "loss": 0.2904,
+      "step": 10197
+    },
+    {
+      "epoch": 1.5605202754399388,
+      "grad_norm": 2.340039983969419,
+      "learning_rate": 2.4276965578508905e-06,
+      "loss": 0.2656,
+      "step": 10198
+    },
+    {
+      "epoch": 1.560673297628156,
+      "grad_norm": 2.221132754924476,
+      "learning_rate": 2.426078170284032e-06,
+      "loss": 0.3104,
+      "step": 10199
+    },
+    {
+      "epoch": 1.5608263198163734,
+      "grad_norm": 2.073516952937971,
+      "learning_rate": 2.4244602478613433e-06,
+      "loss": 0.3356,
+      "step": 10200
+    },
+    {
+      "epoch": 1.5609793420045905,
+      "grad_norm": 2.488583914365108,
+      "learning_rate": 2.4228427906821863e-06,
+      "loss": 0.3038,
+      "step": 10201
+    },
+    {
+      "epoch": 1.561132364192808,
+      "grad_norm": 2.445490540566737,
+      "learning_rate": 2.421225798845894e-06,
+      "loss": 0.3105,
+      "step": 10202
+    },
+    {
+      "epoch": 1.5612853863810252,
+      "grad_norm": 2.1391009903185396,
+      "learning_rate": 2.4196092724517763e-06,
+      "loss": 0.2995,
+      "step": 10203
+    },
+    {
+      "epoch": 1.5614384085692425,
+      "grad_norm": 2.044543683226895,
+      "learning_rate": 2.4179932115991056e-06,
+      "loss": 0.2764,
+      "step": 10204
+    },
+    {
+      "epoch": 1.5615914307574599,
+      "grad_norm": 2.046036413929888,
+      "learning_rate": 2.4163776163871334e-06,
+      "loss": 0.2786,
+      "step": 10205
+    },
+    {
+      "epoch": 1.561744452945677,
+      "grad_norm": 2.2680242817870875,
+      "learning_rate": 2.41476248691508e-06,
+      "loss": 0.3337,
+      "step": 10206
+    },
+    {
+      "epoch": 1.5618974751338945,
+      "grad_norm": 2.0380320601865853,
+      "learning_rate": 2.413147823282135e-06,
+      "loss": 0.33,
+      "step": 10207
+    },
+    {
+      "epoch": 1.5620504973221117,
+      "grad_norm": 2.040382014149729,
+      "learning_rate": 2.4115336255874623e-06,
+      "loss": 0.2632,
+      "step": 10208
+    },
+    {
+      "epoch": 1.562203519510329,
+      "grad_norm": 2.2044053792612632,
+      "learning_rate": 2.409919893930196e-06,
+      "loss": 0.2983,
+      "step": 10209
+    },
+    {
+      "epoch": 1.5623565416985463,
+      "grad_norm": 2.277740664709357,
+      "learning_rate": 2.4083066284094415e-06,
+      "loss": 0.3382,
+      "step": 10210
+    },
+    {
+      "epoch": 1.5625095638867634,
+      "grad_norm": 2.373251287168811,
+      "learning_rate": 2.406693829124276e-06,
+      "loss": 0.3217,
+      "step": 10211
+    },
+    {
+      "epoch": 1.562662586074981,
+      "grad_norm": 2.196277994030039,
+      "learning_rate": 2.4050814961737466e-06,
+      "loss": 0.3034,
+      "step": 10212
+    },
+    {
+      "epoch": 1.562815608263198,
+      "grad_norm": 1.998217282132098,
+      "learning_rate": 2.403469629656875e-06,
+      "loss": 0.2597,
+      "step": 10213
+    },
+    {
+      "epoch": 1.5629686304514154,
+      "grad_norm": 1.9683414339934968,
+      "learning_rate": 2.401858229672651e-06,
+      "loss": 0.3089,
+      "step": 10214
+    },
+    {
+      "epoch": 1.5631216526396328,
+      "grad_norm": 2.171320953900138,
+      "learning_rate": 2.4002472963200374e-06,
+      "loss": 0.3134,
+      "step": 10215
+    },
+    {
+      "epoch": 1.5632746748278499,
+      "grad_norm": 1.9236456967890725,
+      "learning_rate": 2.3986368296979643e-06,
+      "loss": 0.2783,
+      "step": 10216
+    },
+    {
+      "epoch": 1.5634276970160674,
+      "grad_norm": 2.1561594723846205,
+      "learning_rate": 2.397026829905347e-06,
+      "loss": 0.3362,
+      "step": 10217
+    },
+    {
+      "epoch": 1.5635807192042845,
+      "grad_norm": 2.0638153028428294,
+      "learning_rate": 2.395417297041052e-06,
+      "loss": 0.2939,
+      "step": 10218
+    },
+    {
+      "epoch": 1.5637337413925019,
+      "grad_norm": 2.075780646463065,
+      "learning_rate": 2.3938082312039267e-06,
+      "loss": 0.3216,
+      "step": 10219
+    },
+    {
+      "epoch": 1.5638867635807192,
+      "grad_norm": 2.147494836310276,
+      "learning_rate": 2.3921996324927988e-06,
+      "loss": 0.3036,
+      "step": 10220
+    },
+    {
+      "epoch": 1.5640397857689365,
+      "grad_norm": 2.494807687598165,
+      "learning_rate": 2.390591501006452e-06,
+      "loss": 0.4179,
+      "step": 10221
+    },
+    {
+      "epoch": 1.5641928079571539,
+      "grad_norm": 2.2762875943975107,
+      "learning_rate": 2.3889838368436445e-06,
+      "loss": 0.331,
+      "step": 10222
+    },
+    {
+      "epoch": 1.564345830145371,
+      "grad_norm": 2.0772333504973384,
+      "learning_rate": 2.3873766401031205e-06,
+      "loss": 0.2941,
+      "step": 10223
+    },
+    {
+      "epoch": 1.5644988523335883,
+      "grad_norm": 2.1005853247996313,
+      "learning_rate": 2.385769910883573e-06,
+      "loss": 0.2845,
+      "step": 10224
+    },
+    {
+      "epoch": 1.5646518745218057,
+      "grad_norm": 2.1720104480820885,
+      "learning_rate": 2.3841636492836838e-06,
+      "loss": 0.3295,
+      "step": 10225
+    },
+    {
+      "epoch": 1.564804896710023,
+      "grad_norm": 2.0423208854212,
+      "learning_rate": 2.382557855402102e-06,
+      "loss": 0.2516,
+      "step": 10226
+    },
+    {
+      "epoch": 1.5649579188982403,
+      "grad_norm": 2.0188972984497027,
+      "learning_rate": 2.3809525293374357e-06,
+      "loss": 0.2452,
+      "step": 10227
+    },
+    {
+      "epoch": 1.5651109410864574,
+      "grad_norm": 2.0508236985553436,
+      "learning_rate": 2.3793476711882836e-06,
+      "loss": 0.2987,
+      "step": 10228
+    },
+    {
+      "epoch": 1.565263963274675,
+      "grad_norm": 2.232455824019605,
+      "learning_rate": 2.377743281053203e-06,
+      "loss": 0.2617,
+      "step": 10229
+    },
+    {
+      "epoch": 1.565416985462892,
+      "grad_norm": 2.2874276475622795,
+      "learning_rate": 2.3761393590307267e-06,
+      "loss": 0.3049,
+      "step": 10230
+    },
+    {
+      "epoch": 1.5655700076511094,
+      "grad_norm": 2.0678605418290714,
+      "learning_rate": 2.374535905219356e-06,
+      "loss": 0.3362,
+      "step": 10231
+    },
+    {
+      "epoch": 1.5657230298393268,
+      "grad_norm": 1.9479610053175493,
+      "learning_rate": 2.3729329197175668e-06,
+      "loss": 0.3345,
+      "step": 10232
+    },
+    {
+      "epoch": 1.5658760520275439,
+      "grad_norm": 2.115965723564736,
+      "learning_rate": 2.371330402623805e-06,
+      "loss": 0.3229,
+      "step": 10233
+    },
+    {
+      "epoch": 1.5660290742157614,
+      "grad_norm": 2.134915489865641,
+      "learning_rate": 2.3697283540364856e-06,
+      "loss": 0.3516,
+      "step": 10234
+    },
+    {
+      "epoch": 1.5661820964039785,
+      "grad_norm": 2.4074271755888987,
+      "learning_rate": 2.368126774053998e-06,
+      "loss": 0.2972,
+      "step": 10235
+    },
+    {
+      "epoch": 1.5663351185921959,
+      "grad_norm": 2.2511960150185435,
+      "learning_rate": 2.3665256627747012e-06,
+      "loss": 0.2815,
+      "step": 10236
+    },
+    {
+      "epoch": 1.5664881407804132,
+      "grad_norm": 2.0602330321925106,
+      "learning_rate": 2.3649250202969233e-06,
+      "loss": 0.2911,
+      "step": 10237
+    },
+    {
+      "epoch": 1.5666411629686303,
+      "grad_norm": 1.927634330204787,
+      "learning_rate": 2.363324846718974e-06,
+      "loss": 0.2757,
+      "step": 10238
+    },
+    {
+      "epoch": 1.5667941851568479,
+      "grad_norm": 2.2909319046053906,
+      "learning_rate": 2.3617251421391172e-06,
+      "loss": 0.2859,
+      "step": 10239
+    },
+    {
+      "epoch": 1.566947207345065,
+      "grad_norm": 1.8964456203975408,
+      "learning_rate": 2.3601259066555982e-06,
+      "loss": 0.2124,
+      "step": 10240
+    },
+    {
+      "epoch": 1.5671002295332823,
+      "grad_norm": 1.9433830612323009,
+      "learning_rate": 2.358527140366641e-06,
+      "loss": 0.2759,
+      "step": 10241
+    },
+    {
+      "epoch": 1.5672532517214997,
+      "grad_norm": 2.1053810325565174,
+      "learning_rate": 2.356928843370422e-06,
+      "loss": 0.3441,
+      "step": 10242
+    },
+    {
+      "epoch": 1.5674062739097168,
+      "grad_norm": 2.1752026446050117,
+      "learning_rate": 2.3553310157651e-06,
+      "loss": 0.3235,
+      "step": 10243
+    },
+    {
+      "epoch": 1.5675592960979343,
+      "grad_norm": 2.1045113271973603,
+      "learning_rate": 2.3537336576488124e-06,
+      "loss": 0.2883,
+      "step": 10244
+    },
+    {
+      "epoch": 1.5677123182861514,
+      "grad_norm": 2.1417921242316855,
+      "learning_rate": 2.3521367691196474e-06,
+      "loss": 0.4201,
+      "step": 10245
+    },
+    {
+      "epoch": 1.5678653404743688,
+      "grad_norm": 2.2202253891721924,
+      "learning_rate": 2.350540350275684e-06,
+      "loss": 0.344,
+      "step": 10246
+    },
+    {
+      "epoch": 1.568018362662586,
+      "grad_norm": 2.051083146052781,
+      "learning_rate": 2.3489444012149665e-06,
+      "loss": 0.3307,
+      "step": 10247
+    },
+    {
+      "epoch": 1.5681713848508032,
+      "grad_norm": 2.3576621657679113,
+      "learning_rate": 2.3473489220354985e-06,
+      "loss": 0.2906,
+      "step": 10248
+    },
+    {
+      "epoch": 1.5683244070390208,
+      "grad_norm": 2.400771637167478,
+      "learning_rate": 2.3457539128352737e-06,
+      "loss": 0.3003,
+      "step": 10249
+    },
+    {
+      "epoch": 1.5684774292272379,
+      "grad_norm": 2.042581493431355,
+      "learning_rate": 2.344159373712247e-06,
+      "loss": 0.2974,
+      "step": 10250
+    },
+    {
+      "epoch": 1.5686304514154552,
+      "grad_norm": 2.346004763894137,
+      "learning_rate": 2.3425653047643373e-06,
+      "loss": 0.3402,
+      "step": 10251
+    },
+    {
+      "epoch": 1.5687834736036725,
+      "grad_norm": 1.6993995033239386,
+      "learning_rate": 2.340971706089451e-06,
+      "loss": 0.2819,
+      "step": 10252
+    },
+    {
+      "epoch": 1.5689364957918899,
+      "grad_norm": 2.3292922099877766,
+      "learning_rate": 2.339378577785455e-06,
+      "loss": 0.3639,
+      "step": 10253
+    },
+    {
+      "epoch": 1.5690895179801072,
+      "grad_norm": 1.7548228300139943,
+      "learning_rate": 2.3377859199501886e-06,
+      "loss": 0.2774,
+      "step": 10254
+    },
+    {
+      "epoch": 1.5692425401683243,
+      "grad_norm": 2.192534611433887,
+      "learning_rate": 2.3361937326814633e-06,
+      "loss": 0.3263,
+      "step": 10255
+    },
+    {
+      "epoch": 1.5693955623565417,
+      "grad_norm": 2.262468415284445,
+      "learning_rate": 2.3346020160770632e-06,
+      "loss": 0.2805,
+      "step": 10256
+    },
+    {
+      "epoch": 1.569548584544759,
+      "grad_norm": 2.10740520424962,
+      "learning_rate": 2.3330107702347393e-06,
+      "loss": 0.3117,
+      "step": 10257
+    },
+    {
+      "epoch": 1.5697016067329763,
+      "grad_norm": 1.9254659842503297,
+      "learning_rate": 2.3314199952522176e-06,
+      "loss": 0.2384,
+      "step": 10258
+    },
+    {
+      "epoch": 1.5698546289211937,
+      "grad_norm": 2.3472990294430174,
+      "learning_rate": 2.3298296912271932e-06,
+      "loss": 0.3014,
+      "step": 10259
+    },
+    {
+      "epoch": 1.5700076511094108,
+      "grad_norm": 2.123204083750397,
+      "learning_rate": 2.328239858257335e-06,
+      "loss": 0.3593,
+      "step": 10260
+    },
+    {
+      "epoch": 1.5701606732976283,
+      "grad_norm": 2.3856229695225633,
+      "learning_rate": 2.326650496440278e-06,
+      "loss": 0.3334,
+      "step": 10261
+    },
+    {
+      "epoch": 1.5703136954858454,
+      "grad_norm": 2.033254452554158,
+      "learning_rate": 2.325061605873632e-06,
+      "loss": 0.3196,
+      "step": 10262
+    },
+    {
+      "epoch": 1.5704667176740628,
+      "grad_norm": 2.384658331345606,
+      "learning_rate": 2.3234731866549778e-06,
+      "loss": 0.2796,
+      "step": 10263
+    },
+    {
+      "epoch": 1.57061973986228,
+      "grad_norm": 1.9582659952780006,
+      "learning_rate": 2.3218852388818657e-06,
+      "loss": 0.2933,
+      "step": 10264
+    },
+    {
+      "epoch": 1.5707727620504972,
+      "grad_norm": 2.1466251525953166,
+      "learning_rate": 2.3202977626518187e-06,
+      "loss": 0.2537,
+      "step": 10265
+    },
+    {
+      "epoch": 1.5709257842387148,
+      "grad_norm": 2.072612975719027,
+      "learning_rate": 2.3187107580623257e-06,
+      "loss": 0.3466,
+      "step": 10266
+    },
+    {
+      "epoch": 1.5710788064269319,
+      "grad_norm": 1.889187195089687,
+      "learning_rate": 2.3171242252108607e-06,
+      "loss": 0.3119,
+      "step": 10267
+    },
+    {
+      "epoch": 1.5712318286151492,
+      "grad_norm": 2.0271223685590263,
+      "learning_rate": 2.3155381641948494e-06,
+      "loss": 0.3211,
+      "step": 10268
+    },
+    {
+      "epoch": 1.5713848508033665,
+      "grad_norm": 2.0742514867228707,
+      "learning_rate": 2.313952575111699e-06,
+      "loss": 0.3236,
+      "step": 10269
+    },
+    {
+      "epoch": 1.5715378729915837,
+      "grad_norm": 2.4083805585663085,
+      "learning_rate": 2.3123674580587942e-06,
+      "loss": 0.3223,
+      "step": 10270
+    },
+    {
+      "epoch": 1.5716908951798012,
+      "grad_norm": 1.7294525210795593,
+      "learning_rate": 2.3107828131334744e-06,
+      "loss": 0.2704,
+      "step": 10271
+    },
+    {
+      "epoch": 1.5718439173680183,
+      "grad_norm": 2.3189317261675297,
+      "learning_rate": 2.30919864043306e-06,
+      "loss": 0.3368,
+      "step": 10272
+    },
+    {
+      "epoch": 1.5719969395562357,
+      "grad_norm": 2.194260512799246,
+      "learning_rate": 2.3076149400548498e-06,
+      "loss": 0.3096,
+      "step": 10273
+    },
+    {
+      "epoch": 1.572149961744453,
+      "grad_norm": 1.8457396636416614,
+      "learning_rate": 2.306031712096093e-06,
+      "loss": 0.2729,
+      "step": 10274
+    },
+    {
+      "epoch": 1.57230298393267,
+      "grad_norm": 2.084375191930075,
+      "learning_rate": 2.3044489566540306e-06,
+      "loss": 0.3001,
+      "step": 10275
+    },
+    {
+      "epoch": 1.5724560061208877,
+      "grad_norm": 1.91765039031427,
+      "learning_rate": 2.3028666738258653e-06,
+      "loss": 0.2202,
+      "step": 10276
+    },
+    {
+      "epoch": 1.5726090283091048,
+      "grad_norm": 2.1171884440790776,
+      "learning_rate": 2.301284863708764e-06,
+      "loss": 0.303,
+      "step": 10277
+    },
+    {
+      "epoch": 1.572762050497322,
+      "grad_norm": 2.1586507600026392,
+      "learning_rate": 2.2997035263998792e-06,
+      "loss": 0.3683,
+      "step": 10278
+    },
+    {
+      "epoch": 1.5729150726855394,
+      "grad_norm": 2.3001386303828015,
+      "learning_rate": 2.298122661996328e-06,
+      "loss": 0.3143,
+      "step": 10279
+    },
+    {
+      "epoch": 1.5730680948737565,
+      "grad_norm": 2.340415887451529,
+      "learning_rate": 2.296542270595188e-06,
+      "loss": 0.3019,
+      "step": 10280
+    },
+    {
+      "epoch": 1.573221117061974,
+      "grad_norm": 2.2975152673572947,
+      "learning_rate": 2.294962352293526e-06,
+      "loss": 0.3047,
+      "step": 10281
+    },
+    {
+      "epoch": 1.5733741392501912,
+      "grad_norm": 1.7612950905997558,
+      "learning_rate": 2.2933829071883673e-06,
+      "loss": 0.231,
+      "step": 10282
+    },
+    {
+      "epoch": 1.5735271614384085,
+      "grad_norm": 2.139252431583115,
+      "learning_rate": 2.291803935376714e-06,
+      "loss": 0.3253,
+      "step": 10283
+    },
+    {
+      "epoch": 1.5736801836266259,
+      "grad_norm": 2.2296485507686836,
+      "learning_rate": 2.2902254369555354e-06,
+      "loss": 0.3236,
+      "step": 10284
+    },
+    {
+      "epoch": 1.5738332058148432,
+      "grad_norm": 2.230439426279258,
+      "learning_rate": 2.2886474120217726e-06,
+      "loss": 0.3202,
+      "step": 10285
+    },
+    {
+      "epoch": 1.5739862280030605,
+      "grad_norm": 2.194048175922532,
+      "learning_rate": 2.287069860672341e-06,
+      "loss": 0.3116,
+      "step": 10286
+    },
+    {
+      "epoch": 1.5741392501912777,
+      "grad_norm": 2.104361793722446,
+      "learning_rate": 2.2854927830041205e-06,
+      "loss": 0.3093,
+      "step": 10287
+    },
+    {
+      "epoch": 1.574292272379495,
+      "grad_norm": 2.190600186554258,
+      "learning_rate": 2.2839161791139685e-06,
+      "loss": 0.2952,
+      "step": 10288
+    },
+    {
+      "epoch": 1.5744452945677123,
+      "grad_norm": 2.2380497722604566,
+      "learning_rate": 2.2823400490987103e-06,
+      "loss": 0.2993,
+      "step": 10289
+    },
+    {
+      "epoch": 1.5745983167559297,
+      "grad_norm": 2.0662226158733907,
+      "learning_rate": 2.2807643930551403e-06,
+      "loss": 0.2838,
+      "step": 10290
+    },
+    {
+      "epoch": 1.574751338944147,
+      "grad_norm": 1.8754728763422386,
+      "learning_rate": 2.279189211080026e-06,
+      "loss": 0.2747,
+      "step": 10291
+    },
+    {
+      "epoch": 1.574904361132364,
+      "grad_norm": 2.239515204953391,
+      "learning_rate": 2.277614503270108e-06,
+      "loss": 0.3475,
+      "step": 10292
+    },
+    {
+      "epoch": 1.5750573833205816,
+      "grad_norm": 1.9798738633771396,
+      "learning_rate": 2.276040269722092e-06,
+      "loss": 0.3008,
+      "step": 10293
+    },
+    {
+      "epoch": 1.5752104055087988,
+      "grad_norm": 2.2716397492829454,
+      "learning_rate": 2.2744665105326603e-06,
+      "loss": 0.3376,
+      "step": 10294
+    },
+    {
+      "epoch": 1.575363427697016,
+      "grad_norm": 2.2329279492017795,
+      "learning_rate": 2.2728932257984613e-06,
+      "loss": 0.3152,
+      "step": 10295
+    },
+    {
+      "epoch": 1.5755164498852334,
+      "grad_norm": 1.999846969202547,
+      "learning_rate": 2.2713204156161193e-06,
+      "loss": 0.3223,
+      "step": 10296
+    },
+    {
+      "epoch": 1.5756694720734505,
+      "grad_norm": 2.186381035023827,
+      "learning_rate": 2.269748080082225e-06,
+      "loss": 0.3592,
+      "step": 10297
+    },
+    {
+      "epoch": 1.575822494261668,
+      "grad_norm": 2.1638727598497742,
+      "learning_rate": 2.268176219293339e-06,
+      "loss": 0.3635,
+      "step": 10298
+    },
+    {
+      "epoch": 1.5759755164498852,
+      "grad_norm": 1.9826793258017292,
+      "learning_rate": 2.2666048333460046e-06,
+      "loss": 0.3428,
+      "step": 10299
+    },
+    {
+      "epoch": 1.5761285386381025,
+      "grad_norm": 2.1235423137566536,
+      "learning_rate": 2.2650339223367167e-06,
+      "loss": 0.2813,
+      "step": 10300
+    },
+    {
+      "epoch": 1.5762815608263199,
+      "grad_norm": 2.150327660568215,
+      "learning_rate": 2.263463486361953e-06,
+      "loss": 0.3257,
+      "step": 10301
+    },
+    {
+      "epoch": 1.576434583014537,
+      "grad_norm": 2.1608998717546952,
+      "learning_rate": 2.2618935255181673e-06,
+      "loss": 0.3238,
+      "step": 10302
+    },
+    {
+      "epoch": 1.5765876052027545,
+      "grad_norm": 2.163568899899916,
+      "learning_rate": 2.2603240399017668e-06,
+      "loss": 0.2742,
+      "step": 10303
+    },
+    {
+      "epoch": 1.5767406273909716,
+      "grad_norm": 2.0529341198278206,
+      "learning_rate": 2.2587550296091477e-06,
+      "loss": 0.308,
+      "step": 10304
+    },
+    {
+      "epoch": 1.576893649579189,
+      "grad_norm": 2.0794813595152464,
+      "learning_rate": 2.2571864947366685e-06,
+      "loss": 0.3075,
+      "step": 10305
+    },
+    {
+      "epoch": 1.5770466717674063,
+      "grad_norm": 2.0384171796033796,
+      "learning_rate": 2.255618435380651e-06,
+      "loss": 0.3338,
+      "step": 10306
+    },
+    {
+      "epoch": 1.5771996939556234,
+      "grad_norm": 2.226350908567524,
+      "learning_rate": 2.2540508516374036e-06,
+      "loss": 0.3045,
+      "step": 10307
+    },
+    {
+      "epoch": 1.577352716143841,
+      "grad_norm": 2.0631861942769785,
+      "learning_rate": 2.2524837436031997e-06,
+      "loss": 0.2344,
+      "step": 10308
+    },
+    {
+      "epoch": 1.577505738332058,
+      "grad_norm": 2.3266694287378527,
+      "learning_rate": 2.2509171113742724e-06,
+      "loss": 0.3964,
+      "step": 10309
+    },
+    {
+      "epoch": 1.5776587605202754,
+      "grad_norm": 1.9944872958494284,
+      "learning_rate": 2.249350955046842e-06,
+      "loss": 0.2898,
+      "step": 10310
+    },
+    {
+      "epoch": 1.5778117827084928,
+      "grad_norm": 1.9506141300360098,
+      "learning_rate": 2.24778527471709e-06,
+      "loss": 0.2996,
+      "step": 10311
+    },
+    {
+      "epoch": 1.5779648048967099,
+      "grad_norm": 1.9074519577115954,
+      "learning_rate": 2.246220070481171e-06,
+      "loss": 0.2978,
+      "step": 10312
+    },
+    {
+      "epoch": 1.5781178270849274,
+      "grad_norm": 1.9938550842973741,
+      "learning_rate": 2.24465534243521e-06,
+      "loss": 0.2799,
+      "step": 10313
+    },
+    {
+      "epoch": 1.5782708492731445,
+      "grad_norm": 1.8507728107065444,
+      "learning_rate": 2.2430910906753045e-06,
+      "loss": 0.3082,
+      "step": 10314
+    },
+    {
+      "epoch": 1.5784238714613619,
+      "grad_norm": 1.9282543196104605,
+      "learning_rate": 2.2415273152975205e-06,
+      "loss": 0.2698,
+      "step": 10315
+    },
+    {
+      "epoch": 1.5785768936495792,
+      "grad_norm": 2.120689019338789,
+      "learning_rate": 2.2399640163978942e-06,
+      "loss": 0.316,
+      "step": 10316
+    },
+    {
+      "epoch": 1.5787299158377963,
+      "grad_norm": 2.193624778585067,
+      "learning_rate": 2.238401194072436e-06,
+      "loss": 0.2874,
+      "step": 10317
+    },
+    {
+      "epoch": 1.5788829380260139,
+      "grad_norm": 2.302872403542241,
+      "learning_rate": 2.2368388484171246e-06,
+      "loss": 0.3184,
+      "step": 10318
+    },
+    {
+      "epoch": 1.579035960214231,
+      "grad_norm": 1.642833529579855,
+      "learning_rate": 2.235276979527905e-06,
+      "loss": 0.2235,
+      "step": 10319
+    },
+    {
+      "epoch": 1.5791889824024483,
+      "grad_norm": 2.2481428159239334,
+      "learning_rate": 2.2337155875007076e-06,
+      "loss": 0.2923,
+      "step": 10320
+    },
+    {
+      "epoch": 1.5793420045906656,
+      "grad_norm": 1.9982154763461255,
+      "learning_rate": 2.232154672431416e-06,
+      "loss": 0.2771,
+      "step": 10321
+    },
+    {
+      "epoch": 1.579495026778883,
+      "grad_norm": 2.013720888615236,
+      "learning_rate": 2.2305942344158906e-06,
+      "loss": 0.276,
+      "step": 10322
+    },
+    {
+      "epoch": 1.5796480489671003,
+      "grad_norm": 2.068002598244875,
+      "learning_rate": 2.2290342735499724e-06,
+      "loss": 0.2944,
+      "step": 10323
+    },
+    {
+      "epoch": 1.5798010711553174,
+      "grad_norm": 1.9030134400531125,
+      "learning_rate": 2.227474789929458e-06,
+      "loss": 0.2344,
+      "step": 10324
+    },
+    {
+      "epoch": 1.5799540933435348,
+      "grad_norm": 2.022291552569888,
+      "learning_rate": 2.225915783650119e-06,
+      "loss": 0.2472,
+      "step": 10325
+    },
+    {
+      "epoch": 1.580107115531752,
+      "grad_norm": 2.24395570623936,
+      "learning_rate": 2.2243572548077107e-06,
+      "loss": 0.3125,
+      "step": 10326
+    },
+    {
+      "epoch": 1.5802601377199694,
+      "grad_norm": 1.9891422743630058,
+      "learning_rate": 2.2227992034979363e-06,
+      "loss": 0.2601,
+      "step": 10327
+    },
+    {
+      "epoch": 1.5804131599081868,
+      "grad_norm": 1.8983232454344225,
+      "learning_rate": 2.2212416298164895e-06,
+      "loss": 0.3107,
+      "step": 10328
+    },
+    {
+      "epoch": 1.5805661820964039,
+      "grad_norm": 1.7352050344116838,
+      "learning_rate": 2.219684533859028e-06,
+      "loss": 0.2608,
+      "step": 10329
+    },
+    {
+      "epoch": 1.5807192042846214,
+      "grad_norm": 2.0679773793018636,
+      "learning_rate": 2.21812791572117e-06,
+      "loss": 0.2791,
+      "step": 10330
+    },
+    {
+      "epoch": 1.5808722264728385,
+      "grad_norm": 2.0633576159015488,
+      "learning_rate": 2.216571775498523e-06,
+      "loss": 0.3305,
+      "step": 10331
+    },
+    {
+      "epoch": 1.5810252486610559,
+      "grad_norm": 2.0864993072287645,
+      "learning_rate": 2.215016113286652e-06,
+      "loss": 0.2381,
+      "step": 10332
+    },
+    {
+      "epoch": 1.5811782708492732,
+      "grad_norm": 1.9834409122072287,
+      "learning_rate": 2.213460929181097e-06,
+      "loss": 0.3014,
+      "step": 10333
+    },
+    {
+      "epoch": 1.5813312930374903,
+      "grad_norm": 2.2354577819111445,
+      "learning_rate": 2.211906223277367e-06,
+      "loss": 0.2997,
+      "step": 10334
+    },
+    {
+      "epoch": 1.5814843152257079,
+      "grad_norm": 2.2939978595458363,
+      "learning_rate": 2.210351995670943e-06,
+      "loss": 0.318,
+      "step": 10335
+    },
+    {
+      "epoch": 1.581637337413925,
+      "grad_norm": 2.019619517073732,
+      "learning_rate": 2.2087982464572755e-06,
+      "loss": 0.3068,
+      "step": 10336
+    },
+    {
+      "epoch": 1.5817903596021423,
+      "grad_norm": 2.1592548402626175,
+      "learning_rate": 2.207244975731788e-06,
+      "loss": 0.255,
+      "step": 10337
+    },
+    {
+      "epoch": 1.5819433817903596,
+      "grad_norm": 2.0626900857490402,
+      "learning_rate": 2.2056921835898717e-06,
+      "loss": 0.2745,
+      "step": 10338
+    },
+    {
+      "epoch": 1.5820964039785768,
+      "grad_norm": 2.318142925818211,
+      "learning_rate": 2.20413987012689e-06,
+      "loss": 0.3219,
+      "step": 10339
+    },
+    {
+      "epoch": 1.5822494261667943,
+      "grad_norm": 1.997112063237294,
+      "learning_rate": 2.2025880354381767e-06,
+      "loss": 0.3089,
+      "step": 10340
+    },
+    {
+      "epoch": 1.5824024483550114,
+      "grad_norm": 2.284517687797388,
+      "learning_rate": 2.2010366796190353e-06,
+      "loss": 0.3464,
+      "step": 10341
+    },
+    {
+      "epoch": 1.5825554705432288,
+      "grad_norm": 2.670105379186421,
+      "learning_rate": 2.1994858027647414e-06,
+      "loss": 0.3209,
+      "step": 10342
+    },
+    {
+      "epoch": 1.582708492731446,
+      "grad_norm": 2.0680620302934094,
+      "learning_rate": 2.197935404970539e-06,
+      "loss": 0.2762,
+      "step": 10343
+    },
+    {
+      "epoch": 1.5828615149196632,
+      "grad_norm": 2.0968791504944897,
+      "learning_rate": 2.1963854863316468e-06,
+      "loss": 0.2899,
+      "step": 10344
+    },
+    {
+      "epoch": 1.5830145371078808,
+      "grad_norm": 2.0214319597523103,
+      "learning_rate": 2.1948360469432494e-06,
+      "loss": 0.3563,
+      "step": 10345
+    },
+    {
+      "epoch": 1.5831675592960979,
+      "grad_norm": 2.131484039985648,
+      "learning_rate": 2.1932870869005042e-06,
+      "loss": 0.3074,
+      "step": 10346
+    },
+    {
+      "epoch": 1.5833205814843152,
+      "grad_norm": 2.103344766881095,
+      "learning_rate": 2.191738606298539e-06,
+      "loss": 0.3058,
+      "step": 10347
+    },
+    {
+      "epoch": 1.5834736036725325,
+      "grad_norm": 2.556940059879879,
+      "learning_rate": 2.1901906052324495e-06,
+      "loss": 0.4076,
+      "step": 10348
+    },
+    {
+      "epoch": 1.5836266258607496,
+      "grad_norm": 2.2405335632188517,
+      "learning_rate": 2.1886430837973115e-06,
+      "loss": 0.3214,
+      "step": 10349
+    },
+    {
+      "epoch": 1.5837796480489672,
+      "grad_norm": 2.14456826606491,
+      "learning_rate": 2.1870960420881584e-06,
+      "loss": 0.3017,
+      "step": 10350
+    },
+    {
+      "epoch": 1.5839326702371843,
+      "grad_norm": 2.2376186462192096,
+      "learning_rate": 2.185549480199999e-06,
+      "loss": 0.2989,
+      "step": 10351
+    },
+    {
+      "epoch": 1.5840856924254016,
+      "grad_norm": 2.05365198255844,
+      "learning_rate": 2.184003398227821e-06,
+      "loss": 0.2639,
+      "step": 10352
+    },
+    {
+      "epoch": 1.584238714613619,
+      "grad_norm": 1.5879002011757888,
+      "learning_rate": 2.182457796266568e-06,
+      "loss": 0.2415,
+      "step": 10353
+    },
+    {
+      "epoch": 1.5843917368018363,
+      "grad_norm": 2.2942998647330173,
+      "learning_rate": 2.180912674411162e-06,
+      "loss": 0.2657,
+      "step": 10354
+    },
+    {
+      "epoch": 1.5845447589900536,
+      "grad_norm": 2.1415116546843294,
+      "learning_rate": 2.1793680327565024e-06,
+      "loss": 0.3094,
+      "step": 10355
+    },
+    {
+      "epoch": 1.5846977811782708,
+      "grad_norm": 1.868653381216291,
+      "learning_rate": 2.177823871397441e-06,
+      "loss": 0.2852,
+      "step": 10356
+    },
+    {
+      "epoch": 1.584850803366488,
+      "grad_norm": 2.309292831698454,
+      "learning_rate": 2.176280190428819e-06,
+      "loss": 0.3522,
+      "step": 10357
+    },
+    {
+      "epoch": 1.5850038255547054,
+      "grad_norm": 2.1710738028332863,
+      "learning_rate": 2.1747369899454386e-06,
+      "loss": 0.3116,
+      "step": 10358
+    },
+    {
+      "epoch": 1.5851568477429228,
+      "grad_norm": 2.0734244476848454,
+      "learning_rate": 2.1731942700420683e-06,
+      "loss": 0.2549,
+      "step": 10359
+    },
+    {
+      "epoch": 1.58530986993114,
+      "grad_norm": 2.351848876345775,
+      "learning_rate": 2.171652030813458e-06,
+      "loss": 0.325,
+      "step": 10360
+    },
+    {
+      "epoch": 1.5854628921193572,
+      "grad_norm": 2.257892033882186,
+      "learning_rate": 2.1701102723543242e-06,
+      "loss": 0.2825,
+      "step": 10361
+    },
+    {
+      "epoch": 1.5856159143075748,
+      "grad_norm": 2.0733014553722517,
+      "learning_rate": 2.1685689947593445e-06,
+      "loss": 0.3186,
+      "step": 10362
+    },
+    {
+      "epoch": 1.5857689364957919,
+      "grad_norm": 2.1921386004476706,
+      "learning_rate": 2.167028198123182e-06,
+      "loss": 0.332,
+      "step": 10363
+    },
+    {
+      "epoch": 1.5859219586840092,
+      "grad_norm": 2.220723693801886,
+      "learning_rate": 2.16548788254046e-06,
+      "loss": 0.3319,
+      "step": 10364
+    },
+    {
+      "epoch": 1.5860749808722265,
+      "grad_norm": 2.0745477917442696,
+      "learning_rate": 2.1639480481057774e-06,
+      "loss": 0.2772,
+      "step": 10365
+    },
+    {
+      "epoch": 1.5862280030604436,
+      "grad_norm": 2.1995867088192798,
+      "learning_rate": 2.162408694913699e-06,
+      "loss": 0.322,
+      "step": 10366
+    },
+    {
+      "epoch": 1.5863810252486612,
+      "grad_norm": 2.029684586828926,
+      "learning_rate": 2.160869823058763e-06,
+      "loss": 0.2683,
+      "step": 10367
+    },
+    {
+      "epoch": 1.5865340474368783,
+      "grad_norm": 2.316636320242865,
+      "learning_rate": 2.1593314326354787e-06,
+      "loss": 0.3374,
+      "step": 10368
+    },
+    {
+      "epoch": 1.5866870696250956,
+      "grad_norm": 2.2433828028014045,
+      "learning_rate": 2.157793523738324e-06,
+      "loss": 0.3329,
+      "step": 10369
+    },
+    {
+      "epoch": 1.586840091813313,
+      "grad_norm": 1.9832856482770373,
+      "learning_rate": 2.1562560964617473e-06,
+      "loss": 0.2735,
+      "step": 10370
+    },
+    {
+      "epoch": 1.58699311400153,
+      "grad_norm": 2.1848410328916987,
+      "learning_rate": 2.1547191509001687e-06,
+      "loss": 0.2742,
+      "step": 10371
+    },
+    {
+      "epoch": 1.5871461361897476,
+      "grad_norm": 2.118690825527389,
+      "learning_rate": 2.1531826871479787e-06,
+      "loss": 0.3074,
+      "step": 10372
+    },
+    {
+      "epoch": 1.5872991583779648,
+      "grad_norm": 2.3314417826081018,
+      "learning_rate": 2.151646705299536e-06,
+      "loss": 0.2814,
+      "step": 10373
+    },
+    {
+      "epoch": 1.587452180566182,
+      "grad_norm": 2.402232417525519,
+      "learning_rate": 2.1501112054491725e-06,
+      "loss": 0.3494,
+      "step": 10374
+    },
+    {
+      "epoch": 1.5876052027543994,
+      "grad_norm": 2.326820597715731,
+      "learning_rate": 2.1485761876911892e-06,
+      "loss": 0.3038,
+      "step": 10375
+    },
+    {
+      "epoch": 1.5877582249426165,
+      "grad_norm": 2.1383455574598713,
+      "learning_rate": 2.1470416521198567e-06,
+      "loss": 0.352,
+      "step": 10376
+    },
+    {
+      "epoch": 1.587911247130834,
+      "grad_norm": 2.2995939874004625,
+      "learning_rate": 2.1455075988294137e-06,
+      "loss": 0.3279,
+      "step": 10377
+    },
+    {
+      "epoch": 1.5880642693190512,
+      "grad_norm": 1.9137006604044404,
+      "learning_rate": 2.143974027914083e-06,
+      "loss": 0.3033,
+      "step": 10378
+    },
+    {
+      "epoch": 1.5882172915072685,
+      "grad_norm": 2.2833225627008074,
+      "learning_rate": 2.142440939468037e-06,
+      "loss": 0.3236,
+      "step": 10379
+    },
+    {
+      "epoch": 1.5883703136954859,
+      "grad_norm": 1.9860651109123406,
+      "learning_rate": 2.1409083335854287e-06,
+      "loss": 0.2383,
+      "step": 10380
+    },
+    {
+      "epoch": 1.588523335883703,
+      "grad_norm": 2.48380903765018,
+      "learning_rate": 2.1393762103603898e-06,
+      "loss": 0.3249,
+      "step": 10381
+    },
+    {
+      "epoch": 1.5886763580719205,
+      "grad_norm": 2.4131936637247184,
+      "learning_rate": 2.1378445698870064e-06,
+      "loss": 0.3444,
+      "step": 10382
+    },
+    {
+      "epoch": 1.5888293802601376,
+      "grad_norm": 2.1244925827371386,
+      "learning_rate": 2.136313412259342e-06,
+      "loss": 0.3347,
+      "step": 10383
+    },
+    {
+      "epoch": 1.588982402448355,
+      "grad_norm": 2.1275395552021252,
+      "learning_rate": 2.134782737571439e-06,
+      "loss": 0.2968,
+      "step": 10384
+    },
+    {
+      "epoch": 1.5891354246365723,
+      "grad_norm": 2.382525615883336,
+      "learning_rate": 2.1332525459172927e-06,
+      "loss": 0.3307,
+      "step": 10385
+    },
+    {
+      "epoch": 1.5892884468247896,
+      "grad_norm": 2.193605115589541,
+      "learning_rate": 2.131722837390885e-06,
+      "loss": 0.2865,
+      "step": 10386
+    },
+    {
+      "epoch": 1.589441469013007,
+      "grad_norm": 2.3407094562764215,
+      "learning_rate": 2.130193612086161e-06,
+      "loss": 0.3284,
+      "step": 10387
+    },
+    {
+      "epoch": 1.589594491201224,
+      "grad_norm": 2.213402841205263,
+      "learning_rate": 2.128664870097028e-06,
+      "loss": 0.31,
+      "step": 10388
+    },
+    {
+      "epoch": 1.5897475133894414,
+      "grad_norm": 2.0222607907984798,
+      "learning_rate": 2.127136611517382e-06,
+      "loss": 0.2872,
+      "step": 10389
+    },
+    {
+      "epoch": 1.5899005355776588,
+      "grad_norm": 2.2442947647359714,
+      "learning_rate": 2.1256088364410775e-06,
+      "loss": 0.2878,
+      "step": 10390
+    },
+    {
+      "epoch": 1.590053557765876,
+      "grad_norm": 2.240624536205986,
+      "learning_rate": 2.1240815449619335e-06,
+      "loss": 0.3405,
+      "step": 10391
+    },
+    {
+      "epoch": 1.5902065799540934,
+      "grad_norm": 2.1876775626139024,
+      "learning_rate": 2.1225547371737564e-06,
+      "loss": 0.3478,
+      "step": 10392
+    },
+    {
+      "epoch": 1.5903596021423105,
+      "grad_norm": 2.162284723354573,
+      "learning_rate": 2.1210284131703084e-06,
+      "loss": 0.3532,
+      "step": 10393
+    },
+    {
+      "epoch": 1.590512624330528,
+      "grad_norm": 2.1405508919523593,
+      "learning_rate": 2.119502573045329e-06,
+      "loss": 0.3358,
+      "step": 10394
+    },
+    {
+      "epoch": 1.5906656465187452,
+      "grad_norm": 1.9600807168130765,
+      "learning_rate": 2.117977216892525e-06,
+      "loss": 0.3341,
+      "step": 10395
+    },
+    {
+      "epoch": 1.5908186687069625,
+      "grad_norm": 2.314495390443742,
+      "learning_rate": 2.1164523448055752e-06,
+      "loss": 0.2895,
+      "step": 10396
+    },
+    {
+      "epoch": 1.5909716908951799,
+      "grad_norm": 2.201921905625236,
+      "learning_rate": 2.114927956878128e-06,
+      "loss": 0.2522,
+      "step": 10397
+    },
+    {
+      "epoch": 1.591124713083397,
+      "grad_norm": 2.268191757936805,
+      "learning_rate": 2.1134040532038e-06,
+      "loss": 0.3256,
+      "step": 10398
+    },
+    {
+      "epoch": 1.5912777352716145,
+      "grad_norm": 2.2869195645309968,
+      "learning_rate": 2.1118806338761823e-06,
+      "loss": 0.3375,
+      "step": 10399
+    },
+    {
+      "epoch": 1.5914307574598316,
+      "grad_norm": 2.0769558928508727,
+      "learning_rate": 2.110357698988834e-06,
+      "loss": 0.2647,
+      "step": 10400
+    },
+    {
+      "epoch": 1.591583779648049,
+      "grad_norm": 2.4711031090770286,
+      "learning_rate": 2.1088352486352802e-06,
+      "loss": 0.3858,
+      "step": 10401
+    },
+    {
+      "epoch": 1.5917368018362663,
+      "grad_norm": 2.119268678002257,
+      "learning_rate": 2.1073132829090305e-06,
+      "loss": 0.2872,
+      "step": 10402
+    },
+    {
+      "epoch": 1.5918898240244834,
+      "grad_norm": 1.9850108207669523,
+      "learning_rate": 2.105791801903545e-06,
+      "loss": 0.3069,
+      "step": 10403
+    },
+    {
+      "epoch": 1.592042846212701,
+      "grad_norm": 2.224819324782824,
+      "learning_rate": 2.104270805712265e-06,
+      "loss": 0.3448,
+      "step": 10404
+    },
+    {
+      "epoch": 1.592195868400918,
+      "grad_norm": 1.873089072531048,
+      "learning_rate": 2.1027502944286083e-06,
+      "loss": 0.2505,
+      "step": 10405
+    },
+    {
+      "epoch": 1.5923488905891354,
+      "grad_norm": 1.86192926757717,
+      "learning_rate": 2.101230268145944e-06,
+      "loss": 0.3051,
+      "step": 10406
+    },
+    {
+      "epoch": 1.5925019127773528,
+      "grad_norm": 1.9442480263553694,
+      "learning_rate": 2.099710726957633e-06,
+      "loss": 0.293,
+      "step": 10407
+    },
+    {
+      "epoch": 1.5926549349655699,
+      "grad_norm": 2.358985236520384,
+      "learning_rate": 2.098191670956995e-06,
+      "loss": 0.3361,
+      "step": 10408
+    },
+    {
+      "epoch": 1.5928079571537874,
+      "grad_norm": 2.114912121413648,
+      "learning_rate": 2.0966731002373132e-06,
+      "loss": 0.3205,
+      "step": 10409
+    },
+    {
+      "epoch": 1.5929609793420045,
+      "grad_norm": 2.324773058036772,
+      "learning_rate": 2.0951550148918566e-06,
+      "loss": 0.3892,
+      "step": 10410
+    },
+    {
+      "epoch": 1.5931140015302219,
+      "grad_norm": 1.8540830139440694,
+      "learning_rate": 2.093637415013857e-06,
+      "loss": 0.2622,
+      "step": 10411
+    },
+    {
+      "epoch": 1.5932670237184392,
+      "grad_norm": 1.901825704329433,
+      "learning_rate": 2.0921203006965086e-06,
+      "loss": 0.2557,
+      "step": 10412
+    },
+    {
+      "epoch": 1.5934200459066563,
+      "grad_norm": 2.1564336487480387,
+      "learning_rate": 2.0906036720329904e-06,
+      "loss": 0.3251,
+      "step": 10413
+    },
+    {
+      "epoch": 1.5935730680948739,
+      "grad_norm": 2.0254893602452304,
+      "learning_rate": 2.0890875291164425e-06,
+      "loss": 0.2756,
+      "step": 10414
+    },
+    {
+      "epoch": 1.593726090283091,
+      "grad_norm": 2.145515414045669,
+      "learning_rate": 2.087571872039977e-06,
+      "loss": 0.311,
+      "step": 10415
+    },
+    {
+      "epoch": 1.5938791124713083,
+      "grad_norm": 2.2101983287594034,
+      "learning_rate": 2.0860567008966783e-06,
+      "loss": 0.3755,
+      "step": 10416
+    },
+    {
+      "epoch": 1.5940321346595256,
+      "grad_norm": 2.5275320370762353,
+      "learning_rate": 2.084542015779595e-06,
+      "loss": 0.3384,
+      "step": 10417
+    },
+    {
+      "epoch": 1.5941851568477428,
+      "grad_norm": 2.058712242376689,
+      "learning_rate": 2.083027816781753e-06,
+      "loss": 0.3568,
+      "step": 10418
+    },
+    {
+      "epoch": 1.5943381790359603,
+      "grad_norm": 1.9173450107481087,
+      "learning_rate": 2.081514103996144e-06,
+      "loss": 0.2783,
+      "step": 10419
+    },
+    {
+      "epoch": 1.5944912012241774,
+      "grad_norm": 2.0613941843193793,
+      "learning_rate": 2.080000877515731e-06,
+      "loss": 0.2779,
+      "step": 10420
+    },
+    {
+      "epoch": 1.5946442234123948,
+      "grad_norm": 2.0254402953785022,
+      "learning_rate": 2.0784881374334464e-06,
+      "loss": 0.2917,
+      "step": 10421
+    },
+    {
+      "epoch": 1.594797245600612,
+      "grad_norm": 2.200732027895814,
+      "learning_rate": 2.076975883842196e-06,
+      "loss": 0.3322,
+      "step": 10422
+    },
+    {
+      "epoch": 1.5949502677888294,
+      "grad_norm": 2.0881731560154146,
+      "learning_rate": 2.075464116834851e-06,
+      "loss": 0.3173,
+      "step": 10423
+    },
+    {
+      "epoch": 1.5951032899770468,
+      "grad_norm": 2.1056631422662258,
+      "learning_rate": 2.0739528365042548e-06,
+      "loss": 0.2872,
+      "step": 10424
+    },
+    {
+      "epoch": 1.5952563121652639,
+      "grad_norm": 2.169035501067074,
+      "learning_rate": 2.0724420429432234e-06,
+      "loss": 0.3098,
+      "step": 10425
+    },
+    {
+      "epoch": 1.5954093343534812,
+      "grad_norm": 2.0153248522385647,
+      "learning_rate": 2.070931736244538e-06,
+      "loss": 0.2683,
+      "step": 10426
+    },
+    {
+      "epoch": 1.5955623565416985,
+      "grad_norm": 2.215459845696938,
+      "learning_rate": 2.0694219165009534e-06,
+      "loss": 0.3179,
+      "step": 10427
+    },
+    {
+      "epoch": 1.5957153787299159,
+      "grad_norm": 2.1656297130536872,
+      "learning_rate": 2.0679125838051926e-06,
+      "loss": 0.3034,
+      "step": 10428
+    },
+    {
+      "epoch": 1.5958684009181332,
+      "grad_norm": 1.9130239058279197,
+      "learning_rate": 2.0664037382499514e-06,
+      "loss": 0.2878,
+      "step": 10429
+    },
+    {
+      "epoch": 1.5960214231063503,
+      "grad_norm": 2.121543196047268,
+      "learning_rate": 2.0648953799278905e-06,
+      "loss": 0.3468,
+      "step": 10430
+    },
+    {
+      "epoch": 1.5961744452945679,
+      "grad_norm": 1.9881410386253346,
+      "learning_rate": 2.0633875089316524e-06,
+      "loss": 0.3357,
+      "step": 10431
+    },
+    {
+      "epoch": 1.596327467482785,
+      "grad_norm": 1.692514957785927,
+      "learning_rate": 2.0618801253538323e-06,
+      "loss": 0.2143,
+      "step": 10432
+    },
+    {
+      "epoch": 1.5964804896710023,
+      "grad_norm": 2.1161285582341782,
+      "learning_rate": 2.060373229287005e-06,
+      "loss": 0.2955,
+      "step": 10433
+    },
+    {
+      "epoch": 1.5966335118592196,
+      "grad_norm": 2.1256187952301437,
+      "learning_rate": 2.058866820823723e-06,
+      "loss": 0.3186,
+      "step": 10434
+    },
+    {
+      "epoch": 1.5967865340474368,
+      "grad_norm": 1.8594388357448477,
+      "learning_rate": 2.0573609000564933e-06,
+      "loss": 0.2979,
+      "step": 10435
+    },
+    {
+      "epoch": 1.5969395562356543,
+      "grad_norm": 2.0752161038423074,
+      "learning_rate": 2.0558554670777987e-06,
+      "loss": 0.2839,
+      "step": 10436
+    },
+    {
+      "epoch": 1.5970925784238714,
+      "grad_norm": 1.8719102046704,
+      "learning_rate": 2.054350521980104e-06,
+      "loss": 0.2496,
+      "step": 10437
+    },
+    {
+      "epoch": 1.5972456006120888,
+      "grad_norm": 2.2235478483167306,
+      "learning_rate": 2.052846064855821e-06,
+      "loss": 0.3171,
+      "step": 10438
+    },
+    {
+      "epoch": 1.597398622800306,
+      "grad_norm": 2.0084905451553143,
+      "learning_rate": 2.051342095797354e-06,
+      "loss": 0.2823,
+      "step": 10439
+    },
+    {
+      "epoch": 1.5975516449885232,
+      "grad_norm": 2.183017243080984,
+      "learning_rate": 2.049838614897067e-06,
+      "loss": 0.2913,
+      "step": 10440
+    },
+    {
+      "epoch": 1.5977046671767408,
+      "grad_norm": 2.010797629418803,
+      "learning_rate": 2.048335622247286e-06,
+      "loss": 0.3119,
+      "step": 10441
+    },
+    {
+      "epoch": 1.5978576893649579,
+      "grad_norm": 1.6995369473672757,
+      "learning_rate": 2.0468331179403245e-06,
+      "loss": 0.233,
+      "step": 10442
+    },
+    {
+      "epoch": 1.5980107115531752,
+      "grad_norm": 2.253316101580062,
+      "learning_rate": 2.045331102068454e-06,
+      "loss": 0.3347,
+      "step": 10443
+    },
+    {
+      "epoch": 1.5981637337413925,
+      "grad_norm": 2.064383868203437,
+      "learning_rate": 2.0438295747239203e-06,
+      "loss": 0.2746,
+      "step": 10444
+    },
+    {
+      "epoch": 1.5983167559296096,
+      "grad_norm": 2.2317677875246336,
+      "learning_rate": 2.0423285359989366e-06,
+      "loss": 0.2718,
+      "step": 10445
+    },
+    {
+      "epoch": 1.5984697781178272,
+      "grad_norm": 2.0477416135369517,
+      "learning_rate": 2.0408279859856874e-06,
+      "loss": 0.2805,
+      "step": 10446
+    },
+    {
+      "epoch": 1.5986228003060443,
+      "grad_norm": 1.957177005619269,
+      "learning_rate": 2.0393279247763287e-06,
+      "loss": 0.2795,
+      "step": 10447
+    },
+    {
+      "epoch": 1.5987758224942616,
+      "grad_norm": 2.3488538209818706,
+      "learning_rate": 2.0378283524629837e-06,
+      "loss": 0.3327,
+      "step": 10448
+    },
+    {
+      "epoch": 1.598928844682479,
+      "grad_norm": 1.9422180459270484,
+      "learning_rate": 2.036329269137749e-06,
+      "loss": 0.297,
+      "step": 10449
+    },
+    {
+      "epoch": 1.599081866870696,
+      "grad_norm": 2.2860116966764186,
+      "learning_rate": 2.0348306748926873e-06,
+      "loss": 0.3003,
+      "step": 10450
+    },
+    {
+      "epoch": 1.5992348890589136,
+      "grad_norm": 2.24122516023989,
+      "learning_rate": 2.033332569819834e-06,
+      "loss": 0.3337,
+      "step": 10451
+    },
+    {
+      "epoch": 1.5993879112471308,
+      "grad_norm": 2.1824835991354146,
+      "learning_rate": 2.0318349540111924e-06,
+      "loss": 0.2964,
+      "step": 10452
+    },
+    {
+      "epoch": 1.599540933435348,
+      "grad_norm": 2.3276355546648695,
+      "learning_rate": 2.030337827558738e-06,
+      "loss": 0.3569,
+      "step": 10453
+    },
+    {
+      "epoch": 1.5996939556235654,
+      "grad_norm": 2.0981843178955915,
+      "learning_rate": 2.0288411905544158e-06,
+      "loss": 0.3176,
+      "step": 10454
+    },
+    {
+      "epoch": 1.5998469778117828,
+      "grad_norm": 2.2062338938417296,
+      "learning_rate": 2.0273450430901396e-06,
+      "loss": 0.2734,
+      "step": 10455
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.643436905327408,
+      "learning_rate": 2.0258493852577933e-06,
+      "loss": 0.329,
+      "step": 10456
+    },
+    {
+      "epoch": 1.6001530221882172,
+      "grad_norm": 1.964606132543307,
+      "learning_rate": 2.0243542171492314e-06,
+      "loss": 0.2629,
+      "step": 10457
+    },
+    {
+      "epoch": 1.6003060443764345,
+      "grad_norm": 2.0319908205624424,
+      "learning_rate": 2.0228595388562776e-06,
+      "loss": 0.2485,
+      "step": 10458
+    },
+    {
+      "epoch": 1.6004590665646519,
+      "grad_norm": 2.157070383503065,
+      "learning_rate": 2.0213653504707243e-06,
+      "loss": 0.2879,
+      "step": 10459
+    },
+    {
+      "epoch": 1.6006120887528692,
+      "grad_norm": 2.0638106425645106,
+      "learning_rate": 2.019871652084342e-06,
+      "loss": 0.3038,
+      "step": 10460
+    },
+    {
+      "epoch": 1.6007651109410865,
+      "grad_norm": 2.413922284247371,
+      "learning_rate": 2.0183784437888577e-06,
+      "loss": 0.42,
+      "step": 10461
+    },
+    {
+      "epoch": 1.6009181331293036,
+      "grad_norm": 2.1985129144555353,
+      "learning_rate": 2.0168857256759745e-06,
+      "loss": 0.3118,
+      "step": 10462
+    },
+    {
+      "epoch": 1.6010711553175212,
+      "grad_norm": 2.2967862222343305,
+      "learning_rate": 2.0153934978373745e-06,
+      "loss": 0.3459,
+      "step": 10463
+    },
+    {
+      "epoch": 1.6012241775057383,
+      "grad_norm": 2.1976923086941325,
+      "learning_rate": 2.0139017603646925e-06,
+      "loss": 0.3005,
+      "step": 10464
+    },
+    {
+      "epoch": 1.6013771996939556,
+      "grad_norm": 1.9323315456919672,
+      "learning_rate": 2.012410513349543e-06,
+      "loss": 0.2619,
+      "step": 10465
+    },
+    {
+      "epoch": 1.601530221882173,
+      "grad_norm": 2.7793989397342838,
+      "learning_rate": 2.010919756883517e-06,
+      "loss": 0.3693,
+      "step": 10466
+    },
+    {
+      "epoch": 1.60168324407039,
+      "grad_norm": 2.1105341449495194,
+      "learning_rate": 2.009429491058157e-06,
+      "loss": 0.291,
+      "step": 10467
+    },
+    {
+      "epoch": 1.6018362662586076,
+      "grad_norm": 2.212110475794262,
+      "learning_rate": 2.0079397159649938e-06,
+      "loss": 0.3009,
+      "step": 10468
+    },
+    {
+      "epoch": 1.6019892884468248,
+      "grad_norm": 2.2904429553853576,
+      "learning_rate": 2.0064504316955204e-06,
+      "loss": 0.2809,
+      "step": 10469
+    },
+    {
+      "epoch": 1.602142310635042,
+      "grad_norm": 2.2952348089169443,
+      "learning_rate": 2.0049616383411906e-06,
+      "loss": 0.3095,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6022953328232594,
+      "grad_norm": 2.0420024184363377,
+      "learning_rate": 2.0034733359934476e-06,
+      "loss": 0.3069,
+      "step": 10471
+    },
+    {
+      "epoch": 1.6024483550114765,
+      "grad_norm": 2.087684753510197,
+      "learning_rate": 2.001985524743689e-06,
+      "loss": 0.2769,
+      "step": 10472
+    },
+    {
+      "epoch": 1.602601377199694,
+      "grad_norm": 2.3146003021955575,
+      "learning_rate": 2.0004982046832875e-06,
+      "loss": 0.3284,
+      "step": 10473
+    },
+    {
+      "epoch": 1.6027543993879112,
+      "grad_norm": 2.3341543365189037,
+      "learning_rate": 1.9990113759035856e-06,
+      "loss": 0.2941,
+      "step": 10474
+    },
+    {
+      "epoch": 1.6029074215761285,
+      "grad_norm": 2.42343083766534,
+      "learning_rate": 1.9975250384958954e-06,
+      "loss": 0.3841,
+      "step": 10475
+    },
+    {
+      "epoch": 1.6030604437643459,
+      "grad_norm": 2.0000642760572793,
+      "learning_rate": 1.9960391925514975e-06,
+      "loss": 0.2387,
+      "step": 10476
+    },
+    {
+      "epoch": 1.603213465952563,
+      "grad_norm": 2.106691781994404,
+      "learning_rate": 1.9945538381616456e-06,
+      "loss": 0.2961,
+      "step": 10477
+    },
+    {
+      "epoch": 1.6033664881407805,
+      "grad_norm": 2.0613737551747042,
+      "learning_rate": 1.9930689754175604e-06,
+      "loss": 0.3309,
+      "step": 10478
+    },
+    {
+      "epoch": 1.6035195103289976,
+      "grad_norm": 2.4398182852589105,
+      "learning_rate": 1.9915846044104313e-06,
+      "loss": 0.3037,
+      "step": 10479
+    },
+    {
+      "epoch": 1.603672532517215,
+      "grad_norm": 2.1090739150374,
+      "learning_rate": 1.9901007252314185e-06,
+      "loss": 0.2903,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6038255547054323,
+      "grad_norm": 2.144499019468979,
+      "learning_rate": 1.988617337971661e-06,
+      "loss": 0.31,
+      "step": 10481
+    },
+    {
+      "epoch": 1.6039785768936494,
+      "grad_norm": 2.026883977576024,
+      "learning_rate": 1.9871344427222504e-06,
+      "loss": 0.3783,
+      "step": 10482
+    },
+    {
+      "epoch": 1.604131599081867,
+      "grad_norm": 2.411293659913249,
+      "learning_rate": 1.9856520395742575e-06,
+      "loss": 0.3794,
+      "step": 10483
+    },
+    {
+      "epoch": 1.604284621270084,
+      "grad_norm": 2.0058040325716755,
+      "learning_rate": 1.9841701286187297e-06,
+      "loss": 0.2296,
+      "step": 10484
+    },
+    {
+      "epoch": 1.6044376434583014,
+      "grad_norm": 2.0780361687783278,
+      "learning_rate": 1.982688709946671e-06,
+      "loss": 0.2714,
+      "step": 10485
+    },
+    {
+      "epoch": 1.6045906656465188,
+      "grad_norm": 2.0576034473812266,
+      "learning_rate": 1.9812077836490595e-06,
+      "loss": 0.3636,
+      "step": 10486
+    },
+    {
+      "epoch": 1.604743687834736,
+      "grad_norm": 1.9414393221409934,
+      "learning_rate": 1.979727349816852e-06,
+      "loss": 0.3066,
+      "step": 10487
+    },
+    {
+      "epoch": 1.6048967100229534,
+      "grad_norm": 2.4956241751854003,
+      "learning_rate": 1.9782474085409597e-06,
+      "loss": 0.3205,
+      "step": 10488
+    },
+    {
+      "epoch": 1.6050497322111705,
+      "grad_norm": 2.1620399546925717,
+      "learning_rate": 1.9767679599122767e-06,
+      "loss": 0.3238,
+      "step": 10489
+    },
+    {
+      "epoch": 1.6052027543993879,
+      "grad_norm": 2.3001838451511665,
+      "learning_rate": 1.9752890040216644e-06,
+      "loss": 0.3707,
+      "step": 10490
+    },
+    {
+      "epoch": 1.6053557765876052,
+      "grad_norm": 2.1205275312048237,
+      "learning_rate": 1.9738105409599405e-06,
+      "loss": 0.3203,
+      "step": 10491
+    },
+    {
+      "epoch": 1.6055087987758225,
+      "grad_norm": 2.3261888986198023,
+      "learning_rate": 1.972332570817913e-06,
+      "loss": 0.3607,
+      "step": 10492
+    },
+    {
+      "epoch": 1.6056618209640399,
+      "grad_norm": 2.2311365973332733,
+      "learning_rate": 1.97085509368635e-06,
+      "loss": 0.2721,
+      "step": 10493
+    },
+    {
+      "epoch": 1.605814843152257,
+      "grad_norm": 2.0157940050326655,
+      "learning_rate": 1.9693781096559794e-06,
+      "loss": 0.2747,
+      "step": 10494
+    },
+    {
+      "epoch": 1.6059678653404745,
+      "grad_norm": 1.9952768637916527,
+      "learning_rate": 1.9679016188175193e-06,
+      "loss": 0.2393,
+      "step": 10495
+    },
+    {
+      "epoch": 1.6061208875286916,
+      "grad_norm": 2.3814074053620207,
+      "learning_rate": 1.9664256212616417e-06,
+      "loss": 0.3275,
+      "step": 10496
+    },
+    {
+      "epoch": 1.606273909716909,
+      "grad_norm": 2.105466025822687,
+      "learning_rate": 1.9649501170789943e-06,
+      "loss": 0.2923,
+      "step": 10497
+    },
+    {
+      "epoch": 1.6064269319051263,
+      "grad_norm": 1.8103329861441821,
+      "learning_rate": 1.963475106360193e-06,
+      "loss": 0.2546,
+      "step": 10498
+    },
+    {
+      "epoch": 1.6065799540933434,
+      "grad_norm": 1.981345827940942,
+      "learning_rate": 1.962000589195825e-06,
+      "loss": 0.2861,
+      "step": 10499
+    },
+    {
+      "epoch": 1.606732976281561,
+      "grad_norm": 2.031188520976353,
+      "learning_rate": 1.960526565676445e-06,
+      "loss": 0.3322,
+      "step": 10500
+    },
+    {
+      "epoch": 1.606885998469778,
+      "grad_norm": 2.073844327814389,
+      "learning_rate": 1.9590530358925796e-06,
+      "loss": 0.3111,
+      "step": 10501
+    },
+    {
+      "epoch": 1.6070390206579954,
+      "grad_norm": 2.224144695075106,
+      "learning_rate": 1.9575799999347247e-06,
+      "loss": 0.3317,
+      "step": 10502
+    },
+    {
+      "epoch": 1.6071920428462128,
+      "grad_norm": 2.2658544063551456,
+      "learning_rate": 1.9561074578933424e-06,
+      "loss": 0.4216,
+      "step": 10503
+    },
+    {
+      "epoch": 1.6073450650344299,
+      "grad_norm": 2.3868575362093907,
+      "learning_rate": 1.9546354098588693e-06,
+      "loss": 0.3201,
+      "step": 10504
+    },
+    {
+      "epoch": 1.6074980872226474,
+      "grad_norm": 2.052778963043461,
+      "learning_rate": 1.9531638559217104e-06,
+      "loss": 0.3202,
+      "step": 10505
+    },
+    {
+      "epoch": 1.6076511094108645,
+      "grad_norm": 2.027431317380972,
+      "learning_rate": 1.951692796172238e-06,
+      "loss": 0.3444,
+      "step": 10506
+    },
+    {
+      "epoch": 1.6078041315990819,
+      "grad_norm": 1.9896009671346127,
+      "learning_rate": 1.9502222307007956e-06,
+      "loss": 0.2569,
+      "step": 10507
+    },
+    {
+      "epoch": 1.6079571537872992,
+      "grad_norm": 2.3888546056899918,
+      "learning_rate": 1.948752159597698e-06,
+      "loss": 0.3455,
+      "step": 10508
+    },
+    {
+      "epoch": 1.6081101759755163,
+      "grad_norm": 2.3012402578354036,
+      "learning_rate": 1.9472825829532226e-06,
+      "loss": 0.3532,
+      "step": 10509
+    },
+    {
+      "epoch": 1.6082631981637339,
+      "grad_norm": 2.214180934529892,
+      "learning_rate": 1.945813500857633e-06,
+      "loss": 0.3095,
+      "step": 10510
+    },
+    {
+      "epoch": 1.608416220351951,
+      "grad_norm": 2.406222516683202,
+      "learning_rate": 1.9443449134011416e-06,
+      "loss": 0.3571,
+      "step": 10511
+    },
+    {
+      "epoch": 1.6085692425401683,
+      "grad_norm": 1.972700407695371,
+      "learning_rate": 1.9428768206739403e-06,
+      "loss": 0.3136,
+      "step": 10512
+    },
+    {
+      "epoch": 1.6087222647283856,
+      "grad_norm": 2.103254444351623,
+      "learning_rate": 1.9414092227661976e-06,
+      "loss": 0.2693,
+      "step": 10513
+    },
+    {
+      "epoch": 1.6088752869166028,
+      "grad_norm": 2.0014590653047346,
+      "learning_rate": 1.939942119768038e-06,
+      "loss": 0.2823,
+      "step": 10514
+    },
+    {
+      "epoch": 1.6090283091048203,
+      "grad_norm": 2.558600077343717,
+      "learning_rate": 1.9384755117695607e-06,
+      "loss": 0.4164,
+      "step": 10515
+    },
+    {
+      "epoch": 1.6091813312930374,
+      "grad_norm": 1.9254914747715335,
+      "learning_rate": 1.9370093988608453e-06,
+      "loss": 0.2698,
+      "step": 10516
+    },
+    {
+      "epoch": 1.6093343534812548,
+      "grad_norm": 2.3909848914674203,
+      "learning_rate": 1.935543781131919e-06,
+      "loss": 0.3229,
+      "step": 10517
+    },
+    {
+      "epoch": 1.609487375669472,
+      "grad_norm": 2.024294601203651,
+      "learning_rate": 1.934078658672801e-06,
+      "loss": 0.2838,
+      "step": 10518
+    },
+    {
+      "epoch": 1.6096403978576894,
+      "grad_norm": 2.5391493263896154,
+      "learning_rate": 1.9326140315734685e-06,
+      "loss": 0.418,
+      "step": 10519
+    },
+    {
+      "epoch": 1.6097934200459068,
+      "grad_norm": 2.267744207696907,
+      "learning_rate": 1.931149899923863e-06,
+      "loss": 0.2905,
+      "step": 10520
+    },
+    {
+      "epoch": 1.6099464422341239,
+      "grad_norm": 1.9815819153211391,
+      "learning_rate": 1.9296862638139103e-06,
+      "loss": 0.2529,
+      "step": 10521
+    },
+    {
+      "epoch": 1.6100994644223412,
+      "grad_norm": 1.8905960118777305,
+      "learning_rate": 1.928223123333498e-06,
+      "loss": 0.2605,
+      "step": 10522
+    },
+    {
+      "epoch": 1.6102524866105585,
+      "grad_norm": 2.1424726322335386,
+      "learning_rate": 1.926760478572476e-06,
+      "loss": 0.322,
+      "step": 10523
+    },
+    {
+      "epoch": 1.6104055087987759,
+      "grad_norm": 2.2819861093889036,
+      "learning_rate": 1.9252983296206784e-06,
+      "loss": 0.2884,
+      "step": 10524
+    },
+    {
+      "epoch": 1.6105585309869932,
+      "grad_norm": 2.0572968938301877,
+      "learning_rate": 1.9238366765678972e-06,
+      "loss": 0.3299,
+      "step": 10525
+    },
+    {
+      "epoch": 1.6107115531752103,
+      "grad_norm": 1.969348700776263,
+      "learning_rate": 1.922375519503902e-06,
+      "loss": 0.2684,
+      "step": 10526
+    },
+    {
+      "epoch": 1.6108645753634276,
+      "grad_norm": 2.202527841852906,
+      "learning_rate": 1.9209148585184244e-06,
+      "loss": 0.3269,
+      "step": 10527
+    },
+    {
+      "epoch": 1.611017597551645,
+      "grad_norm": 1.940551357103671,
+      "learning_rate": 1.9194546937011716e-06,
+      "loss": 0.2926,
+      "step": 10528
+    },
+    {
+      "epoch": 1.6111706197398623,
+      "grad_norm": 2.4080905545015825,
+      "learning_rate": 1.9179950251418168e-06,
+      "loss": 0.3119,
+      "step": 10529
+    },
+    {
+      "epoch": 1.6113236419280796,
+      "grad_norm": 2.1771118306491495,
+      "learning_rate": 1.9165358529300047e-06,
+      "loss": 0.2689,
+      "step": 10530
+    },
+    {
+      "epoch": 1.6114766641162968,
+      "grad_norm": 2.2555773274378237,
+      "learning_rate": 1.915077177155349e-06,
+      "loss": 0.319,
+      "step": 10531
+    },
+    {
+      "epoch": 1.6116296863045143,
+      "grad_norm": 2.0593411592193664,
+      "learning_rate": 1.9136189979074314e-06,
+      "loss": 0.3501,
+      "step": 10532
+    },
+    {
+      "epoch": 1.6117827084927314,
+      "grad_norm": 1.712499352328928,
+      "learning_rate": 1.9121613152758067e-06,
+      "loss": 0.2723,
+      "step": 10533
+    },
+    {
+      "epoch": 1.6119357306809488,
+      "grad_norm": 2.229661991037623,
+      "learning_rate": 1.910704129349994e-06,
+      "loss": 0.3884,
+      "step": 10534
+    },
+    {
+      "epoch": 1.612088752869166,
+      "grad_norm": 2.44309670412781,
+      "learning_rate": 1.909247440219487e-06,
+      "loss": 0.3624,
+      "step": 10535
+    },
+    {
+      "epoch": 1.6122417750573832,
+      "grad_norm": 2.0981658994548904,
+      "learning_rate": 1.907791247973746e-06,
+      "loss": 0.3294,
+      "step": 10536
+    },
+    {
+      "epoch": 1.6123947972456008,
+      "grad_norm": 2.2664865350975956,
+      "learning_rate": 1.906335552702201e-06,
+      "loss": 0.3435,
+      "step": 10537
+    },
+    {
+      "epoch": 1.6125478194338179,
+      "grad_norm": 2.166203828647484,
+      "learning_rate": 1.9048803544942518e-06,
+      "loss": 0.3217,
+      "step": 10538
+    },
+    {
+      "epoch": 1.6127008416220352,
+      "grad_norm": 2.1129488376182204,
+      "learning_rate": 1.9034256534392692e-06,
+      "loss": 0.2719,
+      "step": 10539
+    },
+    {
+      "epoch": 1.6128538638102525,
+      "grad_norm": 1.9130793331631273,
+      "learning_rate": 1.9019714496265906e-06,
+      "loss": 0.2827,
+      "step": 10540
+    },
+    {
+      "epoch": 1.6130068859984696,
+      "grad_norm": 1.9783507773097913,
+      "learning_rate": 1.9005177431455223e-06,
+      "loss": 0.3302,
+      "step": 10541
+    },
+    {
+      "epoch": 1.6131599081866872,
+      "grad_norm": 2.266976194472374,
+      "learning_rate": 1.8990645340853498e-06,
+      "loss": 0.3105,
+      "step": 10542
+    },
+    {
+      "epoch": 1.6133129303749043,
+      "grad_norm": 2.447531198231755,
+      "learning_rate": 1.8976118225353135e-06,
+      "loss": 0.337,
+      "step": 10543
+    },
+    {
+      "epoch": 1.6134659525631216,
+      "grad_norm": 1.9466005333498533,
+      "learning_rate": 1.896159608584629e-06,
+      "loss": 0.2737,
+      "step": 10544
+    },
+    {
+      "epoch": 1.613618974751339,
+      "grad_norm": 2.270820493266619,
+      "learning_rate": 1.8947078923224905e-06,
+      "loss": 0.3383,
+      "step": 10545
+    },
+    {
+      "epoch": 1.613771996939556,
+      "grad_norm": 2.175481962547,
+      "learning_rate": 1.8932566738380432e-06,
+      "loss": 0.2804,
+      "step": 10546
+    },
+    {
+      "epoch": 1.6139250191277736,
+      "grad_norm": 2.3916081886581337,
+      "learning_rate": 1.891805953220419e-06,
+      "loss": 0.3189,
+      "step": 10547
+    },
+    {
+      "epoch": 1.6140780413159908,
+      "grad_norm": 2.199411097369911,
+      "learning_rate": 1.890355730558715e-06,
+      "loss": 0.3678,
+      "step": 10548
+    },
+    {
+      "epoch": 1.614231063504208,
+      "grad_norm": 2.186939011479005,
+      "learning_rate": 1.888906005941984e-06,
+      "loss": 0.3018,
+      "step": 10549
+    },
+    {
+      "epoch": 1.6143840856924254,
+      "grad_norm": 2.1798549473165565,
+      "learning_rate": 1.887456779459269e-06,
+      "loss": 0.2894,
+      "step": 10550
+    },
+    {
+      "epoch": 1.6145371078806425,
+      "grad_norm": 2.273224534202083,
+      "learning_rate": 1.8860080511995727e-06,
+      "loss": 0.3557,
+      "step": 10551
+    },
+    {
+      "epoch": 1.61469013006886,
+      "grad_norm": 2.1166615630191603,
+      "learning_rate": 1.884559821251859e-06,
+      "loss": 0.3081,
+      "step": 10552
+    },
+    {
+      "epoch": 1.6148431522570772,
+      "grad_norm": 1.9044929221608777,
+      "learning_rate": 1.8831120897050759e-06,
+      "loss": 0.2839,
+      "step": 10553
+    },
+    {
+      "epoch": 1.6149961744452945,
+      "grad_norm": 2.1024143177946377,
+      "learning_rate": 1.8816648566481343e-06,
+      "loss": 0.2778,
+      "step": 10554
+    },
+    {
+      "epoch": 1.6151491966335119,
+      "grad_norm": 2.1048915778898163,
+      "learning_rate": 1.8802181221699124e-06,
+      "loss": 0.3376,
+      "step": 10555
+    },
+    {
+      "epoch": 1.6153022188217292,
+      "grad_norm": 2.187077064266125,
+      "learning_rate": 1.8787718863592597e-06,
+      "loss": 0.3192,
+      "step": 10556
+    },
+    {
+      "epoch": 1.6154552410099465,
+      "grad_norm": 1.9487619847012356,
+      "learning_rate": 1.8773261493049965e-06,
+      "loss": 0.2797,
+      "step": 10557
+    },
+    {
+      "epoch": 1.6156082631981636,
+      "grad_norm": 2.1489496371148378,
+      "learning_rate": 1.8758809110959098e-06,
+      "loss": 0.3082,
+      "step": 10558
+    },
+    {
+      "epoch": 1.615761285386381,
+      "grad_norm": 1.877140278626581,
+      "learning_rate": 1.8744361718207593e-06,
+      "loss": 0.3091,
+      "step": 10559
+    },
+    {
+      "epoch": 1.6159143075745983,
+      "grad_norm": 2.4222748132697367,
+      "learning_rate": 1.87299193156827e-06,
+      "loss": 0.321,
+      "step": 10560
+    },
+    {
+      "epoch": 1.6160673297628156,
+      "grad_norm": 2.402037261974164,
+      "learning_rate": 1.8715481904271393e-06,
+      "loss": 0.3186,
+      "step": 10561
+    },
+    {
+      "epoch": 1.616220351951033,
+      "grad_norm": 1.842256510948127,
+      "learning_rate": 1.8701049484860312e-06,
+      "loss": 0.274,
+      "step": 10562
+    },
+    {
+      "epoch": 1.61637337413925,
+      "grad_norm": 2.1205220163507255,
+      "learning_rate": 1.8686622058335867e-06,
+      "loss": 0.2792,
+      "step": 10563
+    },
+    {
+      "epoch": 1.6165263963274676,
+      "grad_norm": 2.161144153109064,
+      "learning_rate": 1.8672199625584042e-06,
+      "loss": 0.3228,
+      "step": 10564
+    },
+    {
+      "epoch": 1.6166794185156848,
+      "grad_norm": 2.288586202317026,
+      "learning_rate": 1.8657782187490558e-06,
+      "loss": 0.3729,
+      "step": 10565
+    },
+    {
+      "epoch": 1.616832440703902,
+      "grad_norm": 2.110696657148073,
+      "learning_rate": 1.8643369744940942e-06,
+      "loss": 0.312,
+      "step": 10566
+    },
+    {
+      "epoch": 1.6169854628921194,
+      "grad_norm": 2.0382296277093466,
+      "learning_rate": 1.862896229882023e-06,
+      "loss": 0.2529,
+      "step": 10567
+    },
+    {
+      "epoch": 1.6171384850803365,
+      "grad_norm": 2.040328917756369,
+      "learning_rate": 1.8614559850013247e-06,
+      "loss": 0.2704,
+      "step": 10568
+    },
+    {
+      "epoch": 1.617291507268554,
+      "grad_norm": 2.0325733210458905,
+      "learning_rate": 1.8600162399404576e-06,
+      "loss": 0.2916,
+      "step": 10569
+    },
+    {
+      "epoch": 1.6174445294567712,
+      "grad_norm": 1.8749942403353452,
+      "learning_rate": 1.8585769947878318e-06,
+      "loss": 0.2755,
+      "step": 10570
+    },
+    {
+      "epoch": 1.6175975516449885,
+      "grad_norm": 1.9462632302198313,
+      "learning_rate": 1.8571382496318446e-06,
+      "loss": 0.2756,
+      "step": 10571
+    },
+    {
+      "epoch": 1.6177505738332059,
+      "grad_norm": 2.0840998504651407,
+      "learning_rate": 1.8557000045608553e-06,
+      "loss": 0.2953,
+      "step": 10572
+    },
+    {
+      "epoch": 1.617903596021423,
+      "grad_norm": 2.3883921872480305,
+      "learning_rate": 1.8542622596631831e-06,
+      "loss": 0.3609,
+      "step": 10573
+    },
+    {
+      "epoch": 1.6180566182096405,
+      "grad_norm": 2.256482256007678,
+      "learning_rate": 1.8528250150271365e-06,
+      "loss": 0.342,
+      "step": 10574
+    },
+    {
+      "epoch": 1.6182096403978576,
+      "grad_norm": 2.3045778230254768,
+      "learning_rate": 1.8513882707409791e-06,
+      "loss": 0.3323,
+      "step": 10575
+    },
+    {
+      "epoch": 1.618362662586075,
+      "grad_norm": 2.3223204819305505,
+      "learning_rate": 1.8499520268929406e-06,
+      "loss": 0.3327,
+      "step": 10576
+    },
+    {
+      "epoch": 1.6185156847742923,
+      "grad_norm": 2.1041983316848696,
+      "learning_rate": 1.8485162835712334e-06,
+      "loss": 0.327,
+      "step": 10577
+    },
+    {
+      "epoch": 1.6186687069625094,
+      "grad_norm": 1.9349892975599834,
+      "learning_rate": 1.8470810408640305e-06,
+      "loss": 0.2856,
+      "step": 10578
+    },
+    {
+      "epoch": 1.618821729150727,
+      "grad_norm": 2.1178930440291057,
+      "learning_rate": 1.8456462988594748e-06,
+      "loss": 0.3653,
+      "step": 10579
+    },
+    {
+      "epoch": 1.618974751338944,
+      "grad_norm": 2.1031709608253695,
+      "learning_rate": 1.8442120576456802e-06,
+      "loss": 0.2744,
+      "step": 10580
+    },
+    {
+      "epoch": 1.6191277735271614,
+      "grad_norm": 2.478206559028707,
+      "learning_rate": 1.8427783173107283e-06,
+      "loss": 0.3079,
+      "step": 10581
+    },
+    {
+      "epoch": 1.6192807957153788,
+      "grad_norm": 1.8838818778239124,
+      "learning_rate": 1.8413450779426723e-06,
+      "loss": 0.2817,
+      "step": 10582
+    },
+    {
+      "epoch": 1.6194338179035959,
+      "grad_norm": 1.912703665415085,
+      "learning_rate": 1.8399123396295305e-06,
+      "loss": 0.3116,
+      "step": 10583
+    },
+    {
+      "epoch": 1.6195868400918134,
+      "grad_norm": 2.2889397537082345,
+      "learning_rate": 1.8384801024592957e-06,
+      "loss": 0.3419,
+      "step": 10584
+    },
+    {
+      "epoch": 1.6197398622800305,
+      "grad_norm": 2.012218099198962,
+      "learning_rate": 1.8370483665199246e-06,
+      "loss": 0.247,
+      "step": 10585
+    },
+    {
+      "epoch": 1.6198928844682479,
+      "grad_norm": 2.6818823744928593,
+      "learning_rate": 1.8356171318993477e-06,
+      "loss": 0.3282,
+      "step": 10586
+    },
+    {
+      "epoch": 1.6200459066564652,
+      "grad_norm": 2.2120147507952304,
+      "learning_rate": 1.8341863986854624e-06,
+      "loss": 0.3462,
+      "step": 10587
+    },
+    {
+      "epoch": 1.6201989288446825,
+      "grad_norm": 2.052125974963057,
+      "learning_rate": 1.8327561669661343e-06,
+      "loss": 0.3083,
+      "step": 10588
+    },
+    {
+      "epoch": 1.6203519510328999,
+      "grad_norm": 1.9840591498263909,
+      "learning_rate": 1.8313264368292005e-06,
+      "loss": 0.2652,
+      "step": 10589
+    },
+    {
+      "epoch": 1.620504973221117,
+      "grad_norm": 2.4214798106894975,
+      "learning_rate": 1.8298972083624667e-06,
+      "loss": 0.348,
+      "step": 10590
+    },
+    {
+      "epoch": 1.6206579954093343,
+      "grad_norm": 1.8997166893876836,
+      "learning_rate": 1.8284684816537045e-06,
+      "loss": 0.3332,
+      "step": 10591
+    },
+    {
+      "epoch": 1.6208110175975516,
+      "grad_norm": 2.369531878415244,
+      "learning_rate": 1.8270402567906654e-06,
+      "loss": 0.3397,
+      "step": 10592
+    },
+    {
+      "epoch": 1.620964039785769,
+      "grad_norm": 2.170929041681945,
+      "learning_rate": 1.825612533861053e-06,
+      "loss": 0.2398,
+      "step": 10593
+    },
+    {
+      "epoch": 1.6211170619739863,
+      "grad_norm": 1.9795611289660413,
+      "learning_rate": 1.8241853129525522e-06,
+      "loss": 0.2896,
+      "step": 10594
+    },
+    {
+      "epoch": 1.6212700841622034,
+      "grad_norm": 2.155517297499104,
+      "learning_rate": 1.8227585941528192e-06,
+      "loss": 0.2697,
+      "step": 10595
+    },
+    {
+      "epoch": 1.621423106350421,
+      "grad_norm": 2.228766404040802,
+      "learning_rate": 1.8213323775494684e-06,
+      "loss": 0.2509,
+      "step": 10596
+    },
+    {
+      "epoch": 1.621576128538638,
+      "grad_norm": 2.187183161177902,
+      "learning_rate": 1.81990666323009e-06,
+      "loss": 0.3133,
+      "step": 10597
+    },
+    {
+      "epoch": 1.6217291507268554,
+      "grad_norm": 1.8590016521156796,
+      "learning_rate": 1.8184814512822479e-06,
+      "loss": 0.2984,
+      "step": 10598
+    },
+    {
+      "epoch": 1.6218821729150728,
+      "grad_norm": 2.134756508469202,
+      "learning_rate": 1.8170567417934615e-06,
+      "loss": 0.3212,
+      "step": 10599
+    },
+    {
+      "epoch": 1.6220351951032899,
+      "grad_norm": 1.975640103428586,
+      "learning_rate": 1.815632534851235e-06,
+      "loss": 0.344,
+      "step": 10600
+    },
+    {
+      "epoch": 1.6221882172915074,
+      "grad_norm": 1.9895208241922568,
+      "learning_rate": 1.8142088305430339e-06,
+      "loss": 0.3345,
+      "step": 10601
+    },
+    {
+      "epoch": 1.6223412394797245,
+      "grad_norm": 2.0141924746839344,
+      "learning_rate": 1.8127856289562873e-06,
+      "loss": 0.2954,
+      "step": 10602
+    },
+    {
+      "epoch": 1.6224942616679419,
+      "grad_norm": 2.392543783330661,
+      "learning_rate": 1.8113629301784052e-06,
+      "loss": 0.3212,
+      "step": 10603
+    },
+    {
+      "epoch": 1.6226472838561592,
+      "grad_norm": 2.2038116103794976,
+      "learning_rate": 1.8099407342967635e-06,
+      "loss": 0.2888,
+      "step": 10604
+    },
+    {
+      "epoch": 1.6228003060443763,
+      "grad_norm": 2.0600551971159806,
+      "learning_rate": 1.8085190413986953e-06,
+      "loss": 0.3045,
+      "step": 10605
+    },
+    {
+      "epoch": 1.6229533282325939,
+      "grad_norm": 1.9894692711668345,
+      "learning_rate": 1.807097851571521e-06,
+      "loss": 0.3068,
+      "step": 10606
+    },
+    {
+      "epoch": 1.623106350420811,
+      "grad_norm": 2.2340507250131147,
+      "learning_rate": 1.8056771649025173e-06,
+      "loss": 0.3865,
+      "step": 10607
+    },
+    {
+      "epoch": 1.6232593726090283,
+      "grad_norm": 1.8290540957109105,
+      "learning_rate": 1.8042569814789367e-06,
+      "loss": 0.245,
+      "step": 10608
+    },
+    {
+      "epoch": 1.6234123947972456,
+      "grad_norm": 2.3069059295735155,
+      "learning_rate": 1.8028373013879964e-06,
+      "loss": 0.3339,
+      "step": 10609
+    },
+    {
+      "epoch": 1.6235654169854628,
+      "grad_norm": 2.1424771142080723,
+      "learning_rate": 1.801418124716884e-06,
+      "loss": 0.2891,
+      "step": 10610
+    },
+    {
+      "epoch": 1.6237184391736803,
+      "grad_norm": 2.4355336467958035,
+      "learning_rate": 1.7999994515527586e-06,
+      "loss": 0.3452,
+      "step": 10611
+    },
+    {
+      "epoch": 1.6238714613618974,
+      "grad_norm": 2.112129584337102,
+      "learning_rate": 1.7985812819827452e-06,
+      "loss": 0.2674,
+      "step": 10612
+    },
+    {
+      "epoch": 1.6240244835501148,
+      "grad_norm": 1.8521996070155238,
+      "learning_rate": 1.7971636160939388e-06,
+      "loss": 0.263,
+      "step": 10613
+    },
+    {
+      "epoch": 1.624177505738332,
+      "grad_norm": 2.216065359992707,
+      "learning_rate": 1.7957464539734048e-06,
+      "loss": 0.3601,
+      "step": 10614
+    },
+    {
+      "epoch": 1.6243305279265492,
+      "grad_norm": 1.7979366955047242,
+      "learning_rate": 1.7943297957081762e-06,
+      "loss": 0.2903,
+      "step": 10615
+    },
+    {
+      "epoch": 1.6244835501147668,
+      "grad_norm": 2.39989702737361,
+      "learning_rate": 1.7929136413852567e-06,
+      "loss": 0.428,
+      "step": 10616
+    },
+    {
+      "epoch": 1.6246365723029839,
+      "grad_norm": 2.2140572585003557,
+      "learning_rate": 1.7914979910916152e-06,
+      "loss": 0.31,
+      "step": 10617
+    },
+    {
+      "epoch": 1.6247895944912012,
+      "grad_norm": 2.4071733588786715,
+      "learning_rate": 1.7900828449141949e-06,
+      "loss": 0.3316,
+      "step": 10618
+    },
+    {
+      "epoch": 1.6249426166794185,
+      "grad_norm": 2.1107088272132812,
+      "learning_rate": 1.788668202939904e-06,
+      "loss": 0.3253,
+      "step": 10619
+    },
+    {
+      "epoch": 1.6250956388676359,
+      "grad_norm": 2.166748268503527,
+      "learning_rate": 1.7872540652556192e-06,
+      "loss": 0.3268,
+      "step": 10620
+    },
+    {
+      "epoch": 1.6252486610558532,
+      "grad_norm": 2.0274159462473382,
+      "learning_rate": 1.785840431948196e-06,
+      "loss": 0.2423,
+      "step": 10621
+    },
+    {
+      "epoch": 1.6254016832440703,
+      "grad_norm": 2.2267309500004413,
+      "learning_rate": 1.7844273031044435e-06,
+      "loss": 0.2898,
+      "step": 10622
+    },
+    {
+      "epoch": 1.6255547054322876,
+      "grad_norm": 2.1327367988472776,
+      "learning_rate": 1.783014678811147e-06,
+      "loss": 0.2974,
+      "step": 10623
+    },
+    {
+      "epoch": 1.625707727620505,
+      "grad_norm": 2.054036296303469,
+      "learning_rate": 1.781602559155069e-06,
+      "loss": 0.303,
+      "step": 10624
+    },
+    {
+      "epoch": 1.6258607498087223,
+      "grad_norm": 1.9841387608214618,
+      "learning_rate": 1.7801909442229258e-06,
+      "loss": 0.2797,
+      "step": 10625
+    },
+    {
+      "epoch": 1.6260137719969396,
+      "grad_norm": 1.964548815517046,
+      "learning_rate": 1.7787798341014107e-06,
+      "loss": 0.2808,
+      "step": 10626
+    },
+    {
+      "epoch": 1.6261667941851567,
+      "grad_norm": 2.1037655301339906,
+      "learning_rate": 1.7773692288771927e-06,
+      "loss": 0.2973,
+      "step": 10627
+    },
+    {
+      "epoch": 1.6263198163733743,
+      "grad_norm": 2.024239937854239,
+      "learning_rate": 1.7759591286368915e-06,
+      "loss": 0.2694,
+      "step": 10628
+    },
+    {
+      "epoch": 1.6264728385615914,
+      "grad_norm": 2.094301076866119,
+      "learning_rate": 1.774549533467116e-06,
+      "loss": 0.2871,
+      "step": 10629
+    },
+    {
+      "epoch": 1.6266258607498087,
+      "grad_norm": 1.9001204144586141,
+      "learning_rate": 1.773140443454434e-06,
+      "loss": 0.2899,
+      "step": 10630
+    },
+    {
+      "epoch": 1.626778882938026,
+      "grad_norm": 2.0305966737469117,
+      "learning_rate": 1.7717318586853772e-06,
+      "loss": 0.3259,
+      "step": 10631
+    },
+    {
+      "epoch": 1.6269319051262432,
+      "grad_norm": 2.132174412861507,
+      "learning_rate": 1.7703237792464567e-06,
+      "loss": 0.3068,
+      "step": 10632
+    },
+    {
+      "epoch": 1.6270849273144607,
+      "grad_norm": 2.267922670169627,
+      "learning_rate": 1.7689162052241515e-06,
+      "loss": 0.3581,
+      "step": 10633
+    },
+    {
+      "epoch": 1.6272379495026779,
+      "grad_norm": 2.0521196074770414,
+      "learning_rate": 1.767509136704897e-06,
+      "loss": 0.2833,
+      "step": 10634
+    },
+    {
+      "epoch": 1.6273909716908952,
+      "grad_norm": 1.95384614994354,
+      "learning_rate": 1.7661025737751148e-06,
+      "loss": 0.2801,
+      "step": 10635
+    },
+    {
+      "epoch": 1.6275439938791125,
+      "grad_norm": 2.073529431779787,
+      "learning_rate": 1.7646965165211837e-06,
+      "loss": 0.2925,
+      "step": 10636
+    },
+    {
+      "epoch": 1.6276970160673296,
+      "grad_norm": 1.681491994873035,
+      "learning_rate": 1.7632909650294571e-06,
+      "loss": 0.2322,
+      "step": 10637
+    },
+    {
+      "epoch": 1.6278500382555472,
+      "grad_norm": 2.013831593099779,
+      "learning_rate": 1.7618859193862547e-06,
+      "loss": 0.246,
+      "step": 10638
+    },
+    {
+      "epoch": 1.6280030604437643,
+      "grad_norm": 2.222352955584193,
+      "learning_rate": 1.7604813796778652e-06,
+      "loss": 0.3341,
+      "step": 10639
+    },
+    {
+      "epoch": 1.6281560826319816,
+      "grad_norm": 2.0854233819651253,
+      "learning_rate": 1.7590773459905475e-06,
+      "loss": 0.2933,
+      "step": 10640
+    },
+    {
+      "epoch": 1.628309104820199,
+      "grad_norm": 2.1521488694164495,
+      "learning_rate": 1.7576738184105292e-06,
+      "loss": 0.3917,
+      "step": 10641
+    },
+    {
+      "epoch": 1.628462127008416,
+      "grad_norm": 2.0691690612534135,
+      "learning_rate": 1.7562707970240046e-06,
+      "loss": 0.2824,
+      "step": 10642
+    },
+    {
+      "epoch": 1.6286151491966336,
+      "grad_norm": 2.4170362752448065,
+      "learning_rate": 1.7548682819171414e-06,
+      "loss": 0.4363,
+      "step": 10643
+    },
+    {
+      "epoch": 1.6287681713848507,
+      "grad_norm": 2.362068837715601,
+      "learning_rate": 1.7534662731760687e-06,
+      "loss": 0.3469,
+      "step": 10644
+    },
+    {
+      "epoch": 1.628921193573068,
+      "grad_norm": 2.3419814264762713,
+      "learning_rate": 1.7520647708868977e-06,
+      "loss": 0.3054,
+      "step": 10645
+    },
+    {
+      "epoch": 1.6290742157612854,
+      "grad_norm": 2.180124099872857,
+      "learning_rate": 1.7506637751356936e-06,
+      "loss": 0.309,
+      "step": 10646
+    },
+    {
+      "epoch": 1.6292272379495025,
+      "grad_norm": 2.1503373897058373,
+      "learning_rate": 1.7492632860084957e-06,
+      "loss": 0.3065,
+      "step": 10647
+    },
+    {
+      "epoch": 1.62938026013772,
+      "grad_norm": 1.7921363423401993,
+      "learning_rate": 1.7478633035913217e-06,
+      "loss": 0.275,
+      "step": 10648
+    },
+    {
+      "epoch": 1.6295332823259372,
+      "grad_norm": 2.052364874867607,
+      "learning_rate": 1.7464638279701385e-06,
+      "loss": 0.3333,
+      "step": 10649
+    },
+    {
+      "epoch": 1.6296863045141545,
+      "grad_norm": 2.027999984299312,
+      "learning_rate": 1.7450648592309039e-06,
+      "loss": 0.2747,
+      "step": 10650
+    },
+    {
+      "epoch": 1.6298393267023719,
+      "grad_norm": 1.9461536696532593,
+      "learning_rate": 1.7436663974595314e-06,
+      "loss": 0.2983,
+      "step": 10651
+    },
+    {
+      "epoch": 1.629992348890589,
+      "grad_norm": 2.1762010864690153,
+      "learning_rate": 1.742268442741899e-06,
+      "loss": 0.3249,
+      "step": 10652
+    },
+    {
+      "epoch": 1.6301453710788065,
+      "grad_norm": 1.848456650261626,
+      "learning_rate": 1.7408709951638692e-06,
+      "loss": 0.2466,
+      "step": 10653
+    },
+    {
+      "epoch": 1.6302983932670236,
+      "grad_norm": 2.1106357287709314,
+      "learning_rate": 1.7394740548112644e-06,
+      "loss": 0.2985,
+      "step": 10654
+    },
+    {
+      "epoch": 1.630451415455241,
+      "grad_norm": 2.2833521016179117,
+      "learning_rate": 1.738077621769867e-06,
+      "loss": 0.2894,
+      "step": 10655
+    },
+    {
+      "epoch": 1.6306044376434583,
+      "grad_norm": 2.0606386731477,
+      "learning_rate": 1.7366816961254463e-06,
+      "loss": 0.2897,
+      "step": 10656
+    },
+    {
+      "epoch": 1.6307574598316756,
+      "grad_norm": 1.9264653530292764,
+      "learning_rate": 1.7352862779637302e-06,
+      "loss": 0.3205,
+      "step": 10657
+    },
+    {
+      "epoch": 1.630910482019893,
+      "grad_norm": 2.0009189375636045,
+      "learning_rate": 1.7338913673704138e-06,
+      "loss": 0.2725,
+      "step": 10658
+    },
+    {
+      "epoch": 1.63106350420811,
+      "grad_norm": 2.1654583638020024,
+      "learning_rate": 1.7324969644311672e-06,
+      "loss": 0.2803,
+      "step": 10659
+    },
+    {
+      "epoch": 1.6312165263963274,
+      "grad_norm": 1.6338114436592361,
+      "learning_rate": 1.7311030692316244e-06,
+      "loss": 0.2143,
+      "step": 10660
+    },
+    {
+      "epoch": 1.6313695485845447,
+      "grad_norm": 2.2733662110358415,
+      "learning_rate": 1.72970968185739e-06,
+      "loss": 0.295,
+      "step": 10661
+    },
+    {
+      "epoch": 1.631522570772762,
+      "grad_norm": 2.1242848843223703,
+      "learning_rate": 1.7283168023940366e-06,
+      "loss": 0.3347,
+      "step": 10662
+    },
+    {
+      "epoch": 1.6316755929609794,
+      "grad_norm": 2.0020305645316747,
+      "learning_rate": 1.7269244309271083e-06,
+      "loss": 0.263,
+      "step": 10663
+    },
+    {
+      "epoch": 1.6318286151491965,
+      "grad_norm": 2.4292633335719582,
+      "learning_rate": 1.7255325675421154e-06,
+      "loss": 0.3645,
+      "step": 10664
+    },
+    {
+      "epoch": 1.631981637337414,
+      "grad_norm": 2.2540963885960483,
+      "learning_rate": 1.7241412123245372e-06,
+      "loss": 0.3293,
+      "step": 10665
+    },
+    {
+      "epoch": 1.6321346595256312,
+      "grad_norm": 1.7674776566718227,
+      "learning_rate": 1.722750365359822e-06,
+      "loss": 0.2674,
+      "step": 10666
+    },
+    {
+      "epoch": 1.6322876817138485,
+      "grad_norm": 2.0795766976304413,
+      "learning_rate": 1.7213600267333884e-06,
+      "loss": 0.351,
+      "step": 10667
+    },
+    {
+      "epoch": 1.6324407039020659,
+      "grad_norm": 2.539896141459289,
+      "learning_rate": 1.7199701965306214e-06,
+      "loss": 0.3219,
+      "step": 10668
+    },
+    {
+      "epoch": 1.632593726090283,
+      "grad_norm": 2.151179350291576,
+      "learning_rate": 1.718580874836877e-06,
+      "loss": 0.299,
+      "step": 10669
+    },
+    {
+      "epoch": 1.6327467482785005,
+      "grad_norm": 2.59810278830666,
+      "learning_rate": 1.717192061737477e-06,
+      "loss": 0.3523,
+      "step": 10670
+    },
+    {
+      "epoch": 1.6328997704667176,
+      "grad_norm": 1.9421911778296748,
+      "learning_rate": 1.7158037573177156e-06,
+      "loss": 0.2264,
+      "step": 10671
+    },
+    {
+      "epoch": 1.633052792654935,
+      "grad_norm": 2.4319925226909413,
+      "learning_rate": 1.714415961662853e-06,
+      "loss": 0.3519,
+      "step": 10672
+    },
+    {
+      "epoch": 1.6332058148431523,
+      "grad_norm": 2.196304094029701,
+      "learning_rate": 1.7130286748581183e-06,
+      "loss": 0.2792,
+      "step": 10673
+    },
+    {
+      "epoch": 1.6333588370313694,
+      "grad_norm": 1.8814963990285223,
+      "learning_rate": 1.7116418969887149e-06,
+      "loss": 0.254,
+      "step": 10674
+    },
+    {
+      "epoch": 1.633511859219587,
+      "grad_norm": 2.163967544140257,
+      "learning_rate": 1.7102556281398053e-06,
+      "loss": 0.3448,
+      "step": 10675
+    },
+    {
+      "epoch": 1.633664881407804,
+      "grad_norm": 2.3130978412784096,
+      "learning_rate": 1.7088698683965243e-06,
+      "loss": 0.3146,
+      "step": 10676
+    },
+    {
+      "epoch": 1.6338179035960214,
+      "grad_norm": 1.9372326688003392,
+      "learning_rate": 1.7074846178439853e-06,
+      "loss": 0.2683,
+      "step": 10677
+    },
+    {
+      "epoch": 1.6339709257842387,
+      "grad_norm": 2.136531609644606,
+      "learning_rate": 1.7060998765672531e-06,
+      "loss": 0.2711,
+      "step": 10678
+    },
+    {
+      "epoch": 1.6341239479724559,
+      "grad_norm": 2.2288583019578354,
+      "learning_rate": 1.7047156446513723e-06,
+      "loss": 0.3449,
+      "step": 10679
+    },
+    {
+      "epoch": 1.6342769701606734,
+      "grad_norm": 1.9287035630634422,
+      "learning_rate": 1.7033319221813593e-06,
+      "loss": 0.279,
+      "step": 10680
+    },
+    {
+      "epoch": 1.6344299923488905,
+      "grad_norm": 1.9373764849343669,
+      "learning_rate": 1.7019487092421838e-06,
+      "loss": 0.311,
+      "step": 10681
+    },
+    {
+      "epoch": 1.6345830145371079,
+      "grad_norm": 2.1086222036034785,
+      "learning_rate": 1.7005660059188034e-06,
+      "loss": 0.3082,
+      "step": 10682
+    },
+    {
+      "epoch": 1.6347360367253252,
+      "grad_norm": 1.992491784010115,
+      "learning_rate": 1.6991838122961358e-06,
+      "loss": 0.2944,
+      "step": 10683
+    },
+    {
+      "epoch": 1.6348890589135423,
+      "grad_norm": 1.948514553292706,
+      "learning_rate": 1.6978021284590562e-06,
+      "loss": 0.3036,
+      "step": 10684
+    },
+    {
+      "epoch": 1.6350420811017599,
+      "grad_norm": 2.3051469233814905,
+      "learning_rate": 1.6964209544924304e-06,
+      "loss": 0.302,
+      "step": 10685
+    },
+    {
+      "epoch": 1.635195103289977,
+      "grad_norm": 2.0805217135051053,
+      "learning_rate": 1.695040290481077e-06,
+      "loss": 0.2704,
+      "step": 10686
+    },
+    {
+      "epoch": 1.6353481254781943,
+      "grad_norm": 2.130342610927485,
+      "learning_rate": 1.6936601365097883e-06,
+      "loss": 0.3191,
+      "step": 10687
+    },
+    {
+      "epoch": 1.6355011476664116,
+      "grad_norm": 2.1838518937601763,
+      "learning_rate": 1.692280492663325e-06,
+      "loss": 0.2934,
+      "step": 10688
+    },
+    {
+      "epoch": 1.635654169854629,
+      "grad_norm": 1.6849116928654915,
+      "learning_rate": 1.6909013590264178e-06,
+      "loss": 0.1704,
+      "step": 10689
+    },
+    {
+      "epoch": 1.6358071920428463,
+      "grad_norm": 2.2691614114392835,
+      "learning_rate": 1.689522735683763e-06,
+      "loss": 0.285,
+      "step": 10690
+    },
+    {
+      "epoch": 1.6359602142310634,
+      "grad_norm": 2.0610432754287547,
+      "learning_rate": 1.6881446227200272e-06,
+      "loss": 0.2869,
+      "step": 10691
+    },
+    {
+      "epoch": 1.6361132364192807,
+      "grad_norm": 2.220578815693314,
+      "learning_rate": 1.686767020219846e-06,
+      "loss": 0.263,
+      "step": 10692
+    },
+    {
+      "epoch": 1.636266258607498,
+      "grad_norm": 2.237510475416876,
+      "learning_rate": 1.6853899282678243e-06,
+      "loss": 0.3804,
+      "step": 10693
+    },
+    {
+      "epoch": 1.6364192807957154,
+      "grad_norm": 2.3743835475077018,
+      "learning_rate": 1.6840133469485342e-06,
+      "loss": 0.3615,
+      "step": 10694
+    },
+    {
+      "epoch": 1.6365723029839327,
+      "grad_norm": 2.0430532081213353,
+      "learning_rate": 1.682637276346517e-06,
+      "loss": 0.3135,
+      "step": 10695
+    },
+    {
+      "epoch": 1.6367253251721499,
+      "grad_norm": 2.203751506438757,
+      "learning_rate": 1.681261716546282e-06,
+      "loss": 0.3146,
+      "step": 10696
+    },
+    {
+      "epoch": 1.6368783473603674,
+      "grad_norm": 2.203912076629452,
+      "learning_rate": 1.6798866676323078e-06,
+      "loss": 0.2949,
+      "step": 10697
+    },
+    {
+      "epoch": 1.6370313695485845,
+      "grad_norm": 2.1840574837792697,
+      "learning_rate": 1.6785121296890416e-06,
+      "loss": 0.2636,
+      "step": 10698
+    },
+    {
+      "epoch": 1.6371843917368019,
+      "grad_norm": 1.9430320860711479,
+      "learning_rate": 1.6771381028009003e-06,
+      "loss": 0.2496,
+      "step": 10699
+    },
+    {
+      "epoch": 1.6373374139250192,
+      "grad_norm": 2.8639026682037336,
+      "learning_rate": 1.6757645870522665e-06,
+      "loss": 0.2489,
+      "step": 10700
+    },
+    {
+      "epoch": 1.6374904361132363,
+      "grad_norm": 1.9705293616477655,
+      "learning_rate": 1.6743915825274948e-06,
+      "loss": 0.3078,
+      "step": 10701
+    },
+    {
+      "epoch": 1.6376434583014539,
+      "grad_norm": 2.1633250353549585,
+      "learning_rate": 1.6730190893109032e-06,
+      "loss": 0.303,
+      "step": 10702
+    },
+    {
+      "epoch": 1.637796480489671,
+      "grad_norm": 2.1591337070007635,
+      "learning_rate": 1.671647107486789e-06,
+      "loss": 0.2965,
+      "step": 10703
+    },
+    {
+      "epoch": 1.6379495026778883,
+      "grad_norm": 2.228367036532636,
+      "learning_rate": 1.6702756371394046e-06,
+      "loss": 0.3568,
+      "step": 10704
+    },
+    {
+      "epoch": 1.6381025248661056,
+      "grad_norm": 2.035251795089624,
+      "learning_rate": 1.668904678352977e-06,
+      "loss": 0.2833,
+      "step": 10705
+    },
+    {
+      "epoch": 1.6382555470543227,
+      "grad_norm": 2.1454705760944965,
+      "learning_rate": 1.6675342312117094e-06,
+      "loss": 0.3327,
+      "step": 10706
+    },
+    {
+      "epoch": 1.6384085692425403,
+      "grad_norm": 1.845039677873378,
+      "learning_rate": 1.6661642957997592e-06,
+      "loss": 0.2323,
+      "step": 10707
+    },
+    {
+      "epoch": 1.6385615914307574,
+      "grad_norm": 2.128005587177012,
+      "learning_rate": 1.6647948722012586e-06,
+      "loss": 0.3273,
+      "step": 10708
+    },
+    {
+      "epoch": 1.6387146136189747,
+      "grad_norm": 2.483060924862472,
+      "learning_rate": 1.6634259605003178e-06,
+      "loss": 0.3343,
+      "step": 10709
+    },
+    {
+      "epoch": 1.638867635807192,
+      "grad_norm": 1.8476913614331738,
+      "learning_rate": 1.6620575607809963e-06,
+      "loss": 0.241,
+      "step": 10710
+    },
+    {
+      "epoch": 1.6390206579954092,
+      "grad_norm": 1.873225854131616,
+      "learning_rate": 1.6606896731273414e-06,
+      "loss": 0.2667,
+      "step": 10711
+    },
+    {
+      "epoch": 1.6391736801836267,
+      "grad_norm": 2.0053587873486225,
+      "learning_rate": 1.6593222976233591e-06,
+      "loss": 0.3382,
+      "step": 10712
+    },
+    {
+      "epoch": 1.6393267023718439,
+      "grad_norm": 2.381386238829408,
+      "learning_rate": 1.6579554343530192e-06,
+      "loss": 0.3578,
+      "step": 10713
+    },
+    {
+      "epoch": 1.6394797245600612,
+      "grad_norm": 2.0143214868755015,
+      "learning_rate": 1.6565890834002718e-06,
+      "loss": 0.2752,
+      "step": 10714
+    },
+    {
+      "epoch": 1.6396327467482785,
+      "grad_norm": 2.0593276322342104,
+      "learning_rate": 1.6552232448490314e-06,
+      "loss": 0.3165,
+      "step": 10715
+    },
+    {
+      "epoch": 1.6397857689364956,
+      "grad_norm": 2.1151191992071348,
+      "learning_rate": 1.6538579187831715e-06,
+      "loss": 0.3537,
+      "step": 10716
+    },
+    {
+      "epoch": 1.6399387911247132,
+      "grad_norm": 2.2127685314861774,
+      "learning_rate": 1.6524931052865501e-06,
+      "loss": 0.2841,
+      "step": 10717
+    },
+    {
+      "epoch": 1.6400918133129303,
+      "grad_norm": 2.193174920060312,
+      "learning_rate": 1.6511288044429818e-06,
+      "loss": 0.2926,
+      "step": 10718
+    },
+    {
+      "epoch": 1.6402448355011476,
+      "grad_norm": 2.010676753287093,
+      "learning_rate": 1.6497650163362556e-06,
+      "loss": 0.2297,
+      "step": 10719
+    },
+    {
+      "epoch": 1.640397857689365,
+      "grad_norm": 2.2276182035195373,
+      "learning_rate": 1.6484017410501251e-06,
+      "loss": 0.3368,
+      "step": 10720
+    },
+    {
+      "epoch": 1.6405508798775823,
+      "grad_norm": 2.036294692508882,
+      "learning_rate": 1.647038978668316e-06,
+      "loss": 0.2753,
+      "step": 10721
+    },
+    {
+      "epoch": 1.6407039020657996,
+      "grad_norm": 1.9936948700499255,
+      "learning_rate": 1.6456767292745201e-06,
+      "loss": 0.2486,
+      "step": 10722
+    },
+    {
+      "epoch": 1.6408569242540167,
+      "grad_norm": 2.0439149869667412,
+      "learning_rate": 1.644314992952395e-06,
+      "loss": 0.27,
+      "step": 10723
+    },
+    {
+      "epoch": 1.641009946442234,
+      "grad_norm": 2.1513599382318205,
+      "learning_rate": 1.6429537697855802e-06,
+      "loss": 0.3388,
+      "step": 10724
+    },
+    {
+      "epoch": 1.6411629686304514,
+      "grad_norm": 2.5542635210630666,
+      "learning_rate": 1.6415930598576647e-06,
+      "loss": 0.3008,
+      "step": 10725
+    },
+    {
+      "epoch": 1.6413159908186687,
+      "grad_norm": 2.1803917893956997,
+      "learning_rate": 1.6402328632522147e-06,
+      "loss": 0.3298,
+      "step": 10726
+    },
+    {
+      "epoch": 1.641469013006886,
+      "grad_norm": 2.0606574091597727,
+      "learning_rate": 1.6388731800527725e-06,
+      "loss": 0.2659,
+      "step": 10727
+    },
+    {
+      "epoch": 1.6416220351951032,
+      "grad_norm": 2.5470074979530706,
+      "learning_rate": 1.6375140103428355e-06,
+      "loss": 0.3826,
+      "step": 10728
+    },
+    {
+      "epoch": 1.6417750573833207,
+      "grad_norm": 1.9157050495130552,
+      "learning_rate": 1.636155354205875e-06,
+      "loss": 0.2514,
+      "step": 10729
+    },
+    {
+      "epoch": 1.6419280795715379,
+      "grad_norm": 2.312294859804467,
+      "learning_rate": 1.6347972117253386e-06,
+      "loss": 0.3057,
+      "step": 10730
+    },
+    {
+      "epoch": 1.6420811017597552,
+      "grad_norm": 2.2051790746379005,
+      "learning_rate": 1.633439582984625e-06,
+      "loss": 0.3248,
+      "step": 10731
+    },
+    {
+      "epoch": 1.6422341239479725,
+      "grad_norm": 2.1339386560489295,
+      "learning_rate": 1.6320824680671199e-06,
+      "loss": 0.2573,
+      "step": 10732
+    },
+    {
+      "epoch": 1.6423871461361896,
+      "grad_norm": 2.2289140984293496,
+      "learning_rate": 1.6307258670561676e-06,
+      "loss": 0.2824,
+      "step": 10733
+    },
+    {
+      "epoch": 1.6425401683244072,
+      "grad_norm": 1.9393319401090998,
+      "learning_rate": 1.6293697800350761e-06,
+      "loss": 0.238,
+      "step": 10734
+    },
+    {
+      "epoch": 1.6426931905126243,
+      "grad_norm": 2.1686155278174,
+      "learning_rate": 1.6280142070871352e-06,
+      "loss": 0.2616,
+      "step": 10735
+    },
+    {
+      "epoch": 1.6428462127008416,
+      "grad_norm": 1.712168777957528,
+      "learning_rate": 1.626659148295595e-06,
+      "loss": 0.2245,
+      "step": 10736
+    },
+    {
+      "epoch": 1.642999234889059,
+      "grad_norm": 2.2094323994725866,
+      "learning_rate": 1.6253046037436693e-06,
+      "loss": 0.314,
+      "step": 10737
+    },
+    {
+      "epoch": 1.643152257077276,
+      "grad_norm": 1.836421884614005,
+      "learning_rate": 1.6239505735145511e-06,
+      "loss": 0.2484,
+      "step": 10738
+    },
+    {
+      "epoch": 1.6433052792654936,
+      "grad_norm": 2.1152088410801375,
+      "learning_rate": 1.6225970576913963e-06,
+      "loss": 0.3049,
+      "step": 10739
+    },
+    {
+      "epoch": 1.6434583014537107,
+      "grad_norm": 2.0485816420127,
+      "learning_rate": 1.6212440563573284e-06,
+      "loss": 0.2677,
+      "step": 10740
+    },
+    {
+      "epoch": 1.643611323641928,
+      "grad_norm": 2.003129581012719,
+      "learning_rate": 1.6198915695954408e-06,
+      "loss": 0.2557,
+      "step": 10741
+    },
+    {
+      "epoch": 1.6437643458301454,
+      "grad_norm": 1.7381668985451146,
+      "learning_rate": 1.6185395974887952e-06,
+      "loss": 0.2338,
+      "step": 10742
+    },
+    {
+      "epoch": 1.6439173680183625,
+      "grad_norm": 2.0542402267329876,
+      "learning_rate": 1.6171881401204215e-06,
+      "loss": 0.2303,
+      "step": 10743
+    },
+    {
+      "epoch": 1.64407039020658,
+      "grad_norm": 2.116449407226027,
+      "learning_rate": 1.6158371975733178e-06,
+      "loss": 0.2836,
+      "step": 10744
+    },
+    {
+      "epoch": 1.6442234123947972,
+      "grad_norm": 2.3970780751820944,
+      "learning_rate": 1.6144867699304512e-06,
+      "loss": 0.3241,
+      "step": 10745
+    },
+    {
+      "epoch": 1.6443764345830145,
+      "grad_norm": 1.993119568552427,
+      "learning_rate": 1.6131368572747564e-06,
+      "loss": 0.2613,
+      "step": 10746
+    },
+    {
+      "epoch": 1.6445294567712319,
+      "grad_norm": 2.109118877668952,
+      "learning_rate": 1.611787459689136e-06,
+      "loss": 0.2621,
+      "step": 10747
+    },
+    {
+      "epoch": 1.644682478959449,
+      "grad_norm": 2.1555885076869563,
+      "learning_rate": 1.6104385772564625e-06,
+      "loss": 0.2521,
+      "step": 10748
+    },
+    {
+      "epoch": 1.6448355011476665,
+      "grad_norm": 2.425326814567966,
+      "learning_rate": 1.6090902100595774e-06,
+      "loss": 0.3719,
+      "step": 10749
+    },
+    {
+      "epoch": 1.6449885233358836,
+      "grad_norm": 2.1643534965090483,
+      "learning_rate": 1.6077423581812869e-06,
+      "loss": 0.3239,
+      "step": 10750
+    },
+    {
+      "epoch": 1.645141545524101,
+      "grad_norm": 1.9450833734330513,
+      "learning_rate": 1.606395021704369e-06,
+      "loss": 0.2902,
+      "step": 10751
+    },
+    {
+      "epoch": 1.6452945677123183,
+      "grad_norm": 2.1136401370200053,
+      "learning_rate": 1.6050482007115687e-06,
+      "loss": 0.3279,
+      "step": 10752
+    },
+    {
+      "epoch": 1.6454475899005354,
+      "grad_norm": 2.065969679238715,
+      "learning_rate": 1.6037018952855998e-06,
+      "loss": 0.2373,
+      "step": 10753
+    },
+    {
+      "epoch": 1.645600612088753,
+      "grad_norm": 2.377631789499768,
+      "learning_rate": 1.6023561055091441e-06,
+      "loss": 0.2707,
+      "step": 10754
+    },
+    {
+      "epoch": 1.64575363427697,
+      "grad_norm": 1.8720142905869108,
+      "learning_rate": 1.6010108314648498e-06,
+      "loss": 0.2948,
+      "step": 10755
+    },
+    {
+      "epoch": 1.6459066564651874,
+      "grad_norm": 2.0211503929517196,
+      "learning_rate": 1.5996660732353409e-06,
+      "loss": 0.2744,
+      "step": 10756
+    },
+    {
+      "epoch": 1.6460596786534047,
+      "grad_norm": 1.8960009862980085,
+      "learning_rate": 1.5983218309031989e-06,
+      "loss": 0.296,
+      "step": 10757
+    },
+    {
+      "epoch": 1.646212700841622,
+      "grad_norm": 2.3533780465336545,
+      "learning_rate": 1.5969781045509792e-06,
+      "loss": 0.3347,
+      "step": 10758
+    },
+    {
+      "epoch": 1.6463657230298394,
+      "grad_norm": 1.8612954689985926,
+      "learning_rate": 1.595634894261211e-06,
+      "loss": 0.2747,
+      "step": 10759
+    },
+    {
+      "epoch": 1.6465187452180565,
+      "grad_norm": 2.2717995170800385,
+      "learning_rate": 1.5942922001163775e-06,
+      "loss": 0.3192,
+      "step": 10760
+    },
+    {
+      "epoch": 1.6466717674062739,
+      "grad_norm": 2.157957154061846,
+      "learning_rate": 1.5929500221989448e-06,
+      "loss": 0.3263,
+      "step": 10761
+    },
+    {
+      "epoch": 1.6468247895944912,
+      "grad_norm": 2.285376952361559,
+      "learning_rate": 1.5916083605913435e-06,
+      "loss": 0.312,
+      "step": 10762
+    },
+    {
+      "epoch": 1.6469778117827085,
+      "grad_norm": 2.137538885073889,
+      "learning_rate": 1.5902672153759613e-06,
+      "loss": 0.273,
+      "step": 10763
+    },
+    {
+      "epoch": 1.6471308339709259,
+      "grad_norm": 2.1379714062032216,
+      "learning_rate": 1.58892658663517e-06,
+      "loss": 0.3213,
+      "step": 10764
+    },
+    {
+      "epoch": 1.647283856159143,
+      "grad_norm": 1.8926610476687151,
+      "learning_rate": 1.5875864744513048e-06,
+      "loss": 0.2281,
+      "step": 10765
+    },
+    {
+      "epoch": 1.6474368783473605,
+      "grad_norm": 2.3552325347343146,
+      "learning_rate": 1.5862468789066587e-06,
+      "loss": 0.2999,
+      "step": 10766
+    },
+    {
+      "epoch": 1.6475899005355776,
+      "grad_norm": 2.0190965406363457,
+      "learning_rate": 1.5849078000835083e-06,
+      "loss": 0.2834,
+      "step": 10767
+    },
+    {
+      "epoch": 1.647742922723795,
+      "grad_norm": 1.8876247283806757,
+      "learning_rate": 1.583569238064091e-06,
+      "loss": 0.251,
+      "step": 10768
+    },
+    {
+      "epoch": 1.6478959449120123,
+      "grad_norm": 2.0416052122157824,
+      "learning_rate": 1.582231192930611e-06,
+      "loss": 0.3513,
+      "step": 10769
+    },
+    {
+      "epoch": 1.6480489671002294,
+      "grad_norm": 1.949524372805771,
+      "learning_rate": 1.580893664765245e-06,
+      "loss": 0.2473,
+      "step": 10770
+    },
+    {
+      "epoch": 1.648201989288447,
+      "grad_norm": 2.0017501920680085,
+      "learning_rate": 1.5795566536501339e-06,
+      "loss": 0.3034,
+      "step": 10771
+    },
+    {
+      "epoch": 1.648355011476664,
+      "grad_norm": 2.3663624170073243,
+      "learning_rate": 1.5782201596673908e-06,
+      "loss": 0.3085,
+      "step": 10772
+    },
+    {
+      "epoch": 1.6485080336648814,
+      "grad_norm": 1.9866391576555056,
+      "learning_rate": 1.5768841828990934e-06,
+      "loss": 0.2308,
+      "step": 10773
+    },
+    {
+      "epoch": 1.6486610558530987,
+      "grad_norm": 1.9819239180672437,
+      "learning_rate": 1.5755487234272892e-06,
+      "loss": 0.3391,
+      "step": 10774
+    },
+    {
+      "epoch": 1.6488140780413159,
+      "grad_norm": 2.210194731319105,
+      "learning_rate": 1.5742137813339942e-06,
+      "loss": 0.2937,
+      "step": 10775
+    },
+    {
+      "epoch": 1.6489671002295334,
+      "grad_norm": 2.347026374742834,
+      "learning_rate": 1.5728793567011934e-06,
+      "loss": 0.3295,
+      "step": 10776
+    },
+    {
+      "epoch": 1.6491201224177505,
+      "grad_norm": 2.005811317294377,
+      "learning_rate": 1.5715454496108384e-06,
+      "loss": 0.2935,
+      "step": 10777
+    },
+    {
+      "epoch": 1.6492731446059679,
+      "grad_norm": 2.153939130371352,
+      "learning_rate": 1.57021206014485e-06,
+      "loss": 0.3447,
+      "step": 10778
+    },
+    {
+      "epoch": 1.6494261667941852,
+      "grad_norm": 2.067837021033556,
+      "learning_rate": 1.5688791883851152e-06,
+      "loss": 0.273,
+      "step": 10779
+    },
+    {
+      "epoch": 1.6495791889824023,
+      "grad_norm": 2.0123842804250747,
+      "learning_rate": 1.5675468344134936e-06,
+      "loss": 0.2966,
+      "step": 10780
+    },
+    {
+      "epoch": 1.6497322111706199,
+      "grad_norm": 2.142451040406505,
+      "learning_rate": 1.566214998311808e-06,
+      "loss": 0.2641,
+      "step": 10781
+    },
+    {
+      "epoch": 1.649885233358837,
+      "grad_norm": 2.073097588130826,
+      "learning_rate": 1.5648836801618527e-06,
+      "loss": 0.27,
+      "step": 10782
+    },
+    {
+      "epoch": 1.6500382555470543,
+      "grad_norm": 2.483413467106734,
+      "learning_rate": 1.5635528800453892e-06,
+      "loss": 0.3521,
+      "step": 10783
+    },
+    {
+      "epoch": 1.6501912777352716,
+      "grad_norm": 2.1582050153024945,
+      "learning_rate": 1.562222598044144e-06,
+      "loss": 0.2844,
+      "step": 10784
+    },
+    {
+      "epoch": 1.6503442999234887,
+      "grad_norm": 2.176784737134013,
+      "learning_rate": 1.5608928342398232e-06,
+      "loss": 0.2599,
+      "step": 10785
+    },
+    {
+      "epoch": 1.6504973221117063,
+      "grad_norm": 2.5654567717802217,
+      "learning_rate": 1.5595635887140847e-06,
+      "loss": 0.3485,
+      "step": 10786
+    },
+    {
+      "epoch": 1.6506503442999234,
+      "grad_norm": 2.190836740951483,
+      "learning_rate": 1.5582348615485632e-06,
+      "loss": 0.2981,
+      "step": 10787
+    },
+    {
+      "epoch": 1.6508033664881407,
+      "grad_norm": 1.7488115769482173,
+      "learning_rate": 1.5569066528248676e-06,
+      "loss": 0.2574,
+      "step": 10788
+    },
+    {
+      "epoch": 1.650956388676358,
+      "grad_norm": 1.919495686623538,
+      "learning_rate": 1.5555789626245599e-06,
+      "loss": 0.2921,
+      "step": 10789
+    },
+    {
+      "epoch": 1.6511094108645754,
+      "grad_norm": 2.268063202918329,
+      "learning_rate": 1.5542517910291843e-06,
+      "loss": 0.3139,
+      "step": 10790
+    },
+    {
+      "epoch": 1.6512624330527927,
+      "grad_norm": 2.049890249473881,
+      "learning_rate": 1.5529251381202492e-06,
+      "loss": 0.301,
+      "step": 10791
+    },
+    {
+      "epoch": 1.6514154552410099,
+      "grad_norm": 2.2210056204592146,
+      "learning_rate": 1.5515990039792217e-06,
+      "loss": 0.2149,
+      "step": 10792
+    },
+    {
+      "epoch": 1.6515684774292272,
+      "grad_norm": 1.8970844610815365,
+      "learning_rate": 1.5502733886875521e-06,
+      "loss": 0.2351,
+      "step": 10793
+    },
+    {
+      "epoch": 1.6517214996174445,
+      "grad_norm": 1.987111276832506,
+      "learning_rate": 1.5489482923266519e-06,
+      "loss": 0.2797,
+      "step": 10794
+    },
+    {
+      "epoch": 1.6518745218056619,
+      "grad_norm": 2.1966652245652547,
+      "learning_rate": 1.5476237149778928e-06,
+      "loss": 0.3025,
+      "step": 10795
+    },
+    {
+      "epoch": 1.6520275439938792,
+      "grad_norm": 1.7725625214287049,
+      "learning_rate": 1.5462996567226296e-06,
+      "loss": 0.2398,
+      "step": 10796
+    },
+    {
+      "epoch": 1.6521805661820963,
+      "grad_norm": 2.095813332604148,
+      "learning_rate": 1.5449761176421752e-06,
+      "loss": 0.3259,
+      "step": 10797
+    },
+    {
+      "epoch": 1.6523335883703139,
+      "grad_norm": 2.1249341291903736,
+      "learning_rate": 1.5436530978178132e-06,
+      "loss": 0.3051,
+      "step": 10798
+    },
+    {
+      "epoch": 1.652486610558531,
+      "grad_norm": 2.414881397404002,
+      "learning_rate": 1.5423305973307966e-06,
+      "loss": 0.355,
+      "step": 10799
+    },
+    {
+      "epoch": 1.6526396327467483,
+      "grad_norm": 2.101162420996157,
+      "learning_rate": 1.541008616262345e-06,
+      "loss": 0.3226,
+      "step": 10800
+    },
+    {
+      "epoch": 1.6527926549349656,
+      "grad_norm": 2.1531043110040433,
+      "learning_rate": 1.5396871546936453e-06,
+      "loss": 0.2871,
+      "step": 10801
+    },
+    {
+      "epoch": 1.6529456771231827,
+      "grad_norm": 2.235894912765083,
+      "learning_rate": 1.5383662127058552e-06,
+      "loss": 0.2943,
+      "step": 10802
+    },
+    {
+      "epoch": 1.6530986993114003,
+      "grad_norm": 2.3541968452917006,
+      "learning_rate": 1.5370457903800973e-06,
+      "loss": 0.3262,
+      "step": 10803
+    },
+    {
+      "epoch": 1.6532517214996174,
+      "grad_norm": 1.954556761150049,
+      "learning_rate": 1.5357258877974645e-06,
+      "loss": 0.2851,
+      "step": 10804
+    },
+    {
+      "epoch": 1.6534047436878347,
+      "grad_norm": 2.526920108336962,
+      "learning_rate": 1.5344065050390155e-06,
+      "loss": 0.3662,
+      "step": 10805
+    },
+    {
+      "epoch": 1.653557765876052,
+      "grad_norm": 2.179565146522856,
+      "learning_rate": 1.5330876421857855e-06,
+      "loss": 0.3278,
+      "step": 10806
+    },
+    {
+      "epoch": 1.6537107880642692,
+      "grad_norm": 2.074137636510761,
+      "learning_rate": 1.531769299318764e-06,
+      "loss": 0.3092,
+      "step": 10807
+    },
+    {
+      "epoch": 1.6538638102524867,
+      "grad_norm": 2.525170469543575,
+      "learning_rate": 1.530451476518915e-06,
+      "loss": 0.3664,
+      "step": 10808
+    },
+    {
+      "epoch": 1.6540168324407039,
+      "grad_norm": 2.3703063766098875,
+      "learning_rate": 1.5291341738671782e-06,
+      "loss": 0.3352,
+      "step": 10809
+    },
+    {
+      "epoch": 1.6541698546289212,
+      "grad_norm": 2.3707433160320037,
+      "learning_rate": 1.5278173914444472e-06,
+      "loss": 0.3119,
+      "step": 10810
+    },
+    {
+      "epoch": 1.6543228768171385,
+      "grad_norm": 2.281609337415562,
+      "learning_rate": 1.5265011293315923e-06,
+      "loss": 0.2825,
+      "step": 10811
+    },
+    {
+      "epoch": 1.6544758990053556,
+      "grad_norm": 2.3882911358515,
+      "learning_rate": 1.5251853876094557e-06,
+      "loss": 0.2972,
+      "step": 10812
+    },
+    {
+      "epoch": 1.6546289211935732,
+      "grad_norm": 2.0222330062798695,
+      "learning_rate": 1.523870166358834e-06,
+      "loss": 0.3287,
+      "step": 10813
+    },
+    {
+      "epoch": 1.6547819433817903,
+      "grad_norm": 2.1272748346654566,
+      "learning_rate": 1.5225554656605057e-06,
+      "loss": 0.2728,
+      "step": 10814
+    },
+    {
+      "epoch": 1.6549349655700076,
+      "grad_norm": 2.120587577042761,
+      "learning_rate": 1.5212412855952131e-06,
+      "loss": 0.3192,
+      "step": 10815
+    },
+    {
+      "epoch": 1.655087987758225,
+      "grad_norm": 2.259649605135267,
+      "learning_rate": 1.519927626243658e-06,
+      "loss": 0.3069,
+      "step": 10816
+    },
+    {
+      "epoch": 1.655241009946442,
+      "grad_norm": 1.7862660695016823,
+      "learning_rate": 1.5186144876865228e-06,
+      "loss": 0.2747,
+      "step": 10817
+    },
+    {
+      "epoch": 1.6553940321346596,
+      "grad_norm": 2.1200121696415875,
+      "learning_rate": 1.517301870004454e-06,
+      "loss": 0.2835,
+      "step": 10818
+    },
+    {
+      "epoch": 1.6555470543228767,
+      "grad_norm": 2.0400460017927635,
+      "learning_rate": 1.515989773278058e-06,
+      "loss": 0.28,
+      "step": 10819
+    },
+    {
+      "epoch": 1.655700076511094,
+      "grad_norm": 1.8889820609466452,
+      "learning_rate": 1.514678197587921e-06,
+      "loss": 0.2624,
+      "step": 10820
+    },
+    {
+      "epoch": 1.6558530986993114,
+      "grad_norm": 1.9934048791905552,
+      "learning_rate": 1.5133671430145913e-06,
+      "loss": 0.2727,
+      "step": 10821
+    },
+    {
+      "epoch": 1.6560061208875287,
+      "grad_norm": 1.9503046841502563,
+      "learning_rate": 1.5120566096385846e-06,
+      "loss": 0.302,
+      "step": 10822
+    },
+    {
+      "epoch": 1.656159143075746,
+      "grad_norm": 1.8025401263324794,
+      "learning_rate": 1.5107465975403868e-06,
+      "loss": 0.2487,
+      "step": 10823
+    },
+    {
+      "epoch": 1.6563121652639632,
+      "grad_norm": 2.2381643970107246,
+      "learning_rate": 1.509437106800451e-06,
+      "loss": 0.3447,
+      "step": 10824
+    },
+    {
+      "epoch": 1.6564651874521805,
+      "grad_norm": 2.213589525316061,
+      "learning_rate": 1.5081281374991975e-06,
+      "loss": 0.3024,
+      "step": 10825
+    },
+    {
+      "epoch": 1.6566182096403979,
+      "grad_norm": 2.1435019979757364,
+      "learning_rate": 1.5068196897170152e-06,
+      "loss": 0.3324,
+      "step": 10826
+    },
+    {
+      "epoch": 1.6567712318286152,
+      "grad_norm": 2.666594275663528,
+      "learning_rate": 1.505511763534262e-06,
+      "loss": 0.3843,
+      "step": 10827
+    },
+    {
+      "epoch": 1.6569242540168325,
+      "grad_norm": 2.273364682415013,
+      "learning_rate": 1.5042043590312616e-06,
+      "loss": 0.2817,
+      "step": 10828
+    },
+    {
+      "epoch": 1.6570772762050496,
+      "grad_norm": 1.972543812149342,
+      "learning_rate": 1.5028974762883065e-06,
+      "loss": 0.3084,
+      "step": 10829
+    },
+    {
+      "epoch": 1.6572302983932672,
+      "grad_norm": 1.9225873877071888,
+      "learning_rate": 1.5015911153856588e-06,
+      "loss": 0.2399,
+      "step": 10830
+    },
+    {
+      "epoch": 1.6573833205814843,
+      "grad_norm": 1.9259450792736108,
+      "learning_rate": 1.5002852764035468e-06,
+      "loss": 0.2644,
+      "step": 10831
+    },
+    {
+      "epoch": 1.6575363427697016,
+      "grad_norm": 2.25980625558073,
+      "learning_rate": 1.4989799594221676e-06,
+      "loss": 0.2851,
+      "step": 10832
+    },
+    {
+      "epoch": 1.657689364957919,
+      "grad_norm": 1.9883780300425007,
+      "learning_rate": 1.4976751645216846e-06,
+      "loss": 0.2178,
+      "step": 10833
+    },
+    {
+      "epoch": 1.657842387146136,
+      "grad_norm": 2.194469310656908,
+      "learning_rate": 1.4963708917822283e-06,
+      "loss": 0.2982,
+      "step": 10834
+    },
+    {
+      "epoch": 1.6579954093343536,
+      "grad_norm": 2.216289901686091,
+      "learning_rate": 1.4950671412839068e-06,
+      "loss": 0.3195,
+      "step": 10835
+    },
+    {
+      "epoch": 1.6581484315225707,
+      "grad_norm": 2.5717793300199836,
+      "learning_rate": 1.4937639131067817e-06,
+      "loss": 0.3651,
+      "step": 10836
+    },
+    {
+      "epoch": 1.658301453710788,
+      "grad_norm": 2.137589206573856,
+      "learning_rate": 1.4924612073308887e-06,
+      "loss": 0.3486,
+      "step": 10837
+    },
+    {
+      "epoch": 1.6584544758990054,
+      "grad_norm": 2.154092460386197,
+      "learning_rate": 1.4911590240362395e-06,
+      "loss": 0.2565,
+      "step": 10838
+    },
+    {
+      "epoch": 1.6586074980872225,
+      "grad_norm": 2.0526315485709814,
+      "learning_rate": 1.489857363302799e-06,
+      "loss": 0.2636,
+      "step": 10839
+    },
+    {
+      "epoch": 1.65876052027544,
+      "grad_norm": 2.214596346525897,
+      "learning_rate": 1.488556225210508e-06,
+      "loss": 0.3014,
+      "step": 10840
+    },
+    {
+      "epoch": 1.6589135424636572,
+      "grad_norm": 2.0302186159970566,
+      "learning_rate": 1.4872556098392798e-06,
+      "loss": 0.306,
+      "step": 10841
+    },
+    {
+      "epoch": 1.6590665646518745,
+      "grad_norm": 1.9744823847536939,
+      "learning_rate": 1.485955517268982e-06,
+      "loss": 0.2682,
+      "step": 10842
+    },
+    {
+      "epoch": 1.6592195868400919,
+      "grad_norm": 1.970294762784387,
+      "learning_rate": 1.4846559475794653e-06,
+      "loss": 0.3153,
+      "step": 10843
+    },
+    {
+      "epoch": 1.659372609028309,
+      "grad_norm": 2.0777672103857747,
+      "learning_rate": 1.4833569008505422e-06,
+      "loss": 0.2852,
+      "step": 10844
+    },
+    {
+      "epoch": 1.6595256312165265,
+      "grad_norm": 2.187430071032332,
+      "learning_rate": 1.4820583771619845e-06,
+      "loss": 0.2573,
+      "step": 10845
+    },
+    {
+      "epoch": 1.6596786534047436,
+      "grad_norm": 2.263106010448733,
+      "learning_rate": 1.4807603765935451e-06,
+      "loss": 0.2328,
+      "step": 10846
+    },
+    {
+      "epoch": 1.659831675592961,
+      "grad_norm": 2.2069090905920876,
+      "learning_rate": 1.4794628992249427e-06,
+      "loss": 0.3041,
+      "step": 10847
+    },
+    {
+      "epoch": 1.6599846977811783,
+      "grad_norm": 2.2081482884286685,
+      "learning_rate": 1.4781659451358498e-06,
+      "loss": 0.3292,
+      "step": 10848
+    },
+    {
+      "epoch": 1.6601377199693954,
+      "grad_norm": 2.1366601858632275,
+      "learning_rate": 1.4768695144059276e-06,
+      "loss": 0.2761,
+      "step": 10849
+    },
+    {
+      "epoch": 1.660290742157613,
+      "grad_norm": 2.106645047341696,
+      "learning_rate": 1.4755736071147897e-06,
+      "loss": 0.3162,
+      "step": 10850
+    },
+    {
+      "epoch": 1.66044376434583,
+      "grad_norm": 2.3656997479027413,
+      "learning_rate": 1.474278223342026e-06,
+      "loss": 0.3084,
+      "step": 10851
+    },
+    {
+      "epoch": 1.6605967865340474,
+      "grad_norm": 2.192837488975101,
+      "learning_rate": 1.4729833631671887e-06,
+      "loss": 0.2746,
+      "step": 10852
+    },
+    {
+      "epoch": 1.6607498087222647,
+      "grad_norm": 2.139378747762477,
+      "learning_rate": 1.4716890266698002e-06,
+      "loss": 0.325,
+      "step": 10853
+    },
+    {
+      "epoch": 1.660902830910482,
+      "grad_norm": 2.3221793138150773,
+      "learning_rate": 1.470395213929352e-06,
+      "loss": 0.3499,
+      "step": 10854
+    },
+    {
+      "epoch": 1.6610558530986994,
+      "grad_norm": 1.8374502022826251,
+      "learning_rate": 1.4691019250253025e-06,
+      "loss": 0.2696,
+      "step": 10855
+    },
+    {
+      "epoch": 1.6612088752869165,
+      "grad_norm": 2.166262856949546,
+      "learning_rate": 1.4678091600370758e-06,
+      "loss": 0.2904,
+      "step": 10856
+    },
+    {
+      "epoch": 1.6613618974751339,
+      "grad_norm": 2.2757916046025533,
+      "learning_rate": 1.4665169190440664e-06,
+      "loss": 0.2943,
+      "step": 10857
+    },
+    {
+      "epoch": 1.6615149196633512,
+      "grad_norm": 1.9638029604739888,
+      "learning_rate": 1.4652252021256375e-06,
+      "loss": 0.2033,
+      "step": 10858
+    },
+    {
+      "epoch": 1.6616679418515685,
+      "grad_norm": 2.235095501903873,
+      "learning_rate": 1.4639340093611165e-06,
+      "loss": 0.3472,
+      "step": 10859
+    },
+    {
+      "epoch": 1.6618209640397859,
+      "grad_norm": 2.1923690145521393,
+      "learning_rate": 1.4626433408298014e-06,
+      "loss": 0.3573,
+      "step": 10860
+    },
+    {
+      "epoch": 1.661973986228003,
+      "grad_norm": 1.927484721477276,
+      "learning_rate": 1.4613531966109561e-06,
+      "loss": 0.2985,
+      "step": 10861
+    },
+    {
+      "epoch": 1.6621270084162203,
+      "grad_norm": 2.0199650132275404,
+      "learning_rate": 1.4600635767838155e-06,
+      "loss": 0.2646,
+      "step": 10862
+    },
+    {
+      "epoch": 1.6622800306044376,
+      "grad_norm": 2.2089164022086027,
+      "learning_rate": 1.458774481427576e-06,
+      "loss": 0.3142,
+      "step": 10863
+    },
+    {
+      "epoch": 1.662433052792655,
+      "grad_norm": 2.0269913480817245,
+      "learning_rate": 1.4574859106214144e-06,
+      "loss": 0.2498,
+      "step": 10864
+    },
+    {
+      "epoch": 1.6625860749808723,
+      "grad_norm": 2.3356780814963773,
+      "learning_rate": 1.4561978644444596e-06,
+      "loss": 0.2692,
+      "step": 10865
+    },
+    {
+      "epoch": 1.6627390971690894,
+      "grad_norm": 2.055522130474578,
+      "learning_rate": 1.4549103429758138e-06,
+      "loss": 0.2867,
+      "step": 10866
+    },
+    {
+      "epoch": 1.662892119357307,
+      "grad_norm": 1.9009789290007288,
+      "learning_rate": 1.4536233462945582e-06,
+      "loss": 0.266,
+      "step": 10867
+    },
+    {
+      "epoch": 1.663045141545524,
+      "grad_norm": 1.8610570330440441,
+      "learning_rate": 1.4523368744797239e-06,
+      "loss": 0.2756,
+      "step": 10868
+    },
+    {
+      "epoch": 1.6631981637337414,
+      "grad_norm": 2.0347304270079927,
+      "learning_rate": 1.4510509276103179e-06,
+      "loss": 0.2777,
+      "step": 10869
+    },
+    {
+      "epoch": 1.6633511859219587,
+      "grad_norm": 2.0341956783646835,
+      "learning_rate": 1.4497655057653237e-06,
+      "loss": 0.2307,
+      "step": 10870
+    },
+    {
+      "epoch": 1.6635042081101759,
+      "grad_norm": 2.103395181152377,
+      "learning_rate": 1.4484806090236736e-06,
+      "loss": 0.2885,
+      "step": 10871
+    },
+    {
+      "epoch": 1.6636572302983934,
+      "grad_norm": 2.4448504211780824,
+      "learning_rate": 1.4471962374642845e-06,
+      "loss": 0.3329,
+      "step": 10872
+    },
+    {
+      "epoch": 1.6638102524866105,
+      "grad_norm": 2.098333249073084,
+      "learning_rate": 1.4459123911660366e-06,
+      "loss": 0.2922,
+      "step": 10873
+    },
+    {
+      "epoch": 1.6639632746748279,
+      "grad_norm": 2.0247796530057687,
+      "learning_rate": 1.444629070207767e-06,
+      "loss": 0.2942,
+      "step": 10874
+    },
+    {
+      "epoch": 1.6641162968630452,
+      "grad_norm": 2.1636751847253457,
+      "learning_rate": 1.443346274668298e-06,
+      "loss": 0.3873,
+      "step": 10875
+    },
+    {
+      "epoch": 1.6642693190512623,
+      "grad_norm": 2.231073006220205,
+      "learning_rate": 1.44206400462641e-06,
+      "loss": 0.3651,
+      "step": 10876
+    },
+    {
+      "epoch": 1.6644223412394799,
+      "grad_norm": 1.656595975610177,
+      "learning_rate": 1.4407822601608457e-06,
+      "loss": 0.2371,
+      "step": 10877
+    },
+    {
+      "epoch": 1.664575363427697,
+      "grad_norm": 1.969119693034,
+      "learning_rate": 1.4395010413503297e-06,
+      "loss": 0.235,
+      "step": 10878
+    },
+    {
+      "epoch": 1.6647283856159143,
+      "grad_norm": 1.8045403787736842,
+      "learning_rate": 1.4382203482735425e-06,
+      "loss": 0.2718,
+      "step": 10879
+    },
+    {
+      "epoch": 1.6648814078041316,
+      "grad_norm": 1.9410478871577124,
+      "learning_rate": 1.4369401810091377e-06,
+      "loss": 0.3118,
+      "step": 10880
+    },
+    {
+      "epoch": 1.6650344299923487,
+      "grad_norm": 2.1581941318540356,
+      "learning_rate": 1.435660539635736e-06,
+      "loss": 0.2823,
+      "step": 10881
+    },
+    {
+      "epoch": 1.6651874521805663,
+      "grad_norm": 2.3185022235247192,
+      "learning_rate": 1.4343814242319243e-06,
+      "loss": 0.4697,
+      "step": 10882
+    },
+    {
+      "epoch": 1.6653404743687834,
+      "grad_norm": 1.7032804581381125,
+      "learning_rate": 1.4331028348762577e-06,
+      "loss": 0.2148,
+      "step": 10883
+    },
+    {
+      "epoch": 1.6654934965570007,
+      "grad_norm": 2.0705076500454287,
+      "learning_rate": 1.43182477164726e-06,
+      "loss": 0.2949,
+      "step": 10884
+    },
+    {
+      "epoch": 1.665646518745218,
+      "grad_norm": 2.0861927158849665,
+      "learning_rate": 1.4305472346234227e-06,
+      "loss": 0.3027,
+      "step": 10885
+    },
+    {
+      "epoch": 1.6657995409334352,
+      "grad_norm": 2.204515921292248,
+      "learning_rate": 1.4292702238832035e-06,
+      "loss": 0.3218,
+      "step": 10886
+    },
+    {
+      "epoch": 1.6659525631216527,
+      "grad_norm": 2.1139004154729792,
+      "learning_rate": 1.4279937395050258e-06,
+      "loss": 0.2753,
+      "step": 10887
+    },
+    {
+      "epoch": 1.6661055853098699,
+      "grad_norm": 1.9279461931220945,
+      "learning_rate": 1.4267177815672906e-06,
+      "loss": 0.2917,
+      "step": 10888
+    },
+    {
+      "epoch": 1.6662586074980872,
+      "grad_norm": 1.877735560651574,
+      "learning_rate": 1.4254423501483538e-06,
+      "loss": 0.2467,
+      "step": 10889
+    },
+    {
+      "epoch": 1.6664116296863045,
+      "grad_norm": 2.3057435678936398,
+      "learning_rate": 1.4241674453265441e-06,
+      "loss": 0.289,
+      "step": 10890
+    },
+    {
+      "epoch": 1.6665646518745219,
+      "grad_norm": 2.1293163309803123,
+      "learning_rate": 1.4228930671801645e-06,
+      "loss": 0.2736,
+      "step": 10891
+    },
+    {
+      "epoch": 1.6667176740627392,
+      "grad_norm": 2.3130308684961656,
+      "learning_rate": 1.421619215787473e-06,
+      "loss": 0.3645,
+      "step": 10892
+    },
+    {
+      "epoch": 1.6668706962509563,
+      "grad_norm": 1.9476549249402393,
+      "learning_rate": 1.4203458912267032e-06,
+      "loss": 0.1831,
+      "step": 10893
+    },
+    {
+      "epoch": 1.6670237184391736,
+      "grad_norm": 1.5150591864711775,
+      "learning_rate": 1.4190730935760589e-06,
+      "loss": 0.2004,
+      "step": 10894
+    },
+    {
+      "epoch": 1.667176740627391,
+      "grad_norm": 2.1337667385478625,
+      "learning_rate": 1.4178008229137008e-06,
+      "loss": 0.3594,
+      "step": 10895
+    },
+    {
+      "epoch": 1.6673297628156083,
+      "grad_norm": 2.0534710712272726,
+      "learning_rate": 1.416529079317771e-06,
+      "loss": 0.2937,
+      "step": 10896
+    },
+    {
+      "epoch": 1.6674827850038256,
+      "grad_norm": 2.0695438008325646,
+      "learning_rate": 1.4152578628663706e-06,
+      "loss": 0.3173,
+      "step": 10897
+    },
+    {
+      "epoch": 1.6676358071920427,
+      "grad_norm": 1.994886244529576,
+      "learning_rate": 1.4139871736375643e-06,
+      "loss": 0.3141,
+      "step": 10898
+    },
+    {
+      "epoch": 1.6677888293802603,
+      "grad_norm": 2.387752398099085,
+      "learning_rate": 1.4127170117093958e-06,
+      "loss": 0.375,
+      "step": 10899
+    },
+    {
+      "epoch": 1.6679418515684774,
+      "grad_norm": 2.190527276810829,
+      "learning_rate": 1.4114473771598702e-06,
+      "loss": 0.2836,
+      "step": 10900
+    },
+    {
+      "epoch": 1.6680948737566947,
+      "grad_norm": 2.670201984085481,
+      "learning_rate": 1.4101782700669597e-06,
+      "loss": 0.3888,
+      "step": 10901
+    },
+    {
+      "epoch": 1.668247895944912,
+      "grad_norm": 2.1367696600688304,
+      "learning_rate": 1.4089096905086053e-06,
+      "loss": 0.3103,
+      "step": 10902
+    },
+    {
+      "epoch": 1.6684009181331292,
+      "grad_norm": 1.7876575207949825,
+      "learning_rate": 1.4076416385627146e-06,
+      "loss": 0.2303,
+      "step": 10903
+    },
+    {
+      "epoch": 1.6685539403213467,
+      "grad_norm": 2.411778716322336,
+      "learning_rate": 1.4063741143071652e-06,
+      "loss": 0.2621,
+      "step": 10904
+    },
+    {
+      "epoch": 1.6687069625095639,
+      "grad_norm": 2.160323074264381,
+      "learning_rate": 1.4051071178197996e-06,
+      "loss": 0.2919,
+      "step": 10905
+    },
+    {
+      "epoch": 1.6688599846977812,
+      "grad_norm": 2.0159113695934696,
+      "learning_rate": 1.4038406491784285e-06,
+      "loss": 0.2663,
+      "step": 10906
+    },
+    {
+      "epoch": 1.6690130068859985,
+      "grad_norm": 1.8991370060184651,
+      "learning_rate": 1.4025747084608322e-06,
+      "loss": 0.2995,
+      "step": 10907
+    },
+    {
+      "epoch": 1.6691660290742156,
+      "grad_norm": 1.9419152106094528,
+      "learning_rate": 1.401309295744756e-06,
+      "loss": 0.2903,
+      "step": 10908
+    },
+    {
+      "epoch": 1.6693190512624332,
+      "grad_norm": 1.7963355100642329,
+      "learning_rate": 1.4000444111079147e-06,
+      "loss": 0.2973,
+      "step": 10909
+    },
+    {
+      "epoch": 1.6694720734506503,
+      "grad_norm": 1.8830633868108124,
+      "learning_rate": 1.3987800546279895e-06,
+      "loss": 0.2105,
+      "step": 10910
+    },
+    {
+      "epoch": 1.6696250956388676,
+      "grad_norm": 2.2856483161991092,
+      "learning_rate": 1.397516226382629e-06,
+      "loss": 0.2671,
+      "step": 10911
+    },
+    {
+      "epoch": 1.669778117827085,
+      "grad_norm": 1.771961190145075,
+      "learning_rate": 1.3962529264494507e-06,
+      "loss": 0.1899,
+      "step": 10912
+    },
+    {
+      "epoch": 1.669931140015302,
+      "grad_norm": 1.8975790556274168,
+      "learning_rate": 1.3949901549060373e-06,
+      "loss": 0.2446,
+      "step": 10913
+    },
+    {
+      "epoch": 1.6700841622035196,
+      "grad_norm": 1.596883540271513,
+      "learning_rate": 1.3937279118299417e-06,
+      "loss": 0.1948,
+      "step": 10914
+    },
+    {
+      "epoch": 1.6702371843917367,
+      "grad_norm": 2.1678140693547583,
+      "learning_rate": 1.392466197298683e-06,
+      "loss": 0.2797,
+      "step": 10915
+    },
+    {
+      "epoch": 1.670390206579954,
+      "grad_norm": 2.5530718585428835,
+      "learning_rate": 1.3912050113897457e-06,
+      "loss": 0.3456,
+      "step": 10916
+    },
+    {
+      "epoch": 1.6705432287681714,
+      "grad_norm": 2.055834230738391,
+      "learning_rate": 1.3899443541805912e-06,
+      "loss": 0.272,
+      "step": 10917
+    },
+    {
+      "epoch": 1.6706962509563885,
+      "grad_norm": 2.2459829700144107,
+      "learning_rate": 1.3886842257486344e-06,
+      "loss": 0.3651,
+      "step": 10918
+    },
+    {
+      "epoch": 1.670849273144606,
+      "grad_norm": 2.1164224680206662,
+      "learning_rate": 1.3874246261712644e-06,
+      "loss": 0.2782,
+      "step": 10919
+    },
+    {
+      "epoch": 1.6710022953328232,
+      "grad_norm": 1.8515975103873454,
+      "learning_rate": 1.3861655555258457e-06,
+      "loss": 0.2176,
+      "step": 10920
+    },
+    {
+      "epoch": 1.6711553175210405,
+      "grad_norm": 2.025232486640509,
+      "learning_rate": 1.3849070138896948e-06,
+      "loss": 0.2984,
+      "step": 10921
+    },
+    {
+      "epoch": 1.6713083397092579,
+      "grad_norm": 2.182041754809988,
+      "learning_rate": 1.3836490013401049e-06,
+      "loss": 0.3218,
+      "step": 10922
+    },
+    {
+      "epoch": 1.6714613618974752,
+      "grad_norm": 2.104884022879343,
+      "learning_rate": 1.3823915179543411e-06,
+      "loss": 0.3369,
+      "step": 10923
+    },
+    {
+      "epoch": 1.6716143840856925,
+      "grad_norm": 2.259995500180198,
+      "learning_rate": 1.381134563809623e-06,
+      "loss": 0.3001,
+      "step": 10924
+    },
+    {
+      "epoch": 1.6717674062739096,
+      "grad_norm": 1.9805033206567582,
+      "learning_rate": 1.3798781389831505e-06,
+      "loss": 0.245,
+      "step": 10925
+    },
+    {
+      "epoch": 1.671920428462127,
+      "grad_norm": 2.108247142452472,
+      "learning_rate": 1.3786222435520846e-06,
+      "loss": 0.2678,
+      "step": 10926
+    },
+    {
+      "epoch": 1.6720734506503443,
+      "grad_norm": 2.2290200835316747,
+      "learning_rate": 1.377366877593551e-06,
+      "loss": 0.2508,
+      "step": 10927
+    },
+    {
+      "epoch": 1.6722264728385616,
+      "grad_norm": 2.2448037125405973,
+      "learning_rate": 1.3761120411846506e-06,
+      "loss": 0.3183,
+      "step": 10928
+    },
+    {
+      "epoch": 1.672379495026779,
+      "grad_norm": 2.168758056304277,
+      "learning_rate": 1.374857734402446e-06,
+      "loss": 0.3377,
+      "step": 10929
+    },
+    {
+      "epoch": 1.672532517214996,
+      "grad_norm": 1.8955123459943806,
+      "learning_rate": 1.3736039573239712e-06,
+      "loss": 0.2873,
+      "step": 10930
+    },
+    {
+      "epoch": 1.6726855394032136,
+      "grad_norm": 2.1207109975957437,
+      "learning_rate": 1.3723507100262234e-06,
+      "loss": 0.2861,
+      "step": 10931
+    },
+    {
+      "epoch": 1.6728385615914307,
+      "grad_norm": 1.9763204898257254,
+      "learning_rate": 1.37109799258617e-06,
+      "loss": 0.2584,
+      "step": 10932
+    },
+    {
+      "epoch": 1.672991583779648,
+      "grad_norm": 2.475408425541011,
+      "learning_rate": 1.3698458050807451e-06,
+      "loss": 0.3446,
+      "step": 10933
+    },
+    {
+      "epoch": 1.6731446059678654,
+      "grad_norm": 2.092829234547134,
+      "learning_rate": 1.3685941475868502e-06,
+      "loss": 0.3435,
+      "step": 10934
+    },
+    {
+      "epoch": 1.6732976281560825,
+      "grad_norm": 2.40580483114001,
+      "learning_rate": 1.3673430201813553e-06,
+      "loss": 0.3217,
+      "step": 10935
+    },
+    {
+      "epoch": 1.6734506503443,
+      "grad_norm": 2.17627753965489,
+      "learning_rate": 1.3660924229410976e-06,
+      "loss": 0.2959,
+      "step": 10936
+    },
+    {
+      "epoch": 1.6736036725325172,
+      "grad_norm": 2.753710906264951,
+      "learning_rate": 1.3648423559428792e-06,
+      "loss": 0.3364,
+      "step": 10937
+    },
+    {
+      "epoch": 1.6737566947207345,
+      "grad_norm": 2.4017439533471885,
+      "learning_rate": 1.3635928192634728e-06,
+      "loss": 0.3363,
+      "step": 10938
+    },
+    {
+      "epoch": 1.6739097169089519,
+      "grad_norm": 2.0006436609312455,
+      "learning_rate": 1.3623438129796173e-06,
+      "loss": 0.3522,
+      "step": 10939
+    },
+    {
+      "epoch": 1.674062739097169,
+      "grad_norm": 2.0945862255681624,
+      "learning_rate": 1.361095337168019e-06,
+      "loss": 0.283,
+      "step": 10940
+    },
+    {
+      "epoch": 1.6742157612853865,
+      "grad_norm": 1.9868929754295699,
+      "learning_rate": 1.3598473919053524e-06,
+      "loss": 0.2842,
+      "step": 10941
+    },
+    {
+      "epoch": 1.6743687834736036,
+      "grad_norm": 1.7594827525908399,
+      "learning_rate": 1.3585999772682568e-06,
+      "loss": 0.2647,
+      "step": 10942
+    },
+    {
+      "epoch": 1.674521805661821,
+      "grad_norm": 2.061653215879647,
+      "learning_rate": 1.3573530933333423e-06,
+      "loss": 0.2889,
+      "step": 10943
+    },
+    {
+      "epoch": 1.6746748278500383,
+      "grad_norm": 2.0181626743821615,
+      "learning_rate": 1.3561067401771854e-06,
+      "loss": 0.3083,
+      "step": 10944
+    },
+    {
+      "epoch": 1.6748278500382554,
+      "grad_norm": 2.2322001593250973,
+      "learning_rate": 1.354860917876325e-06,
+      "loss": 0.3274,
+      "step": 10945
+    },
+    {
+      "epoch": 1.674980872226473,
+      "grad_norm": 2.033061742018054,
+      "learning_rate": 1.3536156265072808e-06,
+      "loss": 0.2822,
+      "step": 10946
+    },
+    {
+      "epoch": 1.67513389441469,
+      "grad_norm": 2.0820930216656577,
+      "learning_rate": 1.352370866146524e-06,
+      "loss": 0.2432,
+      "step": 10947
+    },
+    {
+      "epoch": 1.6752869166029074,
+      "grad_norm": 2.2752337340265045,
+      "learning_rate": 1.3511266368704989e-06,
+      "loss": 0.2551,
+      "step": 10948
+    },
+    {
+      "epoch": 1.6754399387911247,
+      "grad_norm": 2.2610051373444007,
+      "learning_rate": 1.3498829387556268e-06,
+      "loss": 0.3106,
+      "step": 10949
+    },
+    {
+      "epoch": 1.6755929609793418,
+      "grad_norm": 1.7493537756146116,
+      "learning_rate": 1.3486397718782796e-06,
+      "loss": 0.2036,
+      "step": 10950
+    },
+    {
+      "epoch": 1.6757459831675594,
+      "grad_norm": 2.3185756256801606,
+      "learning_rate": 1.3473971363148074e-06,
+      "loss": 0.3175,
+      "step": 10951
+    },
+    {
+      "epoch": 1.6758990053557765,
+      "grad_norm": 2.425341043817578,
+      "learning_rate": 1.3461550321415295e-06,
+      "loss": 0.3125,
+      "step": 10952
+    },
+    {
+      "epoch": 1.6760520275439938,
+      "grad_norm": 2.432317825748467,
+      "learning_rate": 1.3449134594347213e-06,
+      "loss": 0.3109,
+      "step": 10953
+    },
+    {
+      "epoch": 1.6762050497322112,
+      "grad_norm": 2.5743220175885297,
+      "learning_rate": 1.3436724182706396e-06,
+      "loss": 0.305,
+      "step": 10954
+    },
+    {
+      "epoch": 1.6763580719204285,
+      "grad_norm": 1.9770393582495958,
+      "learning_rate": 1.342431908725499e-06,
+      "loss": 0.2984,
+      "step": 10955
+    },
+    {
+      "epoch": 1.6765110941086458,
+      "grad_norm": 2.1248514001509116,
+      "learning_rate": 1.3411919308754794e-06,
+      "loss": 0.3008,
+      "step": 10956
+    },
+    {
+      "epoch": 1.676664116296863,
+      "grad_norm": 2.0468319850996872,
+      "learning_rate": 1.3399524847967405e-06,
+      "loss": 0.2727,
+      "step": 10957
+    },
+    {
+      "epoch": 1.6768171384850803,
+      "grad_norm": 2.036751420201803,
+      "learning_rate": 1.3387135705653986e-06,
+      "loss": 0.274,
+      "step": 10958
+    },
+    {
+      "epoch": 1.6769701606732976,
+      "grad_norm": 1.9180774151723226,
+      "learning_rate": 1.3374751882575354e-06,
+      "loss": 0.2751,
+      "step": 10959
+    },
+    {
+      "epoch": 1.677123182861515,
+      "grad_norm": 2.4520454556972746,
+      "learning_rate": 1.336237337949211e-06,
+      "loss": 0.3382,
+      "step": 10960
+    },
+    {
+      "epoch": 1.6772762050497323,
+      "grad_norm": 1.761323796608222,
+      "learning_rate": 1.3350000197164436e-06,
+      "loss": 0.2501,
+      "step": 10961
+    },
+    {
+      "epoch": 1.6774292272379494,
+      "grad_norm": 2.062316820272847,
+      "learning_rate": 1.3337632336352234e-06,
+      "loss": 0.2693,
+      "step": 10962
+    },
+    {
+      "epoch": 1.677582249426167,
+      "grad_norm": 2.0720852517253214,
+      "learning_rate": 1.3325269797815066e-06,
+      "loss": 0.3164,
+      "step": 10963
+    },
+    {
+      "epoch": 1.677735271614384,
+      "grad_norm": 2.27703499150179,
+      "learning_rate": 1.3312912582312143e-06,
+      "loss": 0.3367,
+      "step": 10964
+    },
+    {
+      "epoch": 1.6778882938026014,
+      "grad_norm": 1.960915018382373,
+      "learning_rate": 1.3300560690602382e-06,
+      "loss": 0.2368,
+      "step": 10965
+    },
+    {
+      "epoch": 1.6780413159908187,
+      "grad_norm": 1.9766829049595693,
+      "learning_rate": 1.328821412344433e-06,
+      "loss": 0.2469,
+      "step": 10966
+    },
+    {
+      "epoch": 1.6781943381790358,
+      "grad_norm": 1.9250997866626893,
+      "learning_rate": 1.3275872881596319e-06,
+      "loss": 0.2875,
+      "step": 10967
+    },
+    {
+      "epoch": 1.6783473603672534,
+      "grad_norm": 1.549837070098533,
+      "learning_rate": 1.3263536965816203e-06,
+      "loss": 0.2024,
+      "step": 10968
+    },
+    {
+      "epoch": 1.6785003825554705,
+      "grad_norm": 1.7642085141734583,
+      "learning_rate": 1.3251206376861569e-06,
+      "loss": 0.248,
+      "step": 10969
+    },
+    {
+      "epoch": 1.6786534047436878,
+      "grad_norm": 2.0836891565839486,
+      "learning_rate": 1.3238881115489755e-06,
+      "loss": 0.2646,
+      "step": 10970
+    },
+    {
+      "epoch": 1.6788064269319052,
+      "grad_norm": 2.008808530117174,
+      "learning_rate": 1.3226561182457642e-06,
+      "loss": 0.3069,
+      "step": 10971
+    },
+    {
+      "epoch": 1.6789594491201223,
+      "grad_norm": 2.25601227776761,
+      "learning_rate": 1.321424657852185e-06,
+      "loss": 0.3359,
+      "step": 10972
+    },
+    {
+      "epoch": 1.6791124713083398,
+      "grad_norm": 2.028677462966108,
+      "learning_rate": 1.3201937304438728e-06,
+      "loss": 0.3192,
+      "step": 10973
+    },
+    {
+      "epoch": 1.679265493496557,
+      "grad_norm": 1.9466678524250804,
+      "learning_rate": 1.3189633360964137e-06,
+      "loss": 0.3153,
+      "step": 10974
+    },
+    {
+      "epoch": 1.6794185156847743,
+      "grad_norm": 2.045679294294252,
+      "learning_rate": 1.3177334748853798e-06,
+      "loss": 0.3224,
+      "step": 10975
+    },
+    {
+      "epoch": 1.6795715378729916,
+      "grad_norm": 1.902697425925076,
+      "learning_rate": 1.3165041468863004e-06,
+      "loss": 0.2409,
+      "step": 10976
+    },
+    {
+      "epoch": 1.6797245600612087,
+      "grad_norm": 1.9497283206798874,
+      "learning_rate": 1.3152753521746676e-06,
+      "loss": 0.2643,
+      "step": 10977
+    },
+    {
+      "epoch": 1.6798775822494263,
+      "grad_norm": 2.1252780328224996,
+      "learning_rate": 1.3140470908259517e-06,
+      "loss": 0.327,
+      "step": 10978
+    },
+    {
+      "epoch": 1.6800306044376434,
+      "grad_norm": 2.0928532425925597,
+      "learning_rate": 1.312819362915585e-06,
+      "loss": 0.2725,
+      "step": 10979
+    },
+    {
+      "epoch": 1.6801836266258607,
+      "grad_norm": 2.2119579928711413,
+      "learning_rate": 1.3115921685189625e-06,
+      "loss": 0.2886,
+      "step": 10980
+    },
+    {
+      "epoch": 1.680336648814078,
+      "grad_norm": 1.7936452479086349,
+      "learning_rate": 1.3103655077114563e-06,
+      "loss": 0.2175,
+      "step": 10981
+    },
+    {
+      "epoch": 1.6804896710022952,
+      "grad_norm": 2.2272606564737445,
+      "learning_rate": 1.3091393805683972e-06,
+      "loss": 0.3237,
+      "step": 10982
+    },
+    {
+      "epoch": 1.6806426931905127,
+      "grad_norm": 2.356171938700306,
+      "learning_rate": 1.3079137871650894e-06,
+      "loss": 0.2936,
+      "step": 10983
+    },
+    {
+      "epoch": 1.6807957153787298,
+      "grad_norm": 2.232352653535342,
+      "learning_rate": 1.306688727576798e-06,
+      "loss": 0.3052,
+      "step": 10984
+    },
+    {
+      "epoch": 1.6809487375669472,
+      "grad_norm": 2.1782319417814153,
+      "learning_rate": 1.3054642018787612e-06,
+      "loss": 0.3149,
+      "step": 10985
+    },
+    {
+      "epoch": 1.6811017597551645,
+      "grad_norm": 2.2406002945095738,
+      "learning_rate": 1.3042402101461804e-06,
+      "loss": 0.2902,
+      "step": 10986
+    },
+    {
+      "epoch": 1.6812547819433816,
+      "grad_norm": 1.9244081104162174,
+      "learning_rate": 1.303016752454226e-06,
+      "loss": 0.2905,
+      "step": 10987
+    },
+    {
+      "epoch": 1.6814078041315992,
+      "grad_norm": 1.797026733705302,
+      "learning_rate": 1.3017938288780363e-06,
+      "loss": 0.3086,
+      "step": 10988
+    },
+    {
+      "epoch": 1.6815608263198163,
+      "grad_norm": 1.9509682390662428,
+      "learning_rate": 1.300571439492715e-06,
+      "loss": 0.2575,
+      "step": 10989
+    },
+    {
+      "epoch": 1.6817138485080336,
+      "grad_norm": 2.45570148479363,
+      "learning_rate": 1.2993495843733339e-06,
+      "loss": 0.3735,
+      "step": 10990
+    },
+    {
+      "epoch": 1.681866870696251,
+      "grad_norm": 1.949368989482816,
+      "learning_rate": 1.298128263594932e-06,
+      "loss": 0.3218,
+      "step": 10991
+    },
+    {
+      "epoch": 1.6820198928844683,
+      "grad_norm": 2.067653056965363,
+      "learning_rate": 1.2969074772325151e-06,
+      "loss": 0.3221,
+      "step": 10992
+    },
+    {
+      "epoch": 1.6821729150726856,
+      "grad_norm": 2.311033815856632,
+      "learning_rate": 1.2956872253610576e-06,
+      "loss": 0.3396,
+      "step": 10993
+    },
+    {
+      "epoch": 1.6823259372609027,
+      "grad_norm": 1.9032585435009628,
+      "learning_rate": 1.294467508055498e-06,
+      "loss": 0.2379,
+      "step": 10994
+    },
+    {
+      "epoch": 1.68247895944912,
+      "grad_norm": 2.2597516199257406,
+      "learning_rate": 1.2932483253907446e-06,
+      "loss": 0.2719,
+      "step": 10995
+    },
+    {
+      "epoch": 1.6826319816373374,
+      "grad_norm": 2.196069288251133,
+      "learning_rate": 1.2920296774416742e-06,
+      "loss": 0.3325,
+      "step": 10996
+    },
+    {
+      "epoch": 1.6827850038255547,
+      "grad_norm": 2.089356908329131,
+      "learning_rate": 1.2908115642831254e-06,
+      "loss": 0.2846,
+      "step": 10997
+    },
+    {
+      "epoch": 1.682938026013772,
+      "grad_norm": 2.500008607734077,
+      "learning_rate": 1.2895939859899075e-06,
+      "loss": 0.3552,
+      "step": 10998
+    },
+    {
+      "epoch": 1.6830910482019892,
+      "grad_norm": 2.214635489148657,
+      "learning_rate": 1.2883769426368032e-06,
+      "loss": 0.3502,
+      "step": 10999
+    },
+    {
+      "epoch": 1.6832440703902067,
+      "grad_norm": 1.915744367799281,
+      "learning_rate": 1.2871604342985478e-06,
+      "loss": 0.2487,
+      "step": 11000
+    },
+    {
+      "epoch": 1.6833970925784238,
+      "grad_norm": 2.2836247432485948,
+      "learning_rate": 1.2859444610498539e-06,
+      "loss": 0.293,
+      "step": 11001
+    },
+    {
+      "epoch": 1.6835501147666412,
+      "grad_norm": 1.904948231183335,
+      "learning_rate": 1.284729022965403e-06,
+      "loss": 0.2842,
+      "step": 11002
+    },
+    {
+      "epoch": 1.6837031369548585,
+      "grad_norm": 2.108321600771051,
+      "learning_rate": 1.2835141201198343e-06,
+      "loss": 0.2656,
+      "step": 11003
+    },
+    {
+      "epoch": 1.6838561591430756,
+      "grad_norm": 2.093179930514413,
+      "learning_rate": 1.2822997525877633e-06,
+      "loss": 0.2816,
+      "step": 11004
+    },
+    {
+      "epoch": 1.6840091813312932,
+      "grad_norm": 2.377483764466178,
+      "learning_rate": 1.281085920443772e-06,
+      "loss": 0.3617,
+      "step": 11005
+    },
+    {
+      "epoch": 1.6841622035195103,
+      "grad_norm": 2.167029247250692,
+      "learning_rate": 1.2798726237623971e-06,
+      "loss": 0.2789,
+      "step": 11006
+    },
+    {
+      "epoch": 1.6843152257077276,
+      "grad_norm": 2.182866340215894,
+      "learning_rate": 1.2786598626181613e-06,
+      "loss": 0.2687,
+      "step": 11007
+    },
+    {
+      "epoch": 1.684468247895945,
+      "grad_norm": 1.9979186288108053,
+      "learning_rate": 1.2774476370855426e-06,
+      "loss": 0.2822,
+      "step": 11008
+    },
+    {
+      "epoch": 1.684621270084162,
+      "grad_norm": 2.1936633301966597,
+      "learning_rate": 1.276235947238983e-06,
+      "loss": 0.2928,
+      "step": 11009
+    },
+    {
+      "epoch": 1.6847742922723796,
+      "grad_norm": 2.104500943307297,
+      "learning_rate": 1.2750247931529035e-06,
+      "loss": 0.2523,
+      "step": 11010
+    },
+    {
+      "epoch": 1.6849273144605967,
+      "grad_norm": 1.920693123555798,
+      "learning_rate": 1.273814174901684e-06,
+      "loss": 0.3069,
+      "step": 11011
+    },
+    {
+      "epoch": 1.685080336648814,
+      "grad_norm": 2.299114587660011,
+      "learning_rate": 1.272604092559674e-06,
+      "loss": 0.295,
+      "step": 11012
+    },
+    {
+      "epoch": 1.6852333588370314,
+      "grad_norm": 2.1949856289025647,
+      "learning_rate": 1.2713945462011868e-06,
+      "loss": 0.2994,
+      "step": 11013
+    },
+    {
+      "epoch": 1.6853863810252485,
+      "grad_norm": 1.9841400340099198,
+      "learning_rate": 1.2701855359005077e-06,
+      "loss": 0.2852,
+      "step": 11014
+    },
+    {
+      "epoch": 1.685539403213466,
+      "grad_norm": 2.4095612568256866,
+      "learning_rate": 1.2689770617318864e-06,
+      "loss": 0.2926,
+      "step": 11015
+    },
+    {
+      "epoch": 1.6856924254016832,
+      "grad_norm": 2.067956237774031,
+      "learning_rate": 1.2677691237695399e-06,
+      "loss": 0.3413,
+      "step": 11016
+    },
+    {
+      "epoch": 1.6858454475899005,
+      "grad_norm": 2.2597156371173814,
+      "learning_rate": 1.2665617220876513e-06,
+      "loss": 0.3331,
+      "step": 11017
+    },
+    {
+      "epoch": 1.6859984697781178,
+      "grad_norm": 2.1700644455923297,
+      "learning_rate": 1.2653548567603736e-06,
+      "loss": 0.3578,
+      "step": 11018
+    },
+    {
+      "epoch": 1.686151491966335,
+      "grad_norm": 2.0662294654838527,
+      "learning_rate": 1.2641485278618237e-06,
+      "loss": 0.2941,
+      "step": 11019
+    },
+    {
+      "epoch": 1.6863045141545525,
+      "grad_norm": 2.129039860992622,
+      "learning_rate": 1.2629427354660872e-06,
+      "loss": 0.3242,
+      "step": 11020
+    },
+    {
+      "epoch": 1.6864575363427696,
+      "grad_norm": 2.3578673412858,
+      "learning_rate": 1.2617374796472181e-06,
+      "loss": 0.3425,
+      "step": 11021
+    },
+    {
+      "epoch": 1.686610558530987,
+      "grad_norm": 2.057803857866164,
+      "learning_rate": 1.260532760479234e-06,
+      "loss": 0.2395,
+      "step": 11022
+    },
+    {
+      "epoch": 1.6867635807192043,
+      "grad_norm": 2.0059512760022593,
+      "learning_rate": 1.259328578036122e-06,
+      "loss": 0.3397,
+      "step": 11023
+    },
+    {
+      "epoch": 1.6869166029074216,
+      "grad_norm": 2.049651286807147,
+      "learning_rate": 1.2581249323918354e-06,
+      "loss": 0.2789,
+      "step": 11024
+    },
+    {
+      "epoch": 1.687069625095639,
+      "grad_norm": 2.499375308441647,
+      "learning_rate": 1.2569218236202952e-06,
+      "loss": 0.3722,
+      "step": 11025
+    },
+    {
+      "epoch": 1.687222647283856,
+      "grad_norm": 1.960400132055227,
+      "learning_rate": 1.2557192517953897e-06,
+      "loss": 0.2968,
+      "step": 11026
+    },
+    {
+      "epoch": 1.6873756694720734,
+      "grad_norm": 2.1366266710985413,
+      "learning_rate": 1.2545172169909703e-06,
+      "loss": 0.283,
+      "step": 11027
+    },
+    {
+      "epoch": 1.6875286916602907,
+      "grad_norm": 2.039190519536008,
+      "learning_rate": 1.2533157192808644e-06,
+      "loss": 0.3212,
+      "step": 11028
+    },
+    {
+      "epoch": 1.687681713848508,
+      "grad_norm": 2.0949669870871093,
+      "learning_rate": 1.252114758738856e-06,
+      "loss": 0.2679,
+      "step": 11029
+    },
+    {
+      "epoch": 1.6878347360367254,
+      "grad_norm": 2.0416473705727487,
+      "learning_rate": 1.250914335438701e-06,
+      "loss": 0.2916,
+      "step": 11030
+    },
+    {
+      "epoch": 1.6879877582249425,
+      "grad_norm": 2.09341563269475,
+      "learning_rate": 1.2497144494541258e-06,
+      "loss": 0.2946,
+      "step": 11031
+    },
+    {
+      "epoch": 1.68814078041316,
+      "grad_norm": 2.270626399599111,
+      "learning_rate": 1.248515100858817e-06,
+      "loss": 0.331,
+      "step": 11032
+    },
+    {
+      "epoch": 1.6882938026013772,
+      "grad_norm": 2.2390309838409883,
+      "learning_rate": 1.247316289726429e-06,
+      "loss": 0.3092,
+      "step": 11033
+    },
+    {
+      "epoch": 1.6884468247895945,
+      "grad_norm": 2.1983290793456254,
+      "learning_rate": 1.2461180161305919e-06,
+      "loss": 0.3022,
+      "step": 11034
+    },
+    {
+      "epoch": 1.6885998469778118,
+      "grad_norm": 2.0568056713653875,
+      "learning_rate": 1.24492028014489e-06,
+      "loss": 0.3016,
+      "step": 11035
+    },
+    {
+      "epoch": 1.688752869166029,
+      "grad_norm": 2.3206498756421152,
+      "learning_rate": 1.2437230818428846e-06,
+      "loss": 0.3531,
+      "step": 11036
+    },
+    {
+      "epoch": 1.6889058913542465,
+      "grad_norm": 2.0813455867828643,
+      "learning_rate": 1.2425264212981024e-06,
+      "loss": 0.2628,
+      "step": 11037
+    },
+    {
+      "epoch": 1.6890589135424636,
+      "grad_norm": 2.0413458094502785,
+      "learning_rate": 1.241330298584027e-06,
+      "loss": 0.3088,
+      "step": 11038
+    },
+    {
+      "epoch": 1.689211935730681,
+      "grad_norm": 2.1582679198860637,
+      "learning_rate": 1.2401347137741248e-06,
+      "loss": 0.3091,
+      "step": 11039
+    },
+    {
+      "epoch": 1.6893649579188983,
+      "grad_norm": 2.069240644552984,
+      "learning_rate": 1.2389396669418185e-06,
+      "loss": 0.2866,
+      "step": 11040
+    },
+    {
+      "epoch": 1.6895179801071154,
+      "grad_norm": 1.9584455619267858,
+      "learning_rate": 1.2377451581605015e-06,
+      "loss": 0.3061,
+      "step": 11041
+    },
+    {
+      "epoch": 1.689671002295333,
+      "grad_norm": 2.3277588997885412,
+      "learning_rate": 1.2365511875035319e-06,
+      "loss": 0.3369,
+      "step": 11042
+    },
+    {
+      "epoch": 1.68982402448355,
+      "grad_norm": 1.9800477950424002,
+      "learning_rate": 1.2353577550442363e-06,
+      "loss": 0.2405,
+      "step": 11043
+    },
+    {
+      "epoch": 1.6899770466717674,
+      "grad_norm": 2.160902013523082,
+      "learning_rate": 1.2341648608559088e-06,
+      "loss": 0.2896,
+      "step": 11044
+    },
+    {
+      "epoch": 1.6901300688599847,
+      "grad_norm": 2.1642624810194877,
+      "learning_rate": 1.2329725050118091e-06,
+      "loss": 0.2678,
+      "step": 11045
+    },
+    {
+      "epoch": 1.6902830910482018,
+      "grad_norm": 2.209833345083377,
+      "learning_rate": 1.2317806875851646e-06,
+      "loss": 0.2937,
+      "step": 11046
+    },
+    {
+      "epoch": 1.6904361132364194,
+      "grad_norm": 2.402841329362798,
+      "learning_rate": 1.23058940864917e-06,
+      "loss": 0.286,
+      "step": 11047
+    },
+    {
+      "epoch": 1.6905891354246365,
+      "grad_norm": 2.522322338055443,
+      "learning_rate": 1.2293986682769832e-06,
+      "loss": 0.3434,
+      "step": 11048
+    },
+    {
+      "epoch": 1.6907421576128538,
+      "grad_norm": 1.9879433395929236,
+      "learning_rate": 1.2282084665417404e-06,
+      "loss": 0.2557,
+      "step": 11049
+    },
+    {
+      "epoch": 1.6908951798010712,
+      "grad_norm": 2.0561928258750366,
+      "learning_rate": 1.2270188035165277e-06,
+      "loss": 0.2496,
+      "step": 11050
+    },
+    {
+      "epoch": 1.6910482019892883,
+      "grad_norm": 1.9106083750167469,
+      "learning_rate": 1.2258296792744084e-06,
+      "loss": 0.3045,
+      "step": 11051
+    },
+    {
+      "epoch": 1.6912012241775058,
+      "grad_norm": 1.8611621144706574,
+      "learning_rate": 1.224641093888418e-06,
+      "loss": 0.2356,
+      "step": 11052
+    },
+    {
+      "epoch": 1.691354246365723,
+      "grad_norm": 2.020755692551469,
+      "learning_rate": 1.2234530474315453e-06,
+      "loss": 0.2391,
+      "step": 11053
+    },
+    {
+      "epoch": 1.6915072685539403,
+      "grad_norm": 2.3678631524009353,
+      "learning_rate": 1.2222655399767524e-06,
+      "loss": 0.2883,
+      "step": 11054
+    },
+    {
+      "epoch": 1.6916602907421576,
+      "grad_norm": 2.246798245562036,
+      "learning_rate": 1.2210785715969765e-06,
+      "loss": 0.3168,
+      "step": 11055
+    },
+    {
+      "epoch": 1.691813312930375,
+      "grad_norm": 2.1634726407960088,
+      "learning_rate": 1.2198921423651034e-06,
+      "loss": 0.2602,
+      "step": 11056
+    },
+    {
+      "epoch": 1.6919663351185923,
+      "grad_norm": 1.8821433603848743,
+      "learning_rate": 1.218706252354005e-06,
+      "loss": 0.2641,
+      "step": 11057
+    },
+    {
+      "epoch": 1.6921193573068094,
+      "grad_norm": 2.1581611539677863,
+      "learning_rate": 1.2175209016365098e-06,
+      "loss": 0.3223,
+      "step": 11058
+    },
+    {
+      "epoch": 1.6922723794950267,
+      "grad_norm": 2.07129171472233,
+      "learning_rate": 1.2163360902854094e-06,
+      "loss": 0.3231,
+      "step": 11059
+    },
+    {
+      "epoch": 1.692425401683244,
+      "grad_norm": 2.270422976688089,
+      "learning_rate": 1.2151518183734735e-06,
+      "loss": 0.311,
+      "step": 11060
+    },
+    {
+      "epoch": 1.6925784238714614,
+      "grad_norm": 1.8203368474711896,
+      "learning_rate": 1.2139680859734326e-06,
+      "loss": 0.2481,
+      "step": 11061
+    },
+    {
+      "epoch": 1.6927314460596787,
+      "grad_norm": 2.0614554546840553,
+      "learning_rate": 1.2127848931579788e-06,
+      "loss": 0.2811,
+      "step": 11062
+    },
+    {
+      "epoch": 1.6928844682478958,
+      "grad_norm": 2.196661166866587,
+      "learning_rate": 1.2116022399997828e-06,
+      "loss": 0.3167,
+      "step": 11063
+    },
+    {
+      "epoch": 1.6930374904361134,
+      "grad_norm": 2.077107983851203,
+      "learning_rate": 1.2104201265714743e-06,
+      "loss": 0.2713,
+      "step": 11064
+    },
+    {
+      "epoch": 1.6931905126243305,
+      "grad_norm": 2.0755681276903424,
+      "learning_rate": 1.2092385529456497e-06,
+      "loss": 0.2502,
+      "step": 11065
+    },
+    {
+      "epoch": 1.6933435348125478,
+      "grad_norm": 2.22983973981828,
+      "learning_rate": 1.2080575191948763e-06,
+      "loss": 0.2748,
+      "step": 11066
+    },
+    {
+      "epoch": 1.6934965570007652,
+      "grad_norm": 2.2026304230846323,
+      "learning_rate": 1.206877025391684e-06,
+      "loss": 0.2916,
+      "step": 11067
+    },
+    {
+      "epoch": 1.6936495791889823,
+      "grad_norm": 2.1680872094915915,
+      "learning_rate": 1.2056970716085724e-06,
+      "loss": 0.2673,
+      "step": 11068
+    },
+    {
+      "epoch": 1.6938026013771998,
+      "grad_norm": 1.849240136022882,
+      "learning_rate": 1.2045176579180074e-06,
+      "loss": 0.2087,
+      "step": 11069
+    },
+    {
+      "epoch": 1.693955623565417,
+      "grad_norm": 2.0659329721045308,
+      "learning_rate": 1.2033387843924215e-06,
+      "loss": 0.2818,
+      "step": 11070
+    },
+    {
+      "epoch": 1.6941086457536343,
+      "grad_norm": 2.2135172712744176,
+      "learning_rate": 1.202160451104213e-06,
+      "loss": 0.3076,
+      "step": 11071
+    },
+    {
+      "epoch": 1.6942616679418516,
+      "grad_norm": 2.20420719101936,
+      "learning_rate": 1.2009826581257488e-06,
+      "loss": 0.3086,
+      "step": 11072
+    },
+    {
+      "epoch": 1.6944146901300687,
+      "grad_norm": 1.9144814530586822,
+      "learning_rate": 1.1998054055293617e-06,
+      "loss": 0.4203,
+      "step": 11073
+    },
+    {
+      "epoch": 1.6945677123182863,
+      "grad_norm": 2.5558070625916343,
+      "learning_rate": 1.1986286933873503e-06,
+      "loss": 0.2861,
+      "step": 11074
+    },
+    {
+      "epoch": 1.6947207345065034,
+      "grad_norm": 2.2374112894675595,
+      "learning_rate": 1.1974525217719835e-06,
+      "loss": 0.3367,
+      "step": 11075
+    },
+    {
+      "epoch": 1.6948737566947207,
+      "grad_norm": 2.1946913555112637,
+      "learning_rate": 1.1962768907554923e-06,
+      "loss": 0.3348,
+      "step": 11076
+    },
+    {
+      "epoch": 1.695026778882938,
+      "grad_norm": 2.151829815904901,
+      "learning_rate": 1.1951018004100757e-06,
+      "loss": 0.3246,
+      "step": 11077
+    },
+    {
+      "epoch": 1.6951798010711552,
+      "grad_norm": 2.2967753420143513,
+      "learning_rate": 1.1939272508079058e-06,
+      "loss": 0.335,
+      "step": 11078
+    },
+    {
+      "epoch": 1.6953328232593727,
+      "grad_norm": 2.2577136557529838,
+      "learning_rate": 1.192753242021112e-06,
+      "loss": 0.2895,
+      "step": 11079
+    },
+    {
+      "epoch": 1.6954858454475898,
+      "grad_norm": 2.128267511682266,
+      "learning_rate": 1.1915797741217928e-06,
+      "loss": 0.257,
+      "step": 11080
+    },
+    {
+      "epoch": 1.6956388676358072,
+      "grad_norm": 2.4002431250160567,
+      "learning_rate": 1.190406847182023e-06,
+      "loss": 0.3277,
+      "step": 11081
+    },
+    {
+      "epoch": 1.6957918898240245,
+      "grad_norm": 2.2038719394242023,
+      "learning_rate": 1.1892344612738305e-06,
+      "loss": 0.263,
+      "step": 11082
+    },
+    {
+      "epoch": 1.6959449120122416,
+      "grad_norm": 1.8881112103439488,
+      "learning_rate": 1.1880626164692154e-06,
+      "loss": 0.2243,
+      "step": 11083
+    },
+    {
+      "epoch": 1.6960979342004592,
+      "grad_norm": 2.138352754268959,
+      "learning_rate": 1.1868913128401504e-06,
+      "loss": 0.2879,
+      "step": 11084
+    },
+    {
+      "epoch": 1.6962509563886763,
+      "grad_norm": 1.828210483145325,
+      "learning_rate": 1.1857205504585645e-06,
+      "loss": 0.2238,
+      "step": 11085
+    },
+    {
+      "epoch": 1.6964039785768936,
+      "grad_norm": 2.5405874955917516,
+      "learning_rate": 1.1845503293963623e-06,
+      "loss": 0.3505,
+      "step": 11086
+    },
+    {
+      "epoch": 1.696557000765111,
+      "grad_norm": 2.111535006004857,
+      "learning_rate": 1.183380649725413e-06,
+      "loss": 0.2496,
+      "step": 11087
+    },
+    {
+      "epoch": 1.696710022953328,
+      "grad_norm": 2.424057326610335,
+      "learning_rate": 1.1822115115175448e-06,
+      "loss": 0.3544,
+      "step": 11088
+    },
+    {
+      "epoch": 1.6968630451415456,
+      "grad_norm": 2.4592854858100512,
+      "learning_rate": 1.1810429148445645e-06,
+      "loss": 0.2626,
+      "step": 11089
+    },
+    {
+      "epoch": 1.6970160673297627,
+      "grad_norm": 2.042198436206543,
+      "learning_rate": 1.1798748597782417e-06,
+      "loss": 0.2835,
+      "step": 11090
+    },
+    {
+      "epoch": 1.69716908951798,
+      "grad_norm": 2.275740572056835,
+      "learning_rate": 1.1787073463903033e-06,
+      "loss": 0.2623,
+      "step": 11091
+    },
+    {
+      "epoch": 1.6973221117061974,
+      "grad_norm": 2.158557298967889,
+      "learning_rate": 1.1775403747524582e-06,
+      "loss": 0.2947,
+      "step": 11092
+    },
+    {
+      "epoch": 1.6974751338944147,
+      "grad_norm": 1.9757088684466695,
+      "learning_rate": 1.176373944936372e-06,
+      "loss": 0.2579,
+      "step": 11093
+    },
+    {
+      "epoch": 1.697628156082632,
+      "grad_norm": 1.7875604110832137,
+      "learning_rate": 1.1752080570136814e-06,
+      "loss": 0.2421,
+      "step": 11094
+    },
+    {
+      "epoch": 1.6977811782708492,
+      "grad_norm": 2.2847682190407674,
+      "learning_rate": 1.174042711055986e-06,
+      "loss": 0.2785,
+      "step": 11095
+    },
+    {
+      "epoch": 1.6979342004590665,
+      "grad_norm": 2.1586205655942345,
+      "learning_rate": 1.172877907134855e-06,
+      "loss": 0.2986,
+      "step": 11096
+    },
+    {
+      "epoch": 1.6980872226472838,
+      "grad_norm": 2.5074890553104403,
+      "learning_rate": 1.1717136453218236e-06,
+      "loss": 0.3686,
+      "step": 11097
+    },
+    {
+      "epoch": 1.6982402448355012,
+      "grad_norm": 2.01025453080292,
+      "learning_rate": 1.1705499256883934e-06,
+      "loss": 0.2587,
+      "step": 11098
+    },
+    {
+      "epoch": 1.6983932670237185,
+      "grad_norm": 2.3308542569292303,
+      "learning_rate": 1.1693867483060328e-06,
+      "loss": 0.2932,
+      "step": 11099
+    },
+    {
+      "epoch": 1.6985462892119356,
+      "grad_norm": 2.234928679181359,
+      "learning_rate": 1.1682241132461791e-06,
+      "loss": 0.2752,
+      "step": 11100
+    },
+    {
+      "epoch": 1.6986993114001532,
+      "grad_norm": 2.4713138884866472,
+      "learning_rate": 1.1670620205802319e-06,
+      "loss": 0.3389,
+      "step": 11101
+    },
+    {
+      "epoch": 1.6988523335883703,
+      "grad_norm": 2.471201205818322,
+      "learning_rate": 1.1659004703795607e-06,
+      "loss": 0.2308,
+      "step": 11102
+    },
+    {
+      "epoch": 1.6990053557765876,
+      "grad_norm": 1.7215349224563574,
+      "learning_rate": 1.1647394627155006e-06,
+      "loss": 0.1932,
+      "step": 11103
+    },
+    {
+      "epoch": 1.699158377964805,
+      "grad_norm": 2.0922403083721925,
+      "learning_rate": 1.1635789976593536e-06,
+      "loss": 0.2619,
+      "step": 11104
+    },
+    {
+      "epoch": 1.699311400153022,
+      "grad_norm": 2.302314103427617,
+      "learning_rate": 1.1624190752823895e-06,
+      "loss": 0.3499,
+      "step": 11105
+    },
+    {
+      "epoch": 1.6994644223412396,
+      "grad_norm": 1.869357598102231,
+      "learning_rate": 1.1612596956558398e-06,
+      "loss": 0.2962,
+      "step": 11106
+    },
+    {
+      "epoch": 1.6996174445294567,
+      "grad_norm": 2.0397788159015935,
+      "learning_rate": 1.1601008588509143e-06,
+      "loss": 0.2743,
+      "step": 11107
+    },
+    {
+      "epoch": 1.699770466717674,
+      "grad_norm": 2.0520554519115093,
+      "learning_rate": 1.1589425649387743e-06,
+      "loss": 0.2592,
+      "step": 11108
+    },
+    {
+      "epoch": 1.6999234889058914,
+      "grad_norm": 1.7852914422848998,
+      "learning_rate": 1.1577848139905568e-06,
+      "loss": 0.2239,
+      "step": 11109
+    },
+    {
+      "epoch": 1.7000765110941085,
+      "grad_norm": 2.014202409879043,
+      "learning_rate": 1.1566276060773674e-06,
+      "loss": 0.2447,
+      "step": 11110
+    },
+    {
+      "epoch": 1.700229533282326,
+      "grad_norm": 1.989579755348068,
+      "learning_rate": 1.1554709412702714e-06,
+      "loss": 0.2221,
+      "step": 11111
+    },
+    {
+      "epoch": 1.7003825554705432,
+      "grad_norm": 2.090257985515126,
+      "learning_rate": 1.154314819640302e-06,
+      "loss": 0.3446,
+      "step": 11112
+    },
+    {
+      "epoch": 1.7005355776587605,
+      "grad_norm": 1.9125910003610571,
+      "learning_rate": 1.153159241258469e-06,
+      "loss": 0.2864,
+      "step": 11113
+    },
+    {
+      "epoch": 1.7006885998469778,
+      "grad_norm": 2.1613839402350314,
+      "learning_rate": 1.1520042061957315e-06,
+      "loss": 0.2825,
+      "step": 11114
+    },
+    {
+      "epoch": 1.700841622035195,
+      "grad_norm": 2.3295656766209616,
+      "learning_rate": 1.1508497145230314e-06,
+      "loss": 0.3438,
+      "step": 11115
+    },
+    {
+      "epoch": 1.7009946442234125,
+      "grad_norm": 2.2090564568144075,
+      "learning_rate": 1.1496957663112697e-06,
+      "loss": 0.3128,
+      "step": 11116
+    },
+    {
+      "epoch": 1.7011476664116296,
+      "grad_norm": 2.0492676006983297,
+      "learning_rate": 1.1485423616313107e-06,
+      "loss": 0.2841,
+      "step": 11117
+    },
+    {
+      "epoch": 1.701300688599847,
+      "grad_norm": 2.1697518217573113,
+      "learning_rate": 1.1473895005539937e-06,
+      "loss": 0.2625,
+      "step": 11118
+    },
+    {
+      "epoch": 1.7014537107880643,
+      "grad_norm": 1.9730217228325013,
+      "learning_rate": 1.1462371831501207e-06,
+      "loss": 0.2775,
+      "step": 11119
+    },
+    {
+      "epoch": 1.7016067329762814,
+      "grad_norm": 2.088602724372186,
+      "learning_rate": 1.1450854094904551e-06,
+      "loss": 0.3239,
+      "step": 11120
+    },
+    {
+      "epoch": 1.701759755164499,
+      "grad_norm": 2.3255899147858443,
+      "learning_rate": 1.143934179645737e-06,
+      "loss": 0.3086,
+      "step": 11121
+    },
+    {
+      "epoch": 1.701912777352716,
+      "grad_norm": 1.9272329964493535,
+      "learning_rate": 1.1427834936866655e-06,
+      "loss": 0.2355,
+      "step": 11122
+    },
+    {
+      "epoch": 1.7020657995409334,
+      "grad_norm": 1.9946843187147796,
+      "learning_rate": 1.1416333516839107e-06,
+      "loss": 0.2835,
+      "step": 11123
+    },
+    {
+      "epoch": 1.7022188217291507,
+      "grad_norm": 2.173321803166349,
+      "learning_rate": 1.1404837537081049e-06,
+      "loss": 0.2785,
+      "step": 11124
+    },
+    {
+      "epoch": 1.702371843917368,
+      "grad_norm": 2.0278180668942745,
+      "learning_rate": 1.1393346998298505e-06,
+      "loss": 0.266,
+      "step": 11125
+    },
+    {
+      "epoch": 1.7025248661055854,
+      "grad_norm": 2.0094825725857173,
+      "learning_rate": 1.1381861901197144e-06,
+      "loss": 0.2798,
+      "step": 11126
+    },
+    {
+      "epoch": 1.7026778882938025,
+      "grad_norm": 2.098979114428069,
+      "learning_rate": 1.1370382246482337e-06,
+      "loss": 0.2667,
+      "step": 11127
+    },
+    {
+      "epoch": 1.7028309104820198,
+      "grad_norm": 2.0523444053923305,
+      "learning_rate": 1.1358908034859062e-06,
+      "loss": 0.3097,
+      "step": 11128
+    },
+    {
+      "epoch": 1.7029839326702372,
+      "grad_norm": 2.1537059294575456,
+      "learning_rate": 1.1347439267032023e-06,
+      "loss": 0.2674,
+      "step": 11129
+    },
+    {
+      "epoch": 1.7031369548584545,
+      "grad_norm": 2.1006487237785727,
+      "learning_rate": 1.1335975943705523e-06,
+      "loss": 0.2938,
+      "step": 11130
+    },
+    {
+      "epoch": 1.7032899770466718,
+      "grad_norm": 2.3639342633413114,
+      "learning_rate": 1.1324518065583633e-06,
+      "loss": 0.2944,
+      "step": 11131
+    },
+    {
+      "epoch": 1.703442999234889,
+      "grad_norm": 2.082932273742042,
+      "learning_rate": 1.1313065633369968e-06,
+      "loss": 0.2364,
+      "step": 11132
+    },
+    {
+      "epoch": 1.7035960214231065,
+      "grad_norm": 2.188856557820004,
+      "learning_rate": 1.1301618647767876e-06,
+      "loss": 0.3076,
+      "step": 11133
+    },
+    {
+      "epoch": 1.7037490436113236,
+      "grad_norm": 2.000951821270597,
+      "learning_rate": 1.1290177109480406e-06,
+      "loss": 0.2625,
+      "step": 11134
+    },
+    {
+      "epoch": 1.703902065799541,
+      "grad_norm": 2.430335169813233,
+      "learning_rate": 1.127874101921018e-06,
+      "loss": 0.3042,
+      "step": 11135
+    },
+    {
+      "epoch": 1.7040550879877583,
+      "grad_norm": 2.281204762892098,
+      "learning_rate": 1.1267310377659524e-06,
+      "loss": 0.3133,
+      "step": 11136
+    },
+    {
+      "epoch": 1.7042081101759754,
+      "grad_norm": 1.9930439380863587,
+      "learning_rate": 1.1255885185530502e-06,
+      "loss": 0.2929,
+      "step": 11137
+    },
+    {
+      "epoch": 1.704361132364193,
+      "grad_norm": 1.9680645229256954,
+      "learning_rate": 1.12444654435247e-06,
+      "loss": 0.2042,
+      "step": 11138
+    },
+    {
+      "epoch": 1.70451415455241,
+      "grad_norm": 2.0093847715801285,
+      "learning_rate": 1.123305115234351e-06,
+      "loss": 0.349,
+      "step": 11139
+    },
+    {
+      "epoch": 1.7046671767406274,
+      "grad_norm": 2.1094826918613556,
+      "learning_rate": 1.122164231268793e-06,
+      "loss": 0.2989,
+      "step": 11140
+    },
+    {
+      "epoch": 1.7048201989288447,
+      "grad_norm": 2.169821861197907,
+      "learning_rate": 1.1210238925258554e-06,
+      "loss": 0.2632,
+      "step": 11141
+    },
+    {
+      "epoch": 1.7049732211170618,
+      "grad_norm": 2.4069773754922466,
+      "learning_rate": 1.1198840990755777e-06,
+      "loss": 0.3368,
+      "step": 11142
+    },
+    {
+      "epoch": 1.7051262433052794,
+      "grad_norm": 2.017085258934159,
+      "learning_rate": 1.1187448509879562e-06,
+      "loss": 0.2816,
+      "step": 11143
+    },
+    {
+      "epoch": 1.7052792654934965,
+      "grad_norm": 1.9728566967575223,
+      "learning_rate": 1.117606148332957e-06,
+      "loss": 0.284,
+      "step": 11144
+    },
+    {
+      "epoch": 1.7054322876817138,
+      "grad_norm": 1.9431671387155085,
+      "learning_rate": 1.116467991180512e-06,
+      "loss": 0.2646,
+      "step": 11145
+    },
+    {
+      "epoch": 1.7055853098699312,
+      "grad_norm": 1.9854827370978707,
+      "learning_rate": 1.115330379600521e-06,
+      "loss": 0.2952,
+      "step": 11146
+    },
+    {
+      "epoch": 1.7057383320581483,
+      "grad_norm": 2.280487685992356,
+      "learning_rate": 1.114193313662848e-06,
+      "loss": 0.2895,
+      "step": 11147
+    },
+    {
+      "epoch": 1.7058913542463658,
+      "grad_norm": 2.287105129323108,
+      "learning_rate": 1.113056793437326e-06,
+      "loss": 0.2921,
+      "step": 11148
+    },
+    {
+      "epoch": 1.706044376434583,
+      "grad_norm": 2.1288322442429384,
+      "learning_rate": 1.1119208189937514e-06,
+      "loss": 0.2851,
+      "step": 11149
+    },
+    {
+      "epoch": 1.7061973986228003,
+      "grad_norm": 2.0158199871640665,
+      "learning_rate": 1.1107853904018896e-06,
+      "loss": 0.2171,
+      "step": 11150
+    },
+    {
+      "epoch": 1.7063504208110176,
+      "grad_norm": 1.9626837031523319,
+      "learning_rate": 1.1096505077314723e-06,
+      "loss": 0.2358,
+      "step": 11151
+    },
+    {
+      "epoch": 1.7065034429992347,
+      "grad_norm": 2.305121265245296,
+      "learning_rate": 1.1085161710521962e-06,
+      "loss": 0.2749,
+      "step": 11152
+    },
+    {
+      "epoch": 1.7066564651874523,
+      "grad_norm": 2.038485482391815,
+      "learning_rate": 1.1073823804337258e-06,
+      "loss": 0.2631,
+      "step": 11153
+    },
+    {
+      "epoch": 1.7068094873756694,
+      "grad_norm": 2.2882425227783862,
+      "learning_rate": 1.1062491359456917e-06,
+      "loss": 0.3128,
+      "step": 11154
+    },
+    {
+      "epoch": 1.7069625095638867,
+      "grad_norm": 2.2012392539786134,
+      "learning_rate": 1.1051164376576896e-06,
+      "loss": 0.276,
+      "step": 11155
+    },
+    {
+      "epoch": 1.707115531752104,
+      "grad_norm": 2.0394408445761663,
+      "learning_rate": 1.1039842856392856e-06,
+      "loss": 0.307,
+      "step": 11156
+    },
+    {
+      "epoch": 1.7072685539403214,
+      "grad_norm": 2.128922991238398,
+      "learning_rate": 1.102852679960007e-06,
+      "loss": 0.2901,
+      "step": 11157
+    },
+    {
+      "epoch": 1.7074215761285387,
+      "grad_norm": 1.8881752359487909,
+      "learning_rate": 1.101721620689352e-06,
+      "loss": 0.298,
+      "step": 11158
+    },
+    {
+      "epoch": 1.7075745983167558,
+      "grad_norm": 2.0841273220620664,
+      "learning_rate": 1.1005911078967802e-06,
+      "loss": 0.2725,
+      "step": 11159
+    },
+    {
+      "epoch": 1.7077276205049732,
+      "grad_norm": 1.994898698810137,
+      "learning_rate": 1.0994611416517264e-06,
+      "loss": 0.2369,
+      "step": 11160
+    },
+    {
+      "epoch": 1.7078806426931905,
+      "grad_norm": 2.0006084196277003,
+      "learning_rate": 1.0983317220235823e-06,
+      "loss": 0.2582,
+      "step": 11161
+    },
+    {
+      "epoch": 1.7080336648814078,
+      "grad_norm": 2.1415344517825736,
+      "learning_rate": 1.0972028490817077e-06,
+      "loss": 0.2784,
+      "step": 11162
+    },
+    {
+      "epoch": 1.7081866870696252,
+      "grad_norm": 1.8469449110302756,
+      "learning_rate": 1.0960745228954384e-06,
+      "loss": 0.2703,
+      "step": 11163
+    },
+    {
+      "epoch": 1.7083397092578423,
+      "grad_norm": 2.0745692309473167,
+      "learning_rate": 1.0949467435340623e-06,
+      "loss": 0.2781,
+      "step": 11164
+    },
+    {
+      "epoch": 1.7084927314460598,
+      "grad_norm": 2.0757462371321473,
+      "learning_rate": 1.0938195110668425e-06,
+      "loss": 0.3172,
+      "step": 11165
+    },
+    {
+      "epoch": 1.708645753634277,
+      "grad_norm": 1.9539949708197906,
+      "learning_rate": 1.092692825563011e-06,
+      "loss": 0.2808,
+      "step": 11166
+    },
+    {
+      "epoch": 1.7087987758224943,
+      "grad_norm": 1.8315528166629997,
+      "learning_rate": 1.091566687091755e-06,
+      "loss": 0.2459,
+      "step": 11167
+    },
+    {
+      "epoch": 1.7089517980107116,
+      "grad_norm": 1.7884907895676587,
+      "learning_rate": 1.0904410957222411e-06,
+      "loss": 0.234,
+      "step": 11168
+    },
+    {
+      "epoch": 1.7091048201989287,
+      "grad_norm": 2.2466456776934662,
+      "learning_rate": 1.0893160515235957e-06,
+      "loss": 0.3361,
+      "step": 11169
+    },
+    {
+      "epoch": 1.7092578423871463,
+      "grad_norm": 2.4207583570260023,
+      "learning_rate": 1.0881915545649058e-06,
+      "loss": 0.3154,
+      "step": 11170
+    },
+    {
+      "epoch": 1.7094108645753634,
+      "grad_norm": 1.9753841454069814,
+      "learning_rate": 1.0870676049152385e-06,
+      "loss": 0.2679,
+      "step": 11171
+    },
+    {
+      "epoch": 1.7095638867635807,
+      "grad_norm": 2.3190131346279514,
+      "learning_rate": 1.0859442026436185e-06,
+      "loss": 0.2965,
+      "step": 11172
+    },
+    {
+      "epoch": 1.709716908951798,
+      "grad_norm": 2.160003228975004,
+      "learning_rate": 1.0848213478190328e-06,
+      "loss": 0.3465,
+      "step": 11173
+    },
+    {
+      "epoch": 1.7098699311400152,
+      "grad_norm": 2.3473105298032264,
+      "learning_rate": 1.0836990405104475e-06,
+      "loss": 0.3369,
+      "step": 11174
+    },
+    {
+      "epoch": 1.7100229533282327,
+      "grad_norm": 2.014596179323918,
+      "learning_rate": 1.082577280786784e-06,
+      "loss": 0.269,
+      "step": 11175
+    },
+    {
+      "epoch": 1.7101759755164498,
+      "grad_norm": 2.057023392344061,
+      "learning_rate": 1.081456068716935e-06,
+      "loss": 0.3078,
+      "step": 11176
+    },
+    {
+      "epoch": 1.7103289977046672,
+      "grad_norm": 2.444349548800692,
+      "learning_rate": 1.0803354043697588e-06,
+      "loss": 0.3089,
+      "step": 11177
+    },
+    {
+      "epoch": 1.7104820198928845,
+      "grad_norm": 1.9832948768618603,
+      "learning_rate": 1.0792152878140794e-06,
+      "loss": 0.3157,
+      "step": 11178
+    },
+    {
+      "epoch": 1.7106350420811016,
+      "grad_norm": 2.1437773295090126,
+      "learning_rate": 1.078095719118687e-06,
+      "loss": 0.3203,
+      "step": 11179
+    },
+    {
+      "epoch": 1.7107880642693192,
+      "grad_norm": 1.9014695196455216,
+      "learning_rate": 1.0769766983523388e-06,
+      "loss": 0.2788,
+      "step": 11180
+    },
+    {
+      "epoch": 1.7109410864575363,
+      "grad_norm": 1.8859102107863197,
+      "learning_rate": 1.0758582255837591e-06,
+      "loss": 0.2665,
+      "step": 11181
+    },
+    {
+      "epoch": 1.7110941086457536,
+      "grad_norm": 2.0492514664079473,
+      "learning_rate": 1.0747403008816382e-06,
+      "loss": 0.2679,
+      "step": 11182
+    },
+    {
+      "epoch": 1.711247130833971,
+      "grad_norm": 2.0658287803286197,
+      "learning_rate": 1.07362292431463e-06,
+      "loss": 0.2491,
+      "step": 11183
+    },
+    {
+      "epoch": 1.711400153022188,
+      "grad_norm": 2.1805943892160475,
+      "learning_rate": 1.0725060959513578e-06,
+      "loss": 0.3342,
+      "step": 11184
+    },
+    {
+      "epoch": 1.7115531752104056,
+      "grad_norm": 1.9814368426589373,
+      "learning_rate": 1.0713898158604119e-06,
+      "loss": 0.2748,
+      "step": 11185
+    },
+    {
+      "epoch": 1.7117061973986227,
+      "grad_norm": 2.0479335820218343,
+      "learning_rate": 1.0702740841103455e-06,
+      "loss": 0.3067,
+      "step": 11186
+    },
+    {
+      "epoch": 1.71185921958684,
+      "grad_norm": 2.2402698668164467,
+      "learning_rate": 1.0691589007696811e-06,
+      "loss": 0.2936,
+      "step": 11187
+    },
+    {
+      "epoch": 1.7120122417750574,
+      "grad_norm": 2.175527755245013,
+      "learning_rate": 1.0680442659069046e-06,
+      "loss": 0.3096,
+      "step": 11188
+    },
+    {
+      "epoch": 1.7121652639632745,
+      "grad_norm": 2.259874022178763,
+      "learning_rate": 1.0669301795904762e-06,
+      "loss": 0.3045,
+      "step": 11189
+    },
+    {
+      "epoch": 1.712318286151492,
+      "grad_norm": 2.206801471311079,
+      "learning_rate": 1.0658166418888094e-06,
+      "loss": 0.2307,
+      "step": 11190
+    },
+    {
+      "epoch": 1.7124713083397092,
+      "grad_norm": 2.0808032911944663,
+      "learning_rate": 1.0647036528702915e-06,
+      "loss": 0.3301,
+      "step": 11191
+    },
+    {
+      "epoch": 1.7126243305279265,
+      "grad_norm": 2.009233915278547,
+      "learning_rate": 1.06359121260328e-06,
+      "loss": 0.2486,
+      "step": 11192
+    },
+    {
+      "epoch": 1.7127773527161438,
+      "grad_norm": 1.9884306466496928,
+      "learning_rate": 1.0624793211560913e-06,
+      "loss": 0.236,
+      "step": 11193
+    },
+    {
+      "epoch": 1.7129303749043612,
+      "grad_norm": 1.9343940879298103,
+      "learning_rate": 1.0613679785970087e-06,
+      "loss": 0.2245,
+      "step": 11194
+    },
+    {
+      "epoch": 1.7130833970925785,
+      "grad_norm": 2.068900757215732,
+      "learning_rate": 1.06025718499429e-06,
+      "loss": 0.336,
+      "step": 11195
+    },
+    {
+      "epoch": 1.7132364192807956,
+      "grad_norm": 2.107983069680888,
+      "learning_rate": 1.059146940416147e-06,
+      "loss": 0.2883,
+      "step": 11196
+    },
+    {
+      "epoch": 1.713389441469013,
+      "grad_norm": 2.0972134931039657,
+      "learning_rate": 1.0580372449307686e-06,
+      "loss": 0.2593,
+      "step": 11197
+    },
+    {
+      "epoch": 1.7135424636572303,
+      "grad_norm": 2.2987713281363344,
+      "learning_rate": 1.0569280986063058e-06,
+      "loss": 0.3548,
+      "step": 11198
+    },
+    {
+      "epoch": 1.7136954858454476,
+      "grad_norm": 2.059416623438375,
+      "learning_rate": 1.0558195015108708e-06,
+      "loss": 0.2374,
+      "step": 11199
+    },
+    {
+      "epoch": 1.713848508033665,
+      "grad_norm": 2.161279175677584,
+      "learning_rate": 1.0547114537125514e-06,
+      "loss": 0.2702,
+      "step": 11200
+    },
+    {
+      "epoch": 1.714001530221882,
+      "grad_norm": 1.9833873999128637,
+      "learning_rate": 1.0536039552793987e-06,
+      "loss": 0.2869,
+      "step": 11201
+    },
+    {
+      "epoch": 1.7141545524100996,
+      "grad_norm": 1.941243048171417,
+      "learning_rate": 1.0524970062794203e-06,
+      "loss": 0.248,
+      "step": 11202
+    },
+    {
+      "epoch": 1.7143075745983167,
+      "grad_norm": 2.215030856608608,
+      "learning_rate": 1.0513906067806067e-06,
+      "loss": 0.2854,
+      "step": 11203
+    },
+    {
+      "epoch": 1.714460596786534,
+      "grad_norm": 2.09891972731733,
+      "learning_rate": 1.0502847568509023e-06,
+      "loss": 0.2748,
+      "step": 11204
+    },
+    {
+      "epoch": 1.7146136189747514,
+      "grad_norm": 1.9231481988683312,
+      "learning_rate": 1.0491794565582225e-06,
+      "loss": 0.227,
+      "step": 11205
+    },
+    {
+      "epoch": 1.7147666411629685,
+      "grad_norm": 2.1008690413719227,
+      "learning_rate": 1.0480747059704488e-06,
+      "loss": 0.2827,
+      "step": 11206
+    },
+    {
+      "epoch": 1.714919663351186,
+      "grad_norm": 1.9290761821050166,
+      "learning_rate": 1.0469705051554269e-06,
+      "loss": 0.3185,
+      "step": 11207
+    },
+    {
+      "epoch": 1.7150726855394032,
+      "grad_norm": 2.343551545185132,
+      "learning_rate": 1.0458668541809714e-06,
+      "loss": 0.3501,
+      "step": 11208
+    },
+    {
+      "epoch": 1.7152257077276205,
+      "grad_norm": 1.788433718216971,
+      "learning_rate": 1.0447637531148592e-06,
+      "loss": 0.2199,
+      "step": 11209
+    },
+    {
+      "epoch": 1.7153787299158378,
+      "grad_norm": 2.0652724541004863,
+      "learning_rate": 1.0436612020248404e-06,
+      "loss": 0.2759,
+      "step": 11210
+    },
+    {
+      "epoch": 1.715531752104055,
+      "grad_norm": 1.9281228885043027,
+      "learning_rate": 1.0425592009786246e-06,
+      "loss": 0.2673,
+      "step": 11211
+    },
+    {
+      "epoch": 1.7156847742922725,
+      "grad_norm": 2.106942568657105,
+      "learning_rate": 1.0414577500438873e-06,
+      "loss": 0.2637,
+      "step": 11212
+    },
+    {
+      "epoch": 1.7158377964804896,
+      "grad_norm": 2.0071831488097227,
+      "learning_rate": 1.0403568492882786e-06,
+      "loss": 0.2315,
+      "step": 11213
+    },
+    {
+      "epoch": 1.715990818668707,
+      "grad_norm": 2.0902800708364024,
+      "learning_rate": 1.039256498779405e-06,
+      "loss": 0.2754,
+      "step": 11214
+    },
+    {
+      "epoch": 1.7161438408569243,
+      "grad_norm": 1.998381374019673,
+      "learning_rate": 1.0381566985848423e-06,
+      "loss": 0.2503,
+      "step": 11215
+    },
+    {
+      "epoch": 1.7162968630451414,
+      "grad_norm": 2.275807287322682,
+      "learning_rate": 1.0370574487721396e-06,
+      "loss": 0.2725,
+      "step": 11216
+    },
+    {
+      "epoch": 1.716449885233359,
+      "grad_norm": 2.4486035736270746,
+      "learning_rate": 1.0359587494087985e-06,
+      "loss": 0.2977,
+      "step": 11217
+    },
+    {
+      "epoch": 1.716602907421576,
+      "grad_norm": 1.9189903077463175,
+      "learning_rate": 1.0348606005622997e-06,
+      "loss": 0.2505,
+      "step": 11218
+    },
+    {
+      "epoch": 1.7167559296097934,
+      "grad_norm": 2.1347410423324025,
+      "learning_rate": 1.0337630023000856e-06,
+      "loss": 0.287,
+      "step": 11219
+    },
+    {
+      "epoch": 1.7169089517980107,
+      "grad_norm": 2.043348875111778,
+      "learning_rate": 1.032665954689558e-06,
+      "loss": 0.2275,
+      "step": 11220
+    },
+    {
+      "epoch": 1.7170619739862278,
+      "grad_norm": 1.9169361145288593,
+      "learning_rate": 1.031569457798095e-06,
+      "loss": 0.2564,
+      "step": 11221
+    },
+    {
+      "epoch": 1.7172149961744454,
+      "grad_norm": 2.4828468009673257,
+      "learning_rate": 1.0304735116930397e-06,
+      "loss": 0.3424,
+      "step": 11222
+    },
+    {
+      "epoch": 1.7173680183626625,
+      "grad_norm": 1.8955725882532888,
+      "learning_rate": 1.0293781164416905e-06,
+      "loss": 0.3574,
+      "step": 11223
+    },
+    {
+      "epoch": 1.7175210405508798,
+      "grad_norm": 2.096578082910678,
+      "learning_rate": 1.0282832721113267e-06,
+      "loss": 0.351,
+      "step": 11224
+    },
+    {
+      "epoch": 1.7176740627390972,
+      "grad_norm": 1.9892607721302973,
+      "learning_rate": 1.0271889787691846e-06,
+      "loss": 0.2568,
+      "step": 11225
+    },
+    {
+      "epoch": 1.7178270849273145,
+      "grad_norm": 2.2149399450508884,
+      "learning_rate": 1.0260952364824694e-06,
+      "loss": 0.2597,
+      "step": 11226
+    },
+    {
+      "epoch": 1.7179801071155318,
+      "grad_norm": 2.198504328031844,
+      "learning_rate": 1.0250020453183506e-06,
+      "loss": 0.2871,
+      "step": 11227
+    },
+    {
+      "epoch": 1.718133129303749,
+      "grad_norm": 1.8657046369191856,
+      "learning_rate": 1.023909405343968e-06,
+      "loss": 0.2746,
+      "step": 11228
+    },
+    {
+      "epoch": 1.7182861514919663,
+      "grad_norm": 2.522595038063424,
+      "learning_rate": 1.0228173166264233e-06,
+      "loss": 0.2711,
+      "step": 11229
+    },
+    {
+      "epoch": 1.7184391736801836,
+      "grad_norm": 2.0394131673625937,
+      "learning_rate": 1.0217257792327862e-06,
+      "loss": 0.3378,
+      "step": 11230
+    },
+    {
+      "epoch": 1.718592195868401,
+      "grad_norm": 2.3529953949487568,
+      "learning_rate": 1.020634793230092e-06,
+      "loss": 0.3383,
+      "step": 11231
+    },
+    {
+      "epoch": 1.7187452180566183,
+      "grad_norm": 2.491294117442281,
+      "learning_rate": 1.0195443586853415e-06,
+      "loss": 0.4019,
+      "step": 11232
+    },
+    {
+      "epoch": 1.7188982402448354,
+      "grad_norm": 1.903192324592772,
+      "learning_rate": 1.0184544756655047e-06,
+      "loss": 0.2835,
+      "step": 11233
+    },
+    {
+      "epoch": 1.719051262433053,
+      "grad_norm": 2.2662638217854467,
+      "learning_rate": 1.0173651442375154e-06,
+      "loss": 0.3071,
+      "step": 11234
+    },
+    {
+      "epoch": 1.71920428462127,
+      "grad_norm": 2.1920337232874973,
+      "learning_rate": 1.0162763644682715e-06,
+      "loss": 0.3151,
+      "step": 11235
+    },
+    {
+      "epoch": 1.7193573068094874,
+      "grad_norm": 1.8404140119518602,
+      "learning_rate": 1.0151881364246407e-06,
+      "loss": 0.2563,
+      "step": 11236
+    },
+    {
+      "epoch": 1.7195103289977047,
+      "grad_norm": 2.371469181227642,
+      "learning_rate": 1.0141004601734562e-06,
+      "loss": 0.3661,
+      "step": 11237
+    },
+    {
+      "epoch": 1.7196633511859218,
+      "grad_norm": 2.676750375304351,
+      "learning_rate": 1.0130133357815142e-06,
+      "loss": 0.3352,
+      "step": 11238
+    },
+    {
+      "epoch": 1.7198163733741394,
+      "grad_norm": 2.2395476317147556,
+      "learning_rate": 1.0119267633155816e-06,
+      "loss": 0.316,
+      "step": 11239
+    },
+    {
+      "epoch": 1.7199693955623565,
+      "grad_norm": 1.9873318256316577,
+      "learning_rate": 1.0108407428423873e-06,
+      "loss": 0.2805,
+      "step": 11240
+    },
+    {
+      "epoch": 1.7201224177505738,
+      "grad_norm": 2.1685902722728874,
+      "learning_rate": 1.0097552744286277e-06,
+      "loss": 0.3208,
+      "step": 11241
+    },
+    {
+      "epoch": 1.7202754399387912,
+      "grad_norm": 2.0648830031892103,
+      "learning_rate": 1.0086703581409696e-06,
+      "loss": 0.2799,
+      "step": 11242
+    },
+    {
+      "epoch": 1.7204284621270083,
+      "grad_norm": 2.168635886474466,
+      "learning_rate": 1.0075859940460374e-06,
+      "loss": 0.2345,
+      "step": 11243
+    },
+    {
+      "epoch": 1.7205814843152258,
+      "grad_norm": 2.061383471866174,
+      "learning_rate": 1.006502182210426e-06,
+      "loss": 0.3173,
+      "step": 11244
+    },
+    {
+      "epoch": 1.720734506503443,
+      "grad_norm": 2.2431859409567343,
+      "learning_rate": 1.0054189227007027e-06,
+      "loss": 0.3387,
+      "step": 11245
+    },
+    {
+      "epoch": 1.7208875286916603,
+      "grad_norm": 1.9175774670047985,
+      "learning_rate": 1.0043362155833857e-06,
+      "loss": 0.2928,
+      "step": 11246
+    },
+    {
+      "epoch": 1.7210405508798776,
+      "grad_norm": 1.7032224043623452,
+      "learning_rate": 1.0032540609249752e-06,
+      "loss": 0.2237,
+      "step": 11247
+    },
+    {
+      "epoch": 1.7211935730680947,
+      "grad_norm": 2.3240615291172344,
+      "learning_rate": 1.0021724587919302e-06,
+      "loss": 0.3035,
+      "step": 11248
+    },
+    {
+      "epoch": 1.7213465952563123,
+      "grad_norm": 2.1714863936300706,
+      "learning_rate": 1.0010914092506706e-06,
+      "loss": 0.3228,
+      "step": 11249
+    },
+    {
+      "epoch": 1.7214996174445294,
+      "grad_norm": 1.8727184181754257,
+      "learning_rate": 1.0000109123675938e-06,
+      "loss": 0.2461,
+      "step": 11250
+    },
+    {
+      "epoch": 1.7216526396327467,
+      "grad_norm": 1.8880016801810797,
+      "learning_rate": 9.989309682090564e-07,
+      "loss": 0.2402,
+      "step": 11251
+    },
+    {
+      "epoch": 1.721805661820964,
+      "grad_norm": 2.0474369307251807,
+      "learning_rate": 9.978515768413766e-07,
+      "loss": 0.2758,
+      "step": 11252
+    },
+    {
+      "epoch": 1.7219586840091812,
+      "grad_norm": 2.183342201937387,
+      "learning_rate": 9.967727383308501e-07,
+      "loss": 0.3161,
+      "step": 11253
+    },
+    {
+      "epoch": 1.7221117061973987,
+      "grad_norm": 2.1222715461235153,
+      "learning_rate": 9.956944527437308e-07,
+      "loss": 0.3003,
+      "step": 11254
+    },
+    {
+      "epoch": 1.7222647283856158,
+      "grad_norm": 1.9626862552546822,
+      "learning_rate": 9.946167201462399e-07,
+      "loss": 0.2209,
+      "step": 11255
+    },
+    {
+      "epoch": 1.7224177505738332,
+      "grad_norm": 2.2158250843339684,
+      "learning_rate": 9.935395406045655e-07,
+      "loss": 0.2603,
+      "step": 11256
+    },
+    {
+      "epoch": 1.7225707727620505,
+      "grad_norm": 2.039833535788292,
+      "learning_rate": 9.924629141848607e-07,
+      "loss": 0.2453,
+      "step": 11257
+    },
+    {
+      "epoch": 1.7227237949502678,
+      "grad_norm": 2.3722448003118073,
+      "learning_rate": 9.913868409532467e-07,
+      "loss": 0.3115,
+      "step": 11258
+    },
+    {
+      "epoch": 1.7228768171384852,
+      "grad_norm": 2.0730045335293257,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.3535,
+      "step": 11259
+    },
+    {
+      "epoch": 1.7230298393267023,
+      "grad_norm": 2.475071828831205,
+      "learning_rate": 9.892363543185989e-07,
+      "loss": 0.3188,
+      "step": 11260
+    },
+    {
+      "epoch": 1.7231828615149196,
+      "grad_norm": 2.053885122377709,
+      "learning_rate": 9.88161941047635e-07,
+      "loss": 0.2961,
+      "step": 11261
+    },
+    {
+      "epoch": 1.723335883703137,
+      "grad_norm": 1.9970096655581508,
+      "learning_rate": 9.870880812289008e-07,
+      "loss": 0.2553,
+      "step": 11262
+    },
+    {
+      "epoch": 1.7234889058913543,
+      "grad_norm": 2.091151342045369,
+      "learning_rate": 9.860147749283455e-07,
+      "loss": 0.3145,
+      "step": 11263
+    },
+    {
+      "epoch": 1.7236419280795716,
+      "grad_norm": 2.3526633638739427,
+      "learning_rate": 9.849420222118866e-07,
+      "loss": 0.3026,
+      "step": 11264
+    },
+    {
+      "epoch": 1.7237949502677887,
+      "grad_norm": 2.2486718187261983,
+      "learning_rate": 9.838698231454048e-07,
+      "loss": 0.3352,
+      "step": 11265
+    },
+    {
+      "epoch": 1.7239479724560063,
+      "grad_norm": 2.134344080040552,
+      "learning_rate": 9.827981777947481e-07,
+      "loss": 0.2745,
+      "step": 11266
+    },
+    {
+      "epoch": 1.7241009946442234,
+      "grad_norm": 2.13120696519989,
+      "learning_rate": 9.817270862257312e-07,
+      "loss": 0.2227,
+      "step": 11267
+    },
+    {
+      "epoch": 1.7242540168324407,
+      "grad_norm": 1.9544916604476297,
+      "learning_rate": 9.80656548504133e-07,
+      "loss": 0.2757,
+      "step": 11268
+    },
+    {
+      "epoch": 1.724407039020658,
+      "grad_norm": 2.305241911383372,
+      "learning_rate": 9.795865646957015e-07,
+      "loss": 0.2935,
+      "step": 11269
+    },
+    {
+      "epoch": 1.7245600612088752,
+      "grad_norm": 2.217856703258211,
+      "learning_rate": 9.785171348661438e-07,
+      "loss": 0.321,
+      "step": 11270
+    },
+    {
+      "epoch": 1.7247130833970927,
+      "grad_norm": 2.2401253166837325,
+      "learning_rate": 9.774482590811451e-07,
+      "loss": 0.2609,
+      "step": 11271
+    },
+    {
+      "epoch": 1.7248661055853098,
+      "grad_norm": 2.379006393238006,
+      "learning_rate": 9.763799374063442e-07,
+      "loss": 0.3083,
+      "step": 11272
+    },
+    {
+      "epoch": 1.7250191277735272,
+      "grad_norm": 2.0967622566022364,
+      "learning_rate": 9.753121699073487e-07,
+      "loss": 0.2765,
+      "step": 11273
+    },
+    {
+      "epoch": 1.7251721499617445,
+      "grad_norm": 2.2026134591533024,
+      "learning_rate": 9.742449566497424e-07,
+      "loss": 0.2821,
+      "step": 11274
+    },
+    {
+      "epoch": 1.7253251721499616,
+      "grad_norm": 2.0261515121698928,
+      "learning_rate": 9.7317829769906e-07,
+      "loss": 0.3014,
+      "step": 11275
+    },
+    {
+      "epoch": 1.7254781943381792,
+      "grad_norm": 2.019153648269614,
+      "learning_rate": 9.7211219312081e-07,
+      "loss": 0.2442,
+      "step": 11276
+    },
+    {
+      "epoch": 1.7256312165263963,
+      "grad_norm": 2.430551797468562,
+      "learning_rate": 9.71046642980471e-07,
+      "loss": 0.3102,
+      "step": 11277
+    },
+    {
+      "epoch": 1.7257842387146136,
+      "grad_norm": 1.950502258473303,
+      "learning_rate": 9.699816473434753e-07,
+      "loss": 0.3363,
+      "step": 11278
+    },
+    {
+      "epoch": 1.725937260902831,
+      "grad_norm": 2.198363605772209,
+      "learning_rate": 9.68917206275234e-07,
+      "loss": 0.3117,
+      "step": 11279
+    },
+    {
+      "epoch": 1.726090283091048,
+      "grad_norm": 1.9226727381010975,
+      "learning_rate": 9.678533198411188e-07,
+      "loss": 0.2341,
+      "step": 11280
+    },
+    {
+      "epoch": 1.7262433052792656,
+      "grad_norm": 2.200988192635039,
+      "learning_rate": 9.667899881064624e-07,
+      "loss": 0.3561,
+      "step": 11281
+    },
+    {
+      "epoch": 1.7263963274674827,
+      "grad_norm": 2.4511019558072196,
+      "learning_rate": 9.657272111365712e-07,
+      "loss": 0.3339,
+      "step": 11282
+    },
+    {
+      "epoch": 1.7265493496557,
+      "grad_norm": 2.649575488293189,
+      "learning_rate": 9.646649889967152e-07,
+      "loss": 0.3311,
+      "step": 11283
+    },
+    {
+      "epoch": 1.7267023718439174,
+      "grad_norm": 2.0152891336023866,
+      "learning_rate": 9.63603321752129e-07,
+      "loss": 0.2774,
+      "step": 11284
+    },
+    {
+      "epoch": 1.7268553940321345,
+      "grad_norm": 2.0731551227423015,
+      "learning_rate": 9.625422094680126e-07,
+      "loss": 0.2537,
+      "step": 11285
+    },
+    {
+      "epoch": 1.727008416220352,
+      "grad_norm": 1.947496310724743,
+      "learning_rate": 9.614816522095339e-07,
+      "loss": 0.2203,
+      "step": 11286
+    },
+    {
+      "epoch": 1.7271614384085692,
+      "grad_norm": 2.3574204738706803,
+      "learning_rate": 9.604216500418262e-07,
+      "loss": 0.2976,
+      "step": 11287
+    },
+    {
+      "epoch": 1.7273144605967865,
+      "grad_norm": 2.014474596823735,
+      "learning_rate": 9.593622030299876e-07,
+      "loss": 0.339,
+      "step": 11288
+    },
+    {
+      "epoch": 1.7274674827850038,
+      "grad_norm": 1.8008681537701139,
+      "learning_rate": 9.583033112390838e-07,
+      "loss": 0.2211,
+      "step": 11289
+    },
+    {
+      "epoch": 1.7276205049732212,
+      "grad_norm": 2.0716788361002685,
+      "learning_rate": 9.57244974734145e-07,
+      "loss": 0.2579,
+      "step": 11290
+    },
+    {
+      "epoch": 1.7277735271614385,
+      "grad_norm": 2.2269332519553537,
+      "learning_rate": 9.561871935801647e-07,
+      "loss": 0.2713,
+      "step": 11291
+    },
+    {
+      "epoch": 1.7279265493496556,
+      "grad_norm": 2.4042579479130257,
+      "learning_rate": 9.551299678421133e-07,
+      "loss": 0.3419,
+      "step": 11292
+    },
+    {
+      "epoch": 1.728079571537873,
+      "grad_norm": 2.0798852529452336,
+      "learning_rate": 9.540732975849122e-07,
+      "loss": 0.2732,
+      "step": 11293
+    },
+    {
+      "epoch": 1.7282325937260903,
+      "grad_norm": 2.345583325761405,
+      "learning_rate": 9.530171828734558e-07,
+      "loss": 0.2879,
+      "step": 11294
+    },
+    {
+      "epoch": 1.7283856159143076,
+      "grad_norm": 2.1856914083684194,
+      "learning_rate": 9.519616237726104e-07,
+      "loss": 0.2799,
+      "step": 11295
+    },
+    {
+      "epoch": 1.728538638102525,
+      "grad_norm": 1.9641914474915316,
+      "learning_rate": 9.509066203471962e-07,
+      "loss": 0.2643,
+      "step": 11296
+    },
+    {
+      "epoch": 1.728691660290742,
+      "grad_norm": 2.2317078996984523,
+      "learning_rate": 9.498521726620036e-07,
+      "loss": 0.3013,
+      "step": 11297
+    },
+    {
+      "epoch": 1.7288446824789594,
+      "grad_norm": 2.0330881449616705,
+      "learning_rate": 9.487982807817975e-07,
+      "loss": 0.2861,
+      "step": 11298
+    },
+    {
+      "epoch": 1.7289977046671767,
+      "grad_norm": 2.0998242046821702,
+      "learning_rate": 9.477449447712938e-07,
+      "loss": 0.3496,
+      "step": 11299
+    },
+    {
+      "epoch": 1.729150726855394,
+      "grad_norm": 2.57819425032008,
+      "learning_rate": 9.466921646951888e-07,
+      "loss": 0.3567,
+      "step": 11300
+    },
+    {
+      "epoch": 1.7293037490436114,
+      "grad_norm": 2.2586641955898257,
+      "learning_rate": 9.456399406181349e-07,
+      "loss": 0.284,
+      "step": 11301
+    },
+    {
+      "epoch": 1.7294567712318285,
+      "grad_norm": 2.4828821297651373,
+      "learning_rate": 9.445882726047507e-07,
+      "loss": 0.2696,
+      "step": 11302
+    },
+    {
+      "epoch": 1.729609793420046,
+      "grad_norm": 1.850980352969131,
+      "learning_rate": 9.435371607196264e-07,
+      "loss": 0.2184,
+      "step": 11303
+    },
+    {
+      "epoch": 1.7297628156082632,
+      "grad_norm": 2.557017556726781,
+      "learning_rate": 9.424866050273163e-07,
+      "loss": 0.2398,
+      "step": 11304
+    },
+    {
+      "epoch": 1.7299158377964805,
+      "grad_norm": 1.9141775522394715,
+      "learning_rate": 9.41436605592333e-07,
+      "loss": 0.2731,
+      "step": 11305
+    },
+    {
+      "epoch": 1.7300688599846978,
+      "grad_norm": 2.0152008196879665,
+      "learning_rate": 9.403871624791672e-07,
+      "loss": 0.2778,
+      "step": 11306
+    },
+    {
+      "epoch": 1.730221882172915,
+      "grad_norm": 2.533871929801867,
+      "learning_rate": 9.393382757522673e-07,
+      "loss": 0.2577,
+      "step": 11307
+    },
+    {
+      "epoch": 1.7303749043611325,
+      "grad_norm": 2.151577434545533,
+      "learning_rate": 9.382899454760497e-07,
+      "loss": 0.2912,
+      "step": 11308
+    },
+    {
+      "epoch": 1.7305279265493496,
+      "grad_norm": 2.002420979004715,
+      "learning_rate": 9.372421717148961e-07,
+      "loss": 0.2804,
+      "step": 11309
+    },
+    {
+      "epoch": 1.730680948737567,
+      "grad_norm": 1.9167255742661933,
+      "learning_rate": 9.361949545331539e-07,
+      "loss": 0.2903,
+      "step": 11310
+    },
+    {
+      "epoch": 1.7308339709257843,
+      "grad_norm": 1.7526154281756776,
+      "learning_rate": 9.351482939951373e-07,
+      "loss": 0.2414,
+      "step": 11311
+    },
+    {
+      "epoch": 1.7309869931140014,
+      "grad_norm": 2.298302566978914,
+      "learning_rate": 9.341021901651259e-07,
+      "loss": 0.2814,
+      "step": 11312
+    },
+    {
+      "epoch": 1.731140015302219,
+      "grad_norm": 2.151736437727825,
+      "learning_rate": 9.330566431073663e-07,
+      "loss": 0.2432,
+      "step": 11313
+    },
+    {
+      "epoch": 1.731293037490436,
+      "grad_norm": 2.0864141985693156,
+      "learning_rate": 9.320116528860667e-07,
+      "loss": 0.2428,
+      "step": 11314
+    },
+    {
+      "epoch": 1.7314460596786534,
+      "grad_norm": 2.0619824325768095,
+      "learning_rate": 9.309672195654063e-07,
+      "loss": 0.2943,
+      "step": 11315
+    },
+    {
+      "epoch": 1.7315990818668707,
+      "grad_norm": 2.142611756371609,
+      "learning_rate": 9.299233432095267e-07,
+      "loss": 0.2184,
+      "step": 11316
+    },
+    {
+      "epoch": 1.7317521040550878,
+      "grad_norm": 2.1460629954558548,
+      "learning_rate": 9.28880023882538e-07,
+      "loss": 0.336,
+      "step": 11317
+    },
+    {
+      "epoch": 1.7319051262433054,
+      "grad_norm": 2.203744055586528,
+      "learning_rate": 9.27837261648512e-07,
+      "loss": 0.2863,
+      "step": 11318
+    },
+    {
+      "epoch": 1.7320581484315225,
+      "grad_norm": 2.2301200311317104,
+      "learning_rate": 9.267950565714911e-07,
+      "loss": 0.3067,
+      "step": 11319
+    },
+    {
+      "epoch": 1.7322111706197398,
+      "grad_norm": 1.977711646870203,
+      "learning_rate": 9.257534087154796e-07,
+      "loss": 0.2776,
+      "step": 11320
+    },
+    {
+      "epoch": 1.7323641928079572,
+      "grad_norm": 2.1372521289802386,
+      "learning_rate": 9.247123181444517e-07,
+      "loss": 0.2968,
+      "step": 11321
+    },
+    {
+      "epoch": 1.7325172149961743,
+      "grad_norm": 2.118149417733798,
+      "learning_rate": 9.23671784922342e-07,
+      "loss": 0.2988,
+      "step": 11322
+    },
+    {
+      "epoch": 1.7326702371843918,
+      "grad_norm": 2.1422567198478872,
+      "learning_rate": 9.226318091130537e-07,
+      "loss": 0.2463,
+      "step": 11323
+    },
+    {
+      "epoch": 1.732823259372609,
+      "grad_norm": 2.0668194086851748,
+      "learning_rate": 9.215923907804591e-07,
+      "loss": 0.3208,
+      "step": 11324
+    },
+    {
+      "epoch": 1.7329762815608263,
+      "grad_norm": 2.5807343354347423,
+      "learning_rate": 9.205535299883894e-07,
+      "loss": 0.2877,
+      "step": 11325
+    },
+    {
+      "epoch": 1.7331293037490436,
+      "grad_norm": 2.09122486977293,
+      "learning_rate": 9.195152268006457e-07,
+      "loss": 0.2624,
+      "step": 11326
+    },
+    {
+      "epoch": 1.733282325937261,
+      "grad_norm": 2.220035139943399,
+      "learning_rate": 9.184774812809972e-07,
+      "loss": 0.2811,
+      "step": 11327
+    },
+    {
+      "epoch": 1.7334353481254783,
+      "grad_norm": 2.311077102527897,
+      "learning_rate": 9.174402934931704e-07,
+      "loss": 0.3409,
+      "step": 11328
+    },
+    {
+      "epoch": 1.7335883703136954,
+      "grad_norm": 1.845809541515683,
+      "learning_rate": 9.164036635008689e-07,
+      "loss": 0.259,
+      "step": 11329
+    },
+    {
+      "epoch": 1.7337413925019127,
+      "grad_norm": 1.9894773477892849,
+      "learning_rate": 9.15367591367754e-07,
+      "loss": 0.2562,
+      "step": 11330
+    },
+    {
+      "epoch": 1.73389441469013,
+      "grad_norm": 2.343597507954915,
+      "learning_rate": 9.143320771574526e-07,
+      "loss": 0.2868,
+      "step": 11331
+    },
+    {
+      "epoch": 1.7340474368783474,
+      "grad_norm": 2.3318031296705426,
+      "learning_rate": 9.132971209335628e-07,
+      "loss": 0.3158,
+      "step": 11332
+    },
+    {
+      "epoch": 1.7342004590665647,
+      "grad_norm": 2.009158194023136,
+      "learning_rate": 9.122627227596458e-07,
+      "loss": 0.2978,
+      "step": 11333
+    },
+    {
+      "epoch": 1.7343534812547818,
+      "grad_norm": 2.0812300521920086,
+      "learning_rate": 9.112288826992233e-07,
+      "loss": 0.3239,
+      "step": 11334
+    },
+    {
+      "epoch": 1.7345065034429994,
+      "grad_norm": 1.9694566051586355,
+      "learning_rate": 9.10195600815792e-07,
+      "loss": 0.302,
+      "step": 11335
+    },
+    {
+      "epoch": 1.7346595256312165,
+      "grad_norm": 2.1698146071646707,
+      "learning_rate": 9.091628771728078e-07,
+      "loss": 0.3485,
+      "step": 11336
+    },
+    {
+      "epoch": 1.7348125478194338,
+      "grad_norm": 2.2078184817883404,
+      "learning_rate": 9.081307118336957e-07,
+      "loss": 0.2642,
+      "step": 11337
+    },
+    {
+      "epoch": 1.7349655700076512,
+      "grad_norm": 2.0498493991285347,
+      "learning_rate": 9.070991048618438e-07,
+      "loss": 0.269,
+      "step": 11338
+    },
+    {
+      "epoch": 1.7351185921958683,
+      "grad_norm": 1.9622688176545964,
+      "learning_rate": 9.060680563206082e-07,
+      "loss": 0.218,
+      "step": 11339
+    },
+    {
+      "epoch": 1.7352716143840858,
+      "grad_norm": 2.0330487439105545,
+      "learning_rate": 9.05037566273308e-07,
+      "loss": 0.2984,
+      "step": 11340
+    },
+    {
+      "epoch": 1.735424636572303,
+      "grad_norm": 2.174098256735964,
+      "learning_rate": 9.040076347832305e-07,
+      "loss": 0.2946,
+      "step": 11341
+    },
+    {
+      "epoch": 1.7355776587605203,
+      "grad_norm": 1.9391762352970459,
+      "learning_rate": 9.029782619136285e-07,
+      "loss": 0.2331,
+      "step": 11342
+    },
+    {
+      "epoch": 1.7357306809487376,
+      "grad_norm": 1.954971032676284,
+      "learning_rate": 9.01949447727719e-07,
+      "loss": 0.2687,
+      "step": 11343
+    },
+    {
+      "epoch": 1.7358837031369547,
+      "grad_norm": 2.3946998205627383,
+      "learning_rate": 9.009211922886862e-07,
+      "loss": 0.3037,
+      "step": 11344
+    },
+    {
+      "epoch": 1.7360367253251723,
+      "grad_norm": 2.4884493360178195,
+      "learning_rate": 8.998934956596772e-07,
+      "loss": 0.3187,
+      "step": 11345
+    },
+    {
+      "epoch": 1.7361897475133894,
+      "grad_norm": 2.0318519644823403,
+      "learning_rate": 8.988663579038104e-07,
+      "loss": 0.2894,
+      "step": 11346
+    },
+    {
+      "epoch": 1.7363427697016067,
+      "grad_norm": 2.032057463388814,
+      "learning_rate": 8.978397790841631e-07,
+      "loss": 0.2795,
+      "step": 11347
+    },
+    {
+      "epoch": 1.736495791889824,
+      "grad_norm": 1.9829708365779646,
+      "learning_rate": 8.968137592637838e-07,
+      "loss": 0.2309,
+      "step": 11348
+    },
+    {
+      "epoch": 1.7366488140780412,
+      "grad_norm": 1.9451190583188658,
+      "learning_rate": 8.95788298505682e-07,
+      "loss": 0.2661,
+      "step": 11349
+    },
+    {
+      "epoch": 1.7368018362662587,
+      "grad_norm": 2.2877897156250455,
+      "learning_rate": 8.947633968728376e-07,
+      "loss": 0.2577,
+      "step": 11350
+    },
+    {
+      "epoch": 1.7369548584544758,
+      "grad_norm": 2.28996390901099,
+      "learning_rate": 8.937390544281932e-07,
+      "loss": 0.2877,
+      "step": 11351
+    },
+    {
+      "epoch": 1.7371078806426932,
+      "grad_norm": 2.1909353599002626,
+      "learning_rate": 8.927152712346542e-07,
+      "loss": 0.2953,
+      "step": 11352
+    },
+    {
+      "epoch": 1.7372609028309105,
+      "grad_norm": 2.0225927534092505,
+      "learning_rate": 8.916920473551039e-07,
+      "loss": 0.2364,
+      "step": 11353
+    },
+    {
+      "epoch": 1.7374139250191276,
+      "grad_norm": 2.195364625764178,
+      "learning_rate": 8.906693828523738e-07,
+      "loss": 0.2801,
+      "step": 11354
+    },
+    {
+      "epoch": 1.7375669472073452,
+      "grad_norm": 2.198796276058231,
+      "learning_rate": 8.896472777892706e-07,
+      "loss": 0.2753,
+      "step": 11355
+    },
+    {
+      "epoch": 1.7377199693955623,
+      "grad_norm": 1.9924419768513708,
+      "learning_rate": 8.886257322285729e-07,
+      "loss": 0.2271,
+      "step": 11356
+    },
+    {
+      "epoch": 1.7378729915837796,
+      "grad_norm": 1.8575121798898055,
+      "learning_rate": 8.876047462330084e-07,
+      "loss": 0.2319,
+      "step": 11357
+    },
+    {
+      "epoch": 1.738026013771997,
+      "grad_norm": 2.0271128466885022,
+      "learning_rate": 8.865843198652858e-07,
+      "loss": 0.279,
+      "step": 11358
+    },
+    {
+      "epoch": 1.7381790359602143,
+      "grad_norm": 2.020993282119618,
+      "learning_rate": 8.855644531880747e-07,
+      "loss": 0.2546,
+      "step": 11359
+    },
+    {
+      "epoch": 1.7383320581484316,
+      "grad_norm": 2.116368172177348,
+      "learning_rate": 8.84545146264002e-07,
+      "loss": 0.3017,
+      "step": 11360
+    },
+    {
+      "epoch": 1.7384850803366487,
+      "grad_norm": 2.1967153978718184,
+      "learning_rate": 8.835263991556742e-07,
+      "loss": 0.2769,
+      "step": 11361
+    },
+    {
+      "epoch": 1.738638102524866,
+      "grad_norm": 1.9077855748756016,
+      "learning_rate": 8.825082119256557e-07,
+      "loss": 0.2808,
+      "step": 11362
+    },
+    {
+      "epoch": 1.7387911247130834,
+      "grad_norm": 1.8821456287630594,
+      "learning_rate": 8.814905846364719e-07,
+      "loss": 0.2419,
+      "step": 11363
+    },
+    {
+      "epoch": 1.7389441469013007,
+      "grad_norm": 2.3135659597892784,
+      "learning_rate": 8.804735173506251e-07,
+      "loss": 0.3392,
+      "step": 11364
+    },
+    {
+      "epoch": 1.739097169089518,
+      "grad_norm": 1.8049680049774777,
+      "learning_rate": 8.794570101305744e-07,
+      "loss": 0.2129,
+      "step": 11365
+    },
+    {
+      "epoch": 1.7392501912777352,
+      "grad_norm": 2.355594127900009,
+      "learning_rate": 8.784410630387497e-07,
+      "loss": 0.2976,
+      "step": 11366
+    },
+    {
+      "epoch": 1.7394032134659527,
+      "grad_norm": 1.8566497792068795,
+      "learning_rate": 8.774256761375432e-07,
+      "loss": 0.2239,
+      "step": 11367
+    },
+    {
+      "epoch": 1.7395562356541698,
+      "grad_norm": 1.8551089045311278,
+      "learning_rate": 8.764108494893131e-07,
+      "loss": 0.222,
+      "step": 11368
+    },
+    {
+      "epoch": 1.7397092578423872,
+      "grad_norm": 2.11851018960962,
+      "learning_rate": 8.753965831563838e-07,
+      "loss": 0.2562,
+      "step": 11369
+    },
+    {
+      "epoch": 1.7398622800306045,
+      "grad_norm": 2.2494424386365672,
+      "learning_rate": 8.743828772010465e-07,
+      "loss": 0.3498,
+      "step": 11370
+    },
+    {
+      "epoch": 1.7400153022188216,
+      "grad_norm": 2.0475301438582627,
+      "learning_rate": 8.733697316855572e-07,
+      "loss": 0.314,
+      "step": 11371
+    },
+    {
+      "epoch": 1.7401683244070392,
+      "grad_norm": 2.346917348304334,
+      "learning_rate": 8.723571466721348e-07,
+      "loss": 0.2817,
+      "step": 11372
+    },
+    {
+      "epoch": 1.7403213465952563,
+      "grad_norm": 1.9556503799353762,
+      "learning_rate": 8.713451222229674e-07,
+      "loss": 0.2514,
+      "step": 11373
+    },
+    {
+      "epoch": 1.7404743687834736,
+      "grad_norm": 1.909780103481461,
+      "learning_rate": 8.703336584002098e-07,
+      "loss": 0.2623,
+      "step": 11374
+    },
+    {
+      "epoch": 1.740627390971691,
+      "grad_norm": 2.2751671168907084,
+      "learning_rate": 8.693227552659766e-07,
+      "loss": 0.3077,
+      "step": 11375
+    },
+    {
+      "epoch": 1.740780413159908,
+      "grad_norm": 2.295512749361854,
+      "learning_rate": 8.683124128823506e-07,
+      "loss": 0.3109,
+      "step": 11376
+    },
+    {
+      "epoch": 1.7409334353481256,
+      "grad_norm": 2.070742958849601,
+      "learning_rate": 8.673026313113853e-07,
+      "loss": 0.2412,
+      "step": 11377
+    },
+    {
+      "epoch": 1.7410864575363427,
+      "grad_norm": 2.107583228263569,
+      "learning_rate": 8.662934106150922e-07,
+      "loss": 0.2787,
+      "step": 11378
+    },
+    {
+      "epoch": 1.74123947972456,
+      "grad_norm": 2.361860363040369,
+      "learning_rate": 8.652847508554497e-07,
+      "loss": 0.3423,
+      "step": 11379
+    },
+    {
+      "epoch": 1.7413925019127774,
+      "grad_norm": 2.0767444703186912,
+      "learning_rate": 8.642766520944091e-07,
+      "loss": 0.3141,
+      "step": 11380
+    },
+    {
+      "epoch": 1.7415455241009945,
+      "grad_norm": 2.336321547404426,
+      "learning_rate": 8.632691143938754e-07,
+      "loss": 0.3341,
+      "step": 11381
+    },
+    {
+      "epoch": 1.741698546289212,
+      "grad_norm": 2.0940993397565943,
+      "learning_rate": 8.622621378157292e-07,
+      "loss": 0.2899,
+      "step": 11382
+    },
+    {
+      "epoch": 1.7418515684774292,
+      "grad_norm": 2.3391721829229533,
+      "learning_rate": 8.612557224218154e-07,
+      "loss": 0.3682,
+      "step": 11383
+    },
+    {
+      "epoch": 1.7420045906656465,
+      "grad_norm": 2.4228550269184,
+      "learning_rate": 8.602498682739347e-07,
+      "loss": 0.2575,
+      "step": 11384
+    },
+    {
+      "epoch": 1.7421576128538638,
+      "grad_norm": 2.0087616832226405,
+      "learning_rate": 8.592445754338652e-07,
+      "loss": 0.2201,
+      "step": 11385
+    },
+    {
+      "epoch": 1.742310635042081,
+      "grad_norm": 2.1070525512343026,
+      "learning_rate": 8.582398439633466e-07,
+      "loss": 0.2497,
+      "step": 11386
+    },
+    {
+      "epoch": 1.7424636572302985,
+      "grad_norm": 2.08953895418418,
+      "learning_rate": 8.572356739240817e-07,
+      "loss": 0.2649,
+      "step": 11387
+    },
+    {
+      "epoch": 1.7426166794185156,
+      "grad_norm": 2.144426001085784,
+      "learning_rate": 8.562320653777401e-07,
+      "loss": 0.2698,
+      "step": 11388
+    },
+    {
+      "epoch": 1.742769701606733,
+      "grad_norm": 2.063685815560944,
+      "learning_rate": 8.55229018385958e-07,
+      "loss": 0.2662,
+      "step": 11389
+    },
+    {
+      "epoch": 1.7429227237949503,
+      "grad_norm": 2.0046930646619363,
+      "learning_rate": 8.542265330103372e-07,
+      "loss": 0.217,
+      "step": 11390
+    },
+    {
+      "epoch": 1.7430757459831676,
+      "grad_norm": 2.0568254195121796,
+      "learning_rate": 8.53224609312443e-07,
+      "loss": 0.2466,
+      "step": 11391
+    },
+    {
+      "epoch": 1.743228768171385,
+      "grad_norm": 2.342779252411613,
+      "learning_rate": 8.522232473538073e-07,
+      "loss": 0.3322,
+      "step": 11392
+    },
+    {
+      "epoch": 1.743381790359602,
+      "grad_norm": 2.245448570805698,
+      "learning_rate": 8.512224471959296e-07,
+      "loss": 0.2831,
+      "step": 11393
+    },
+    {
+      "epoch": 1.7435348125478194,
+      "grad_norm": 2.3014573795661146,
+      "learning_rate": 8.50222208900271e-07,
+      "loss": 0.2902,
+      "step": 11394
+    },
+    {
+      "epoch": 1.7436878347360367,
+      "grad_norm": 1.9968036086457233,
+      "learning_rate": 8.492225325282599e-07,
+      "loss": 0.2555,
+      "step": 11395
+    },
+    {
+      "epoch": 1.743840856924254,
+      "grad_norm": 1.9143267854598263,
+      "learning_rate": 8.482234181412918e-07,
+      "loss": 0.2421,
+      "step": 11396
+    },
+    {
+      "epoch": 1.7439938791124714,
+      "grad_norm": 1.7719945723268122,
+      "learning_rate": 8.472248658007253e-07,
+      "loss": 0.2534,
+      "step": 11397
+    },
+    {
+      "epoch": 1.7441469013006885,
+      "grad_norm": 2.533973732151635,
+      "learning_rate": 8.46226875567886e-07,
+      "loss": 0.3271,
+      "step": 11398
+    },
+    {
+      "epoch": 1.744299923488906,
+      "grad_norm": 2.2236967026045082,
+      "learning_rate": 8.452294475040645e-07,
+      "loss": 0.3054,
+      "step": 11399
+    },
+    {
+      "epoch": 1.7444529456771232,
+      "grad_norm": 2.1544617140659237,
+      "learning_rate": 8.442325816705154e-07,
+      "loss": 0.2555,
+      "step": 11400
+    },
+    {
+      "epoch": 1.7446059678653405,
+      "grad_norm": 2.3671783149900247,
+      "learning_rate": 8.43236278128462e-07,
+      "loss": 0.3112,
+      "step": 11401
+    },
+    {
+      "epoch": 1.7447589900535578,
+      "grad_norm": 2.125329113033378,
+      "learning_rate": 8.422405369390874e-07,
+      "loss": 0.2517,
+      "step": 11402
+    },
+    {
+      "epoch": 1.744912012241775,
+      "grad_norm": 2.374595888527881,
+      "learning_rate": 8.412453581635505e-07,
+      "loss": 0.3258,
+      "step": 11403
+    },
+    {
+      "epoch": 1.7450650344299925,
+      "grad_norm": 2.1344730686834503,
+      "learning_rate": 8.402507418629646e-07,
+      "loss": 0.3029,
+      "step": 11404
+    },
+    {
+      "epoch": 1.7452180566182096,
+      "grad_norm": 2.2387729700388648,
+      "learning_rate": 8.39256688098411e-07,
+      "loss": 0.2967,
+      "step": 11405
+    },
+    {
+      "epoch": 1.745371078806427,
+      "grad_norm": 1.959205870616328,
+      "learning_rate": 8.382631969309451e-07,
+      "loss": 0.2543,
+      "step": 11406
+    },
+    {
+      "epoch": 1.7455241009946443,
+      "grad_norm": 2.007488255339045,
+      "learning_rate": 8.37270268421575e-07,
+      "loss": 0.2798,
+      "step": 11407
+    },
+    {
+      "epoch": 1.7456771231828614,
+      "grad_norm": 2.152705889958382,
+      "learning_rate": 8.362779026312817e-07,
+      "loss": 0.2928,
+      "step": 11408
+    },
+    {
+      "epoch": 1.745830145371079,
+      "grad_norm": 2.005074106876521,
+      "learning_rate": 8.352860996210133e-07,
+      "loss": 0.2823,
+      "step": 11409
+    },
+    {
+      "epoch": 1.745983167559296,
+      "grad_norm": 2.692710912856644,
+      "learning_rate": 8.342948594516753e-07,
+      "loss": 0.3304,
+      "step": 11410
+    },
+    {
+      "epoch": 1.7461361897475134,
+      "grad_norm": 2.5533912783203205,
+      "learning_rate": 8.333041821841481e-07,
+      "loss": 0.3511,
+      "step": 11411
+    },
+    {
+      "epoch": 1.7462892119357307,
+      "grad_norm": 2.031777309784537,
+      "learning_rate": 8.32314067879274e-07,
+      "loss": 0.2852,
+      "step": 11412
+    },
+    {
+      "epoch": 1.7464422341239478,
+      "grad_norm": 1.735112109022947,
+      "learning_rate": 8.313245165978522e-07,
+      "loss": 0.2149,
+      "step": 11413
+    },
+    {
+      "epoch": 1.7465952563121654,
+      "grad_norm": 1.7149314804299296,
+      "learning_rate": 8.303355284006631e-07,
+      "loss": 0.1944,
+      "step": 11414
+    },
+    {
+      "epoch": 1.7467482785003825,
+      "grad_norm": 2.34769130403817,
+      "learning_rate": 8.293471033484412e-07,
+      "loss": 0.2967,
+      "step": 11415
+    },
+    {
+      "epoch": 1.7469013006885998,
+      "grad_norm": 2.104585496737917,
+      "learning_rate": 8.283592415018871e-07,
+      "loss": 0.3018,
+      "step": 11416
+    },
+    {
+      "epoch": 1.7470543228768172,
+      "grad_norm": 2.1709189972513117,
+      "learning_rate": 8.273719429216732e-07,
+      "loss": 0.3429,
+      "step": 11417
+    },
+    {
+      "epoch": 1.7472073450650343,
+      "grad_norm": 2.033005114625582,
+      "learning_rate": 8.26385207668432e-07,
+      "loss": 0.2769,
+      "step": 11418
+    },
+    {
+      "epoch": 1.7473603672532518,
+      "grad_norm": 2.200352392592713,
+      "learning_rate": 8.25399035802763e-07,
+      "loss": 0.3022,
+      "step": 11419
+    },
+    {
+      "epoch": 1.747513389441469,
+      "grad_norm": 2.016402621506828,
+      "learning_rate": 8.244134273852289e-07,
+      "loss": 0.3039,
+      "step": 11420
+    },
+    {
+      "epoch": 1.7476664116296863,
+      "grad_norm": 1.9838466411531992,
+      "learning_rate": 8.234283824763623e-07,
+      "loss": 0.2549,
+      "step": 11421
+    },
+    {
+      "epoch": 1.7478194338179036,
+      "grad_norm": 1.9775094872488335,
+      "learning_rate": 8.22443901136658e-07,
+      "loss": 0.2676,
+      "step": 11422
+    },
+    {
+      "epoch": 1.7479724560061207,
+      "grad_norm": 1.9390339024939802,
+      "learning_rate": 8.214599834265757e-07,
+      "loss": 0.2548,
+      "step": 11423
+    },
+    {
+      "epoch": 1.7481254781943383,
+      "grad_norm": 2.2990351085347966,
+      "learning_rate": 8.204766294065436e-07,
+      "loss": 0.2939,
+      "step": 11424
+    },
+    {
+      "epoch": 1.7482785003825554,
+      "grad_norm": 2.1371254621600824,
+      "learning_rate": 8.194938391369511e-07,
+      "loss": 0.2853,
+      "step": 11425
+    },
+    {
+      "epoch": 1.7484315225707727,
+      "grad_norm": 1.8783787360270152,
+      "learning_rate": 8.185116126781567e-07,
+      "loss": 0.2473,
+      "step": 11426
+    },
+    {
+      "epoch": 1.74858454475899,
+      "grad_norm": 1.946085021023166,
+      "learning_rate": 8.175299500904821e-07,
+      "loss": 0.2571,
+      "step": 11427
+    },
+    {
+      "epoch": 1.7487375669472074,
+      "grad_norm": 2.2973323888353607,
+      "learning_rate": 8.165488514342157e-07,
+      "loss": 0.3026,
+      "step": 11428
+    },
+    {
+      "epoch": 1.7488905891354247,
+      "grad_norm": 2.2079906556541835,
+      "learning_rate": 8.155683167696105e-07,
+      "loss": 0.3072,
+      "step": 11429
+    },
+    {
+      "epoch": 1.7490436113236418,
+      "grad_norm": 2.1061825703626034,
+      "learning_rate": 8.145883461568837e-07,
+      "loss": 0.287,
+      "step": 11430
+    },
+    {
+      "epoch": 1.7491966335118592,
+      "grad_norm": 2.4976507044518965,
+      "learning_rate": 8.136089396562186e-07,
+      "loss": 0.3586,
+      "step": 11431
+    },
+    {
+      "epoch": 1.7493496557000765,
+      "grad_norm": 2.1563888177726134,
+      "learning_rate": 8.126300973277701e-07,
+      "loss": 0.2401,
+      "step": 11432
+    },
+    {
+      "epoch": 1.7495026778882938,
+      "grad_norm": 2.4782391257527436,
+      "learning_rate": 8.116518192316459e-07,
+      "loss": 0.3337,
+      "step": 11433
+    },
+    {
+      "epoch": 1.7496557000765112,
+      "grad_norm": 2.0657477292079895,
+      "learning_rate": 8.106741054279277e-07,
+      "loss": 0.2599,
+      "step": 11434
+    },
+    {
+      "epoch": 1.7498087222647283,
+      "grad_norm": 2.4612207172641676,
+      "learning_rate": 8.096969559766643e-07,
+      "loss": 0.2549,
+      "step": 11435
+    },
+    {
+      "epoch": 1.7499617444529458,
+      "grad_norm": 2.118689141461657,
+      "learning_rate": 8.087203709378622e-07,
+      "loss": 0.2663,
+      "step": 11436
+    },
+    {
+      "epoch": 1.750114766641163,
+      "grad_norm": 2.19481740548431,
+      "learning_rate": 8.077443503714965e-07,
+      "loss": 0.3231,
+      "step": 11437
+    },
+    {
+      "epoch": 1.7502677888293803,
+      "grad_norm": 2.345923801834366,
+      "learning_rate": 8.06768894337514e-07,
+      "loss": 0.3122,
+      "step": 11438
+    },
+    {
+      "epoch": 1.7504208110175976,
+      "grad_norm": 2.1517185673359727,
+      "learning_rate": 8.057940028958145e-07,
+      "loss": 0.271,
+      "step": 11439
+    },
+    {
+      "epoch": 1.7505738332058147,
+      "grad_norm": 2.047360134838788,
+      "learning_rate": 8.048196761062743e-07,
+      "loss": 0.2962,
+      "step": 11440
+    },
+    {
+      "epoch": 1.7507268553940323,
+      "grad_norm": 2.166641555801906,
+      "learning_rate": 8.038459140287325e-07,
+      "loss": 0.3112,
+      "step": 11441
+    },
+    {
+      "epoch": 1.7508798775822494,
+      "grad_norm": 2.050956415007141,
+      "learning_rate": 8.028727167229844e-07,
+      "loss": 0.2088,
+      "step": 11442
+    },
+    {
+      "epoch": 1.7510328997704667,
+      "grad_norm": 1.7371176140968658,
+      "learning_rate": 8.019000842488045e-07,
+      "loss": 0.2038,
+      "step": 11443
+    },
+    {
+      "epoch": 1.751185921958684,
+      "grad_norm": 2.948667707658099,
+      "learning_rate": 8.00928016665925e-07,
+      "loss": 0.2795,
+      "step": 11444
+    },
+    {
+      "epoch": 1.7513389441469012,
+      "grad_norm": 1.9595754810553292,
+      "learning_rate": 7.999565140340393e-07,
+      "loss": 0.3042,
+      "step": 11445
+    },
+    {
+      "epoch": 1.7514919663351187,
+      "grad_norm": 1.9510168138436947,
+      "learning_rate": 7.989855764128163e-07,
+      "loss": 0.2728,
+      "step": 11446
+    },
+    {
+      "epoch": 1.7516449885233358,
+      "grad_norm": 2.25768939840928,
+      "learning_rate": 7.980152038618838e-07,
+      "loss": 0.2869,
+      "step": 11447
+    },
+    {
+      "epoch": 1.7517980107115532,
+      "grad_norm": 2.322938772528777,
+      "learning_rate": 7.970453964408365e-07,
+      "loss": 0.3438,
+      "step": 11448
+    },
+    {
+      "epoch": 1.7519510328997705,
+      "grad_norm": 2.1999953646735477,
+      "learning_rate": 7.960761542092332e-07,
+      "loss": 0.3005,
+      "step": 11449
+    },
+    {
+      "epoch": 1.7521040550879876,
+      "grad_norm": 1.7289761044406273,
+      "learning_rate": 7.951074772265987e-07,
+      "loss": 0.222,
+      "step": 11450
+    },
+    {
+      "epoch": 1.7522570772762052,
+      "grad_norm": 2.190680943604739,
+      "learning_rate": 7.941393655524243e-07,
+      "loss": 0.3138,
+      "step": 11451
+    },
+    {
+      "epoch": 1.7524100994644223,
+      "grad_norm": 2.425449360415477,
+      "learning_rate": 7.931718192461657e-07,
+      "loss": 0.3134,
+      "step": 11452
+    },
+    {
+      "epoch": 1.7525631216526396,
+      "grad_norm": 1.8979134354611327,
+      "learning_rate": 7.92204838367242e-07,
+      "loss": 0.3067,
+      "step": 11453
+    },
+    {
+      "epoch": 1.752716143840857,
+      "grad_norm": 1.9714110258957196,
+      "learning_rate": 7.912384229750402e-07,
+      "loss": 0.2503,
+      "step": 11454
+    },
+    {
+      "epoch": 1.752869166029074,
+      "grad_norm": 2.2808248096299493,
+      "learning_rate": 7.902725731289107e-07,
+      "loss": 0.2858,
+      "step": 11455
+    },
+    {
+      "epoch": 1.7530221882172916,
+      "grad_norm": 1.7634975671903566,
+      "learning_rate": 7.893072888881748e-07,
+      "loss": 0.2124,
+      "step": 11456
+    },
+    {
+      "epoch": 1.7531752104055087,
+      "grad_norm": 1.990459773666613,
+      "learning_rate": 7.883425703121083e-07,
+      "loss": 0.2959,
+      "step": 11457
+    },
+    {
+      "epoch": 1.753328232593726,
+      "grad_norm": 1.952214898201334,
+      "learning_rate": 7.873784174599586e-07,
+      "loss": 0.2631,
+      "step": 11458
+    },
+    {
+      "epoch": 1.7534812547819434,
+      "grad_norm": 2.8143096353414445,
+      "learning_rate": 7.864148303909447e-07,
+      "loss": 0.3382,
+      "step": 11459
+    },
+    {
+      "epoch": 1.7536342769701607,
+      "grad_norm": 2.1359237551802837,
+      "learning_rate": 7.85451809164236e-07,
+      "loss": 0.233,
+      "step": 11460
+    },
+    {
+      "epoch": 1.753787299158378,
+      "grad_norm": 2.358983205654095,
+      "learning_rate": 7.844893538389808e-07,
+      "loss": 0.2735,
+      "step": 11461
+    },
+    {
+      "epoch": 1.7539403213465952,
+      "grad_norm": 2.186624182602075,
+      "learning_rate": 7.835274644742874e-07,
+      "loss": 0.2891,
+      "step": 11462
+    },
+    {
+      "epoch": 1.7540933435348125,
+      "grad_norm": 2.1643668630024058,
+      "learning_rate": 7.82566141129224e-07,
+      "loss": 0.3178,
+      "step": 11463
+    },
+    {
+      "epoch": 1.7542463657230298,
+      "grad_norm": 2.391814174037397,
+      "learning_rate": 7.816053838628346e-07,
+      "loss": 0.3334,
+      "step": 11464
+    },
+    {
+      "epoch": 1.7543993879112472,
+      "grad_norm": 2.127914629053998,
+      "learning_rate": 7.80645192734123e-07,
+      "loss": 0.3049,
+      "step": 11465
+    },
+    {
+      "epoch": 1.7545524100994645,
+      "grad_norm": 2.3015822780937505,
+      "learning_rate": 7.796855678020531e-07,
+      "loss": 0.3718,
+      "step": 11466
+    },
+    {
+      "epoch": 1.7547054322876816,
+      "grad_norm": 2.1525976489722987,
+      "learning_rate": 7.787265091255636e-07,
+      "loss": 0.2538,
+      "step": 11467
+    },
+    {
+      "epoch": 1.7548584544758992,
+      "grad_norm": 2.2044691852542035,
+      "learning_rate": 7.777680167635537e-07,
+      "loss": 0.2651,
+      "step": 11468
+    },
+    {
+      "epoch": 1.7550114766641163,
+      "grad_norm": 2.046414617913969,
+      "learning_rate": 7.768100907748866e-07,
+      "loss": 0.2679,
+      "step": 11469
+    },
+    {
+      "epoch": 1.7551644988523336,
+      "grad_norm": 2.2233483966772716,
+      "learning_rate": 7.758527312183939e-07,
+      "loss": 0.3119,
+      "step": 11470
+    },
+    {
+      "epoch": 1.755317521040551,
+      "grad_norm": 2.0313291587701934,
+      "learning_rate": 7.748959381528698e-07,
+      "loss": 0.2891,
+      "step": 11471
+    },
+    {
+      "epoch": 1.755470543228768,
+      "grad_norm": 2.393618769392913,
+      "learning_rate": 7.73939711637075e-07,
+      "loss": 0.3505,
+      "step": 11472
+    },
+    {
+      "epoch": 1.7556235654169856,
+      "grad_norm": 2.128114574717289,
+      "learning_rate": 7.729840517297349e-07,
+      "loss": 0.3469,
+      "step": 11473
+    },
+    {
+      "epoch": 1.7557765876052027,
+      "grad_norm": 2.2834270947072324,
+      "learning_rate": 7.720289584895413e-07,
+      "loss": 0.297,
+      "step": 11474
+    },
+    {
+      "epoch": 1.75592960979342,
+      "grad_norm": 2.3363821368142124,
+      "learning_rate": 7.710744319751484e-07,
+      "loss": 0.3029,
+      "step": 11475
+    },
+    {
+      "epoch": 1.7560826319816374,
+      "grad_norm": 2.12918511038581,
+      "learning_rate": 7.701204722451783e-07,
+      "loss": 0.2405,
+      "step": 11476
+    },
+    {
+      "epoch": 1.7562356541698545,
+      "grad_norm": 2.187618497157304,
+      "learning_rate": 7.691670793582185e-07,
+      "loss": 0.2519,
+      "step": 11477
+    },
+    {
+      "epoch": 1.756388676358072,
+      "grad_norm": 2.3802505932165023,
+      "learning_rate": 7.682142533728187e-07,
+      "loss": 0.3085,
+      "step": 11478
+    },
+    {
+      "epoch": 1.7565416985462892,
+      "grad_norm": 2.090745528324452,
+      "learning_rate": 7.672619943474968e-07,
+      "loss": 0.2837,
+      "step": 11479
+    },
+    {
+      "epoch": 1.7566947207345065,
+      "grad_norm": 2.149798905042403,
+      "learning_rate": 7.663103023407337e-07,
+      "loss": 0.269,
+      "step": 11480
+    },
+    {
+      "epoch": 1.7568477429227238,
+      "grad_norm": 2.8280663649694042,
+      "learning_rate": 7.653591774109781e-07,
+      "loss": 0.3053,
+      "step": 11481
+    },
+    {
+      "epoch": 1.757000765110941,
+      "grad_norm": 2.0802449500599085,
+      "learning_rate": 7.644086196166423e-07,
+      "loss": 0.3047,
+      "step": 11482
+    },
+    {
+      "epoch": 1.7571537872991585,
+      "grad_norm": 1.946712250652231,
+      "learning_rate": 7.634586290161017e-07,
+      "loss": 0.2859,
+      "step": 11483
+    },
+    {
+      "epoch": 1.7573068094873756,
+      "grad_norm": 2.1479427054523903,
+      "learning_rate": 7.625092056676997e-07,
+      "loss": 0.2553,
+      "step": 11484
+    },
+    {
+      "epoch": 1.757459831675593,
+      "grad_norm": 2.0637537859831996,
+      "learning_rate": 7.615603496297475e-07,
+      "loss": 0.2663,
+      "step": 11485
+    },
+    {
+      "epoch": 1.7576128538638103,
+      "grad_norm": 2.345797647798903,
+      "learning_rate": 7.606120609605128e-07,
+      "loss": 0.2572,
+      "step": 11486
+    },
+    {
+      "epoch": 1.7577658760520274,
+      "grad_norm": 2.0232236730906106,
+      "learning_rate": 7.596643397182357e-07,
+      "loss": 0.2539,
+      "step": 11487
+    },
+    {
+      "epoch": 1.757918898240245,
+      "grad_norm": 2.1486066998428193,
+      "learning_rate": 7.587171859611219e-07,
+      "loss": 0.333,
+      "step": 11488
+    },
+    {
+      "epoch": 1.758071920428462,
+      "grad_norm": 1.887984788614984,
+      "learning_rate": 7.577705997473372e-07,
+      "loss": 0.2489,
+      "step": 11489
+    },
+    {
+      "epoch": 1.7582249426166794,
+      "grad_norm": 1.8155595847795774,
+      "learning_rate": 7.568245811350139e-07,
+      "loss": 0.2399,
+      "step": 11490
+    },
+    {
+      "epoch": 1.7583779648048967,
+      "grad_norm": 2.2819860715358384,
+      "learning_rate": 7.558791301822555e-07,
+      "loss": 0.3071,
+      "step": 11491
+    },
+    {
+      "epoch": 1.758530986993114,
+      "grad_norm": 2.005584975186744,
+      "learning_rate": 7.549342469471199e-07,
+      "loss": 0.2336,
+      "step": 11492
+    },
+    {
+      "epoch": 1.7586840091813314,
+      "grad_norm": 2.289801482583614,
+      "learning_rate": 7.539899314876409e-07,
+      "loss": 0.3048,
+      "step": 11493
+    },
+    {
+      "epoch": 1.7588370313695485,
+      "grad_norm": 2.121216076992854,
+      "learning_rate": 7.530461838618119e-07,
+      "loss": 0.3218,
+      "step": 11494
+    },
+    {
+      "epoch": 1.7589900535577658,
+      "grad_norm": 2.236459890017379,
+      "learning_rate": 7.521030041275879e-07,
+      "loss": 0.267,
+      "step": 11495
+    },
+    {
+      "epoch": 1.7591430757459832,
+      "grad_norm": 2.2002981567336466,
+      "learning_rate": 7.51160392342899e-07,
+      "loss": 0.2827,
+      "step": 11496
+    },
+    {
+      "epoch": 1.7592960979342005,
+      "grad_norm": 2.0246238778322327,
+      "learning_rate": 7.502183485656311e-07,
+      "loss": 0.2729,
+      "step": 11497
+    },
+    {
+      "epoch": 1.7594491201224178,
+      "grad_norm": 2.2739189494236993,
+      "learning_rate": 7.492768728536404e-07,
+      "loss": 0.2925,
+      "step": 11498
+    },
+    {
+      "epoch": 1.759602142310635,
+      "grad_norm": 1.9264721755974386,
+      "learning_rate": 7.483359652647448e-07,
+      "loss": 0.2605,
+      "step": 11499
+    },
+    {
+      "epoch": 1.7597551644988525,
+      "grad_norm": 2.0289411572643523,
+      "learning_rate": 7.473956258567316e-07,
+      "loss": 0.2722,
+      "step": 11500
+    },
+    {
+      "epoch": 1.7599081866870696,
+      "grad_norm": 1.6723850393620163,
+      "learning_rate": 7.46455854687349e-07,
+      "loss": 0.1741,
+      "step": 11501
+    },
+    {
+      "epoch": 1.760061208875287,
+      "grad_norm": 2.3182455245711857,
+      "learning_rate": 7.45516651814312e-07,
+      "loss": 0.2993,
+      "step": 11502
+    },
+    {
+      "epoch": 1.7602142310635043,
+      "grad_norm": 2.6195480262424002,
+      "learning_rate": 7.44578017295301e-07,
+      "loss": 0.3212,
+      "step": 11503
+    },
+    {
+      "epoch": 1.7603672532517214,
+      "grad_norm": 1.762441987335934,
+      "learning_rate": 7.436399511879622e-07,
+      "loss": 0.2604,
+      "step": 11504
+    },
+    {
+      "epoch": 1.760520275439939,
+      "grad_norm": 1.7135814424977367,
+      "learning_rate": 7.42702453549905e-07,
+      "loss": 0.2559,
+      "step": 11505
+    },
+    {
+      "epoch": 1.760673297628156,
+      "grad_norm": 2.2057167673996303,
+      "learning_rate": 7.417655244387045e-07,
+      "loss": 0.2687,
+      "step": 11506
+    },
+    {
+      "epoch": 1.7608263198163734,
+      "grad_norm": 2.0842795751715477,
+      "learning_rate": 7.408291639119014e-07,
+      "loss": 0.2647,
+      "step": 11507
+    },
+    {
+      "epoch": 1.7609793420045907,
+      "grad_norm": 2.034933453675627,
+      "learning_rate": 7.398933720270018e-07,
+      "loss": 0.3,
+      "step": 11508
+    },
+    {
+      "epoch": 1.7611323641928078,
+      "grad_norm": 2.130052559162876,
+      "learning_rate": 7.389581488414754e-07,
+      "loss": 0.3058,
+      "step": 11509
+    },
+    {
+      "epoch": 1.7612853863810254,
+      "grad_norm": 1.902586949096245,
+      "learning_rate": 7.380234944127595e-07,
+      "loss": 0.2671,
+      "step": 11510
+    },
+    {
+      "epoch": 1.7614384085692425,
+      "grad_norm": 2.0243430655818573,
+      "learning_rate": 7.370894087982528e-07,
+      "loss": 0.2402,
+      "step": 11511
+    },
+    {
+      "epoch": 1.7615914307574598,
+      "grad_norm": 2.0713333991695277,
+      "learning_rate": 7.361558920553225e-07,
+      "loss": 0.2736,
+      "step": 11512
+    },
+    {
+      "epoch": 1.7617444529456772,
+      "grad_norm": 2.6500941540661587,
+      "learning_rate": 7.352229442412984e-07,
+      "loss": 0.3045,
+      "step": 11513
+    },
+    {
+      "epoch": 1.7618974751338943,
+      "grad_norm": 2.2426275693857325,
+      "learning_rate": 7.342905654134802e-07,
+      "loss": 0.2958,
+      "step": 11514
+    },
+    {
+      "epoch": 1.7620504973221118,
+      "grad_norm": 2.187412096709696,
+      "learning_rate": 7.333587556291233e-07,
+      "loss": 0.2794,
+      "step": 11515
+    },
+    {
+      "epoch": 1.762203519510329,
+      "grad_norm": 2.086117649436215,
+      "learning_rate": 7.324275149454563e-07,
+      "loss": 0.2924,
+      "step": 11516
+    },
+    {
+      "epoch": 1.7623565416985463,
+      "grad_norm": 2.3882021880357027,
+      "learning_rate": 7.314968434196734e-07,
+      "loss": 0.272,
+      "step": 11517
+    },
+    {
+      "epoch": 1.7625095638867636,
+      "grad_norm": 2.189237862691424,
+      "learning_rate": 7.305667411089257e-07,
+      "loss": 0.2529,
+      "step": 11518
+    },
+    {
+      "epoch": 1.7626625860749807,
+      "grad_norm": 2.2909008823470076,
+      "learning_rate": 7.296372080703351e-07,
+      "loss": 0.3102,
+      "step": 11519
+    },
+    {
+      "epoch": 1.7628156082631983,
+      "grad_norm": 2.479912838161086,
+      "learning_rate": 7.287082443609938e-07,
+      "loss": 0.3642,
+      "step": 11520
+    },
+    {
+      "epoch": 1.7629686304514154,
+      "grad_norm": 2.396033551508515,
+      "learning_rate": 7.27779850037944e-07,
+      "loss": 0.3179,
+      "step": 11521
+    },
+    {
+      "epoch": 1.7631216526396327,
+      "grad_norm": 2.0599105222120686,
+      "learning_rate": 7.268520251582101e-07,
+      "loss": 0.2572,
+      "step": 11522
+    },
+    {
+      "epoch": 1.76327467482785,
+      "grad_norm": 2.1985313014313226,
+      "learning_rate": 7.259247697787697e-07,
+      "loss": 0.2397,
+      "step": 11523
+    },
+    {
+      "epoch": 1.7634276970160672,
+      "grad_norm": 2.5257026577043353,
+      "learning_rate": 7.249980839565673e-07,
+      "loss": 0.4124,
+      "step": 11524
+    },
+    {
+      "epoch": 1.7635807192042847,
+      "grad_norm": 2.1187741614819218,
+      "learning_rate": 7.240719677485186e-07,
+      "loss": 0.2284,
+      "step": 11525
+    },
+    {
+      "epoch": 1.7637337413925018,
+      "grad_norm": 2.27650166500535,
+      "learning_rate": 7.23146421211498e-07,
+      "loss": 0.3091,
+      "step": 11526
+    },
+    {
+      "epoch": 1.7638867635807192,
+      "grad_norm": 2.1431620058394687,
+      "learning_rate": 7.222214444023468e-07,
+      "loss": 0.2679,
+      "step": 11527
+    },
+    {
+      "epoch": 1.7640397857689365,
+      "grad_norm": 2.2747276067202393,
+      "learning_rate": 7.212970373778705e-07,
+      "loss": 0.2761,
+      "step": 11528
+    },
+    {
+      "epoch": 1.7641928079571538,
+      "grad_norm": 2.3816843053242156,
+      "learning_rate": 7.203732001948427e-07,
+      "loss": 0.3181,
+      "step": 11529
+    },
+    {
+      "epoch": 1.7643458301453712,
+      "grad_norm": 1.9811359822858945,
+      "learning_rate": 7.194499329099991e-07,
+      "loss": 0.241,
+      "step": 11530
+    },
+    {
+      "epoch": 1.7644988523335883,
+      "grad_norm": 2.076332988962881,
+      "learning_rate": 7.185272355800399e-07,
+      "loss": 0.318,
+      "step": 11531
+    },
+    {
+      "epoch": 1.7646518745218056,
+      "grad_norm": 2.5309491908046593,
+      "learning_rate": 7.176051082616331e-07,
+      "loss": 0.3363,
+      "step": 11532
+    },
+    {
+      "epoch": 1.764804896710023,
+      "grad_norm": 1.9335674180540168,
+      "learning_rate": 7.16683551011409e-07,
+      "loss": 0.2686,
+      "step": 11533
+    },
+    {
+      "epoch": 1.7649579188982403,
+      "grad_norm": 1.986235927705645,
+      "learning_rate": 7.157625638859634e-07,
+      "loss": 0.2505,
+      "step": 11534
+    },
+    {
+      "epoch": 1.7651109410864576,
+      "grad_norm": 2.013218268869103,
+      "learning_rate": 7.14842146941862e-07,
+      "loss": 0.2755,
+      "step": 11535
+    },
+    {
+      "epoch": 1.7652639632746747,
+      "grad_norm": 2.5751645961695533,
+      "learning_rate": 7.139223002356266e-07,
+      "loss": 0.2704,
+      "step": 11536
+    },
+    {
+      "epoch": 1.7654169854628923,
+      "grad_norm": 1.9783689343552875,
+      "learning_rate": 7.130030238237484e-07,
+      "loss": 0.2792,
+      "step": 11537
+    },
+    {
+      "epoch": 1.7655700076511094,
+      "grad_norm": 1.9929281594053494,
+      "learning_rate": 7.120843177626879e-07,
+      "loss": 0.2913,
+      "step": 11538
+    },
+    {
+      "epoch": 1.7657230298393267,
+      "grad_norm": 2.3890388963972895,
+      "learning_rate": 7.111661821088633e-07,
+      "loss": 0.3186,
+      "step": 11539
+    },
+    {
+      "epoch": 1.765876052027544,
+      "grad_norm": 2.3135962005718786,
+      "learning_rate": 7.102486169186595e-07,
+      "loss": 0.2763,
+      "step": 11540
+    },
+    {
+      "epoch": 1.7660290742157612,
+      "grad_norm": 1.9568397396801063,
+      "learning_rate": 7.093316222484337e-07,
+      "loss": 0.2673,
+      "step": 11541
+    },
+    {
+      "epoch": 1.7661820964039787,
+      "grad_norm": 1.940537721405934,
+      "learning_rate": 7.08415198154494e-07,
+      "loss": 0.2478,
+      "step": 11542
+    },
+    {
+      "epoch": 1.7663351185921958,
+      "grad_norm": 2.2076985256142763,
+      "learning_rate": 7.074993446931288e-07,
+      "loss": 0.277,
+      "step": 11543
+    },
+    {
+      "epoch": 1.7664881407804132,
+      "grad_norm": 1.9708964282837547,
+      "learning_rate": 7.065840619205821e-07,
+      "loss": 0.2539,
+      "step": 11544
+    },
+    {
+      "epoch": 1.7666411629686305,
+      "grad_norm": 2.2308721958293627,
+      "learning_rate": 7.056693498930611e-07,
+      "loss": 0.3503,
+      "step": 11545
+    },
+    {
+      "epoch": 1.7667941851568476,
+      "grad_norm": 2.3264206303422523,
+      "learning_rate": 7.047552086667464e-07,
+      "loss": 0.3175,
+      "step": 11546
+    },
+    {
+      "epoch": 1.7669472073450652,
+      "grad_norm": 2.2786931204458947,
+      "learning_rate": 7.038416382977797e-07,
+      "loss": 0.2961,
+      "step": 11547
+    },
+    {
+      "epoch": 1.7671002295332823,
+      "grad_norm": 2.401697495954377,
+      "learning_rate": 7.029286388422608e-07,
+      "loss": 0.3503,
+      "step": 11548
+    },
+    {
+      "epoch": 1.7672532517214996,
+      "grad_norm": 2.2087065408054634,
+      "learning_rate": 7.020162103562667e-07,
+      "loss": 0.2838,
+      "step": 11549
+    },
+    {
+      "epoch": 1.767406273909717,
+      "grad_norm": 2.0496835744113904,
+      "learning_rate": 7.011043528958306e-07,
+      "loss": 0.2834,
+      "step": 11550
+    },
+    {
+      "epoch": 1.767559296097934,
+      "grad_norm": 2.0960520988601146,
+      "learning_rate": 7.001930665169543e-07,
+      "loss": 0.2617,
+      "step": 11551
+    },
+    {
+      "epoch": 1.7677123182861516,
+      "grad_norm": 2.1397333109537247,
+      "learning_rate": 6.992823512756019e-07,
+      "loss": 0.3063,
+      "step": 11552
+    },
+    {
+      "epoch": 1.7678653404743687,
+      "grad_norm": 2.28631592306058,
+      "learning_rate": 6.983722072277055e-07,
+      "loss": 0.2877,
+      "step": 11553
+    },
+    {
+      "epoch": 1.768018362662586,
+      "grad_norm": 1.7910000877177028,
+      "learning_rate": 6.974626344291602e-07,
+      "loss": 0.2054,
+      "step": 11554
+    },
+    {
+      "epoch": 1.7681713848508034,
+      "grad_norm": 1.8451767717460463,
+      "learning_rate": 6.965536329358258e-07,
+      "loss": 0.2168,
+      "step": 11555
+    },
+    {
+      "epoch": 1.7683244070390205,
+      "grad_norm": 2.277969473191107,
+      "learning_rate": 6.956452028035287e-07,
+      "loss": 0.2725,
+      "step": 11556
+    },
+    {
+      "epoch": 1.768477429227238,
+      "grad_norm": 1.8798605342727712,
+      "learning_rate": 6.947373440880578e-07,
+      "loss": 0.2392,
+      "step": 11557
+    },
+    {
+      "epoch": 1.7686304514154552,
+      "grad_norm": 2.018158526897199,
+      "learning_rate": 6.938300568451695e-07,
+      "loss": 0.2837,
+      "step": 11558
+    },
+    {
+      "epoch": 1.7687834736036725,
+      "grad_norm": 2.082724016431595,
+      "learning_rate": 6.929233411305836e-07,
+      "loss": 0.297,
+      "step": 11559
+    },
+    {
+      "epoch": 1.7689364957918898,
+      "grad_norm": 2.4526921631636722,
+      "learning_rate": 6.920171969999845e-07,
+      "loss": 0.2738,
+      "step": 11560
+    },
+    {
+      "epoch": 1.7690895179801072,
+      "grad_norm": 2.137846305201505,
+      "learning_rate": 6.911116245090222e-07,
+      "loss": 0.2852,
+      "step": 11561
+    },
+    {
+      "epoch": 1.7692425401683245,
+      "grad_norm": 2.1361546200502093,
+      "learning_rate": 6.90206623713312e-07,
+      "loss": 0.3244,
+      "step": 11562
+    },
+    {
+      "epoch": 1.7693955623565416,
+      "grad_norm": 2.2487698262458062,
+      "learning_rate": 6.893021946684309e-07,
+      "loss": 0.3367,
+      "step": 11563
+    },
+    {
+      "epoch": 1.769548584544759,
+      "grad_norm": 2.0760080354126487,
+      "learning_rate": 6.883983374299296e-07,
+      "loss": 0.3091,
+      "step": 11564
+    },
+    {
+      "epoch": 1.7697016067329763,
+      "grad_norm": 2.10544391417939,
+      "learning_rate": 6.874950520533108e-07,
+      "loss": 0.3197,
+      "step": 11565
+    },
+    {
+      "epoch": 1.7698546289211936,
+      "grad_norm": 2.161015077218226,
+      "learning_rate": 6.86592338594051e-07,
+      "loss": 0.2401,
+      "step": 11566
+    },
+    {
+      "epoch": 1.770007651109411,
+      "grad_norm": 2.291425465652151,
+      "learning_rate": 6.856901971075936e-07,
+      "loss": 0.2838,
+      "step": 11567
+    },
+    {
+      "epoch": 1.770160673297628,
+      "grad_norm": 1.9249319567789467,
+      "learning_rate": 6.847886276493365e-07,
+      "loss": 0.2095,
+      "step": 11568
+    },
+    {
+      "epoch": 1.7703136954858456,
+      "grad_norm": 2.20987764398357,
+      "learning_rate": 6.8388763027465e-07,
+      "loss": 0.2599,
+      "step": 11569
+    },
+    {
+      "epoch": 1.7704667176740627,
+      "grad_norm": 2.3077010605825703,
+      "learning_rate": 6.82987205038872e-07,
+      "loss": 0.3059,
+      "step": 11570
+    },
+    {
+      "epoch": 1.77061973986228,
+      "grad_norm": 2.2411642688887174,
+      "learning_rate": 6.820873519972949e-07,
+      "loss": 0.2815,
+      "step": 11571
+    },
+    {
+      "epoch": 1.7707727620504974,
+      "grad_norm": 2.2461254208083763,
+      "learning_rate": 6.811880712051866e-07,
+      "loss": 0.3286,
+      "step": 11572
+    },
+    {
+      "epoch": 1.7709257842387145,
+      "grad_norm": 2.4101762861822853,
+      "learning_rate": 6.802893627177753e-07,
+      "loss": 0.2896,
+      "step": 11573
+    },
+    {
+      "epoch": 1.771078806426932,
+      "grad_norm": 2.2324724632302635,
+      "learning_rate": 6.793912265902514e-07,
+      "loss": 0.2563,
+      "step": 11574
+    },
+    {
+      "epoch": 1.7712318286151492,
+      "grad_norm": 2.409887429701208,
+      "learning_rate": 6.78493662877775e-07,
+      "loss": 0.337,
+      "step": 11575
+    },
+    {
+      "epoch": 1.7713848508033665,
+      "grad_norm": 2.211937176981524,
+      "learning_rate": 6.77596671635471e-07,
+      "loss": 0.2453,
+      "step": 11576
+    },
+    {
+      "epoch": 1.7715378729915838,
+      "grad_norm": 1.940927053286327,
+      "learning_rate": 6.767002529184219e-07,
+      "loss": 0.2219,
+      "step": 11577
+    },
+    {
+      "epoch": 1.771690895179801,
+      "grad_norm": 2.122117699775585,
+      "learning_rate": 6.758044067816849e-07,
+      "loss": 0.2919,
+      "step": 11578
+    },
+    {
+      "epoch": 1.7718439173680185,
+      "grad_norm": 2.014861880625214,
+      "learning_rate": 6.749091332802748e-07,
+      "loss": 0.2203,
+      "step": 11579
+    },
+    {
+      "epoch": 1.7719969395562356,
+      "grad_norm": 2.134298312181937,
+      "learning_rate": 6.740144324691755e-07,
+      "loss": 0.2862,
+      "step": 11580
+    },
+    {
+      "epoch": 1.772149961744453,
+      "grad_norm": 2.1255099152844896,
+      "learning_rate": 6.73120304403333e-07,
+      "loss": 0.2669,
+      "step": 11581
+    },
+    {
+      "epoch": 1.7723029839326703,
+      "grad_norm": 2.523914791255019,
+      "learning_rate": 6.722267491376599e-07,
+      "loss": 0.3238,
+      "step": 11582
+    },
+    {
+      "epoch": 1.7724560061208874,
+      "grad_norm": 1.9316026331139717,
+      "learning_rate": 6.713337667270325e-07,
+      "loss": 0.2682,
+      "step": 11583
+    },
+    {
+      "epoch": 1.772609028309105,
+      "grad_norm": 2.3472903220055366,
+      "learning_rate": 6.704413572262924e-07,
+      "loss": 0.316,
+      "step": 11584
+    },
+    {
+      "epoch": 1.772762050497322,
+      "grad_norm": 2.054524648699194,
+      "learning_rate": 6.695495206902458e-07,
+      "loss": 0.2599,
+      "step": 11585
+    },
+    {
+      "epoch": 1.7729150726855394,
+      "grad_norm": 1.8121343266405059,
+      "learning_rate": 6.686582571736643e-07,
+      "loss": 0.2625,
+      "step": 11586
+    },
+    {
+      "epoch": 1.7730680948737567,
+      "grad_norm": 2.1207606831716674,
+      "learning_rate": 6.677675667312844e-07,
+      "loss": 0.2734,
+      "step": 11587
+    },
+    {
+      "epoch": 1.7732211170619738,
+      "grad_norm": 2.4304987605550656,
+      "learning_rate": 6.668774494178054e-07,
+      "loss": 0.3168,
+      "step": 11588
+    },
+    {
+      "epoch": 1.7733741392501914,
+      "grad_norm": 1.9828230674015517,
+      "learning_rate": 6.659879052878925e-07,
+      "loss": 0.2348,
+      "step": 11589
+    },
+    {
+      "epoch": 1.7735271614384085,
+      "grad_norm": 2.460569333288936,
+      "learning_rate": 6.650989343961788e-07,
+      "loss": 0.3821,
+      "step": 11590
+    },
+    {
+      "epoch": 1.7736801836266258,
+      "grad_norm": 2.016231564778463,
+      "learning_rate": 6.642105367972573e-07,
+      "loss": 0.2461,
+      "step": 11591
+    },
+    {
+      "epoch": 1.7738332058148432,
+      "grad_norm": 1.9688510542138224,
+      "learning_rate": 6.633227125456887e-07,
+      "loss": 0.242,
+      "step": 11592
+    },
+    {
+      "epoch": 1.7739862280030605,
+      "grad_norm": 1.953541653477738,
+      "learning_rate": 6.624354616959971e-07,
+      "loss": 0.2461,
+      "step": 11593
+    },
+    {
+      "epoch": 1.7741392501912778,
+      "grad_norm": 2.1083010117470486,
+      "learning_rate": 6.615487843026735e-07,
+      "loss": 0.2774,
+      "step": 11594
+    },
+    {
+      "epoch": 1.774292272379495,
+      "grad_norm": 2.1541068089107656,
+      "learning_rate": 6.606626804201688e-07,
+      "loss": 0.2902,
+      "step": 11595
+    },
+    {
+      "epoch": 1.7744452945677123,
+      "grad_norm": 2.285536978638378,
+      "learning_rate": 6.597771501029082e-07,
+      "loss": 0.2622,
+      "step": 11596
+    },
+    {
+      "epoch": 1.7745983167559296,
+      "grad_norm": 2.319317293092176,
+      "learning_rate": 6.588921934052705e-07,
+      "loss": 0.2898,
+      "step": 11597
+    },
+    {
+      "epoch": 1.774751338944147,
+      "grad_norm": 1.993917651888502,
+      "learning_rate": 6.580078103816034e-07,
+      "loss": 0.2264,
+      "step": 11598
+    },
+    {
+      "epoch": 1.7749043611323643,
+      "grad_norm": 1.9890587475176882,
+      "learning_rate": 6.571240010862267e-07,
+      "loss": 0.2461,
+      "step": 11599
+    },
+    {
+      "epoch": 1.7750573833205814,
+      "grad_norm": 1.7738468768909665,
+      "learning_rate": 6.562407655734116e-07,
+      "loss": 0.2365,
+      "step": 11600
+    },
+    {
+      "epoch": 1.775210405508799,
+      "grad_norm": 2.229496564560051,
+      "learning_rate": 6.553581038974055e-07,
+      "loss": 0.3035,
+      "step": 11601
+    },
+    {
+      "epoch": 1.775363427697016,
+      "grad_norm": 2.0336619601974375,
+      "learning_rate": 6.544760161124153e-07,
+      "loss": 0.2633,
+      "step": 11602
+    },
+    {
+      "epoch": 1.7755164498852334,
+      "grad_norm": 2.002039939027311,
+      "learning_rate": 6.535945022726109e-07,
+      "loss": 0.2538,
+      "step": 11603
+    },
+    {
+      "epoch": 1.7756694720734507,
+      "grad_norm": 2.0847731213098033,
+      "learning_rate": 6.527135624321324e-07,
+      "loss": 0.3249,
+      "step": 11604
+    },
+    {
+      "epoch": 1.7758224942616678,
+      "grad_norm": 2.2660603591564517,
+      "learning_rate": 6.51833196645083e-07,
+      "loss": 0.3123,
+      "step": 11605
+    },
+    {
+      "epoch": 1.7759755164498854,
+      "grad_norm": 2.432517369709772,
+      "learning_rate": 6.509534049655231e-07,
+      "loss": 0.2786,
+      "step": 11606
+    },
+    {
+      "epoch": 1.7761285386381025,
+      "grad_norm": 1.8314023909408828,
+      "learning_rate": 6.500741874474903e-07,
+      "loss": 0.2399,
+      "step": 11607
+    },
+    {
+      "epoch": 1.7762815608263198,
+      "grad_norm": 2.32761232065426,
+      "learning_rate": 6.491955441449793e-07,
+      "loss": 0.3315,
+      "step": 11608
+    },
+    {
+      "epoch": 1.7764345830145372,
+      "grad_norm": 2.27982039867596,
+      "learning_rate": 6.483174751119503e-07,
+      "loss": 0.289,
+      "step": 11609
+    },
+    {
+      "epoch": 1.7765876052027543,
+      "grad_norm": 2.112027368923221,
+      "learning_rate": 6.47439980402329e-07,
+      "loss": 0.2515,
+      "step": 11610
+    },
+    {
+      "epoch": 1.7767406273909718,
+      "grad_norm": 2.2189829473835383,
+      "learning_rate": 6.465630600700057e-07,
+      "loss": 0.2844,
+      "step": 11611
+    },
+    {
+      "epoch": 1.776893649579189,
+      "grad_norm": 1.9160559036153286,
+      "learning_rate": 6.456867141688361e-07,
+      "loss": 0.2474,
+      "step": 11612
+    },
+    {
+      "epoch": 1.7770466717674063,
+      "grad_norm": 1.8834851653508435,
+      "learning_rate": 6.448109427526394e-07,
+      "loss": 0.3078,
+      "step": 11613
+    },
+    {
+      "epoch": 1.7771996939556236,
+      "grad_norm": 2.0496891897542695,
+      "learning_rate": 6.439357458752005e-07,
+      "loss": 0.2768,
+      "step": 11614
+    },
+    {
+      "epoch": 1.7773527161438407,
+      "grad_norm": 2.071621909374972,
+      "learning_rate": 6.430611235902684e-07,
+      "loss": 0.2088,
+      "step": 11615
+    },
+    {
+      "epoch": 1.7775057383320583,
+      "grad_norm": 2.405026711662064,
+      "learning_rate": 6.421870759515558e-07,
+      "loss": 0.3089,
+      "step": 11616
+    },
+    {
+      "epoch": 1.7776587605202754,
+      "grad_norm": 2.082176155954287,
+      "learning_rate": 6.413136030127465e-07,
+      "loss": 0.2935,
+      "step": 11617
+    },
+    {
+      "epoch": 1.7778117827084927,
+      "grad_norm": 2.0675961828515295,
+      "learning_rate": 6.404407048274775e-07,
+      "loss": 0.3031,
+      "step": 11618
+    },
+    {
+      "epoch": 1.77796480489671,
+      "grad_norm": 2.359503068353368,
+      "learning_rate": 6.395683814493592e-07,
+      "loss": 0.3404,
+      "step": 11619
+    },
+    {
+      "epoch": 1.7781178270849272,
+      "grad_norm": 2.4968081203038293,
+      "learning_rate": 6.386966329319666e-07,
+      "loss": 0.3414,
+      "step": 11620
+    },
+    {
+      "epoch": 1.7782708492731447,
+      "grad_norm": 2.034343682425657,
+      "learning_rate": 6.378254593288347e-07,
+      "loss": 0.2631,
+      "step": 11621
+    },
+    {
+      "epoch": 1.7784238714613618,
+      "grad_norm": 2.01888433987026,
+      "learning_rate": 6.36954860693465e-07,
+      "loss": 0.2474,
+      "step": 11622
+    },
+    {
+      "epoch": 1.7785768936495792,
+      "grad_norm": 2.239054272129354,
+      "learning_rate": 6.360848370793293e-07,
+      "loss": 0.3453,
+      "step": 11623
+    },
+    {
+      "epoch": 1.7787299158377965,
+      "grad_norm": 1.8388079956214614,
+      "learning_rate": 6.352153885398516e-07,
+      "loss": 0.2245,
+      "step": 11624
+    },
+    {
+      "epoch": 1.7788829380260138,
+      "grad_norm": 2.3830486932199335,
+      "learning_rate": 6.343465151284334e-07,
+      "loss": 0.3301,
+      "step": 11625
+    },
+    {
+      "epoch": 1.7790359602142312,
+      "grad_norm": 2.0508156507592705,
+      "learning_rate": 6.334782168984365e-07,
+      "loss": 0.2308,
+      "step": 11626
+    },
+    {
+      "epoch": 1.7791889824024483,
+      "grad_norm": 2.086349526363687,
+      "learning_rate": 6.326104939031818e-07,
+      "loss": 0.2734,
+      "step": 11627
+    },
+    {
+      "epoch": 1.7793420045906656,
+      "grad_norm": 2.1454319811649882,
+      "learning_rate": 6.317433461959622e-07,
+      "loss": 0.2935,
+      "step": 11628
+    },
+    {
+      "epoch": 1.779495026778883,
+      "grad_norm": 2.3671829733376453,
+      "learning_rate": 6.308767738300358e-07,
+      "loss": 0.3281,
+      "step": 11629
+    },
+    {
+      "epoch": 1.7796480489671003,
+      "grad_norm": 2.192383371880891,
+      "learning_rate": 6.30010776858615e-07,
+      "loss": 0.3156,
+      "step": 11630
+    },
+    {
+      "epoch": 1.7798010711553176,
+      "grad_norm": 2.131210440220201,
+      "learning_rate": 6.291453553348892e-07,
+      "loss": 0.3044,
+      "step": 11631
+    },
+    {
+      "epoch": 1.7799540933435347,
+      "grad_norm": 2.5597526386839844,
+      "learning_rate": 6.282805093120059e-07,
+      "loss": 0.3019,
+      "step": 11632
+    },
+    {
+      "epoch": 1.780107115531752,
+      "grad_norm": 2.3159736097903587,
+      "learning_rate": 6.274162388430794e-07,
+      "loss": 0.2879,
+      "step": 11633
+    },
+    {
+      "epoch": 1.7802601377199694,
+      "grad_norm": 2.4289419941363324,
+      "learning_rate": 6.265525439811881e-07,
+      "loss": 0.354,
+      "step": 11634
+    },
+    {
+      "epoch": 1.7804131599081867,
+      "grad_norm": 2.023080467977148,
+      "learning_rate": 6.256894247793732e-07,
+      "loss": 0.2547,
+      "step": 11635
+    },
+    {
+      "epoch": 1.780566182096404,
+      "grad_norm": 1.9144497659810062,
+      "learning_rate": 6.248268812906433e-07,
+      "loss": 0.2787,
+      "step": 11636
+    },
+    {
+      "epoch": 1.7807192042846212,
+      "grad_norm": 2.104341239947922,
+      "learning_rate": 6.239649135679704e-07,
+      "loss": 0.2601,
+      "step": 11637
+    },
+    {
+      "epoch": 1.7808722264728387,
+      "grad_norm": 1.8563273314684772,
+      "learning_rate": 6.231035216642922e-07,
+      "loss": 0.2237,
+      "step": 11638
+    },
+    {
+      "epoch": 1.7810252486610558,
+      "grad_norm": 1.9084349351630756,
+      "learning_rate": 6.222427056325087e-07,
+      "loss": 0.2434,
+      "step": 11639
+    },
+    {
+      "epoch": 1.7811782708492732,
+      "grad_norm": 2.20964827735286,
+      "learning_rate": 6.213824655254874e-07,
+      "loss": 0.3002,
+      "step": 11640
+    },
+    {
+      "epoch": 1.7813312930374905,
+      "grad_norm": 2.0264820560965933,
+      "learning_rate": 6.205228013960574e-07,
+      "loss": 0.2534,
+      "step": 11641
+    },
+    {
+      "epoch": 1.7814843152257076,
+      "grad_norm": 2.2198244041607924,
+      "learning_rate": 6.196637132970151e-07,
+      "loss": 0.2982,
+      "step": 11642
+    },
+    {
+      "epoch": 1.7816373374139252,
+      "grad_norm": 2.051614021142709,
+      "learning_rate": 6.188052012811207e-07,
+      "loss": 0.2739,
+      "step": 11643
+    },
+    {
+      "epoch": 1.7817903596021423,
+      "grad_norm": 2.148961658639848,
+      "learning_rate": 6.179472654010976e-07,
+      "loss": 0.2943,
+      "step": 11644
+    },
+    {
+      "epoch": 1.7819433817903596,
+      "grad_norm": 1.96673620662787,
+      "learning_rate": 6.170899057096347e-07,
+      "loss": 0.3192,
+      "step": 11645
+    },
+    {
+      "epoch": 1.782096403978577,
+      "grad_norm": 2.0614212600984705,
+      "learning_rate": 6.162331222593898e-07,
+      "loss": 0.335,
+      "step": 11646
+    },
+    {
+      "epoch": 1.782249426166794,
+      "grad_norm": 2.0945513989942457,
+      "learning_rate": 6.153769151029765e-07,
+      "loss": 0.2588,
+      "step": 11647
+    },
+    {
+      "epoch": 1.7824024483550116,
+      "grad_norm": 2.131680603392567,
+      "learning_rate": 6.145212842929793e-07,
+      "loss": 0.3115,
+      "step": 11648
+    },
+    {
+      "epoch": 1.7825554705432287,
+      "grad_norm": 2.169429836682061,
+      "learning_rate": 6.136662298819484e-07,
+      "loss": 0.3033,
+      "step": 11649
+    },
+    {
+      "epoch": 1.782708492731446,
+      "grad_norm": 1.8813273070863772,
+      "learning_rate": 6.128117519223919e-07,
+      "loss": 0.2579,
+      "step": 11650
+    },
+    {
+      "epoch": 1.7828615149196634,
+      "grad_norm": 1.9967126429259028,
+      "learning_rate": 6.119578504667878e-07,
+      "loss": 0.2296,
+      "step": 11651
+    },
+    {
+      "epoch": 1.7830145371078805,
+      "grad_norm": 2.325664541784967,
+      "learning_rate": 6.111045255675808e-07,
+      "loss": 0.262,
+      "step": 11652
+    },
+    {
+      "epoch": 1.783167559296098,
+      "grad_norm": 2.1955597124128086,
+      "learning_rate": 6.102517772771721e-07,
+      "loss": 0.3015,
+      "step": 11653
+    },
+    {
+      "epoch": 1.7833205814843152,
+      "grad_norm": 2.2385583608451967,
+      "learning_rate": 6.093996056479356e-07,
+      "loss": 0.3167,
+      "step": 11654
+    },
+    {
+      "epoch": 1.7834736036725325,
+      "grad_norm": 2.0869614373628185,
+      "learning_rate": 6.085480107322072e-07,
+      "loss": 0.3075,
+      "step": 11655
+    },
+    {
+      "epoch": 1.7836266258607498,
+      "grad_norm": 1.8924749927757958,
+      "learning_rate": 6.076969925822829e-07,
+      "loss": 0.2345,
+      "step": 11656
+    },
+    {
+      "epoch": 1.783779648048967,
+      "grad_norm": 2.347694523952781,
+      "learning_rate": 6.068465512504296e-07,
+      "loss": 0.3419,
+      "step": 11657
+    },
+    {
+      "epoch": 1.7839326702371845,
+      "grad_norm": 2.1575066247384034,
+      "learning_rate": 6.059966867888778e-07,
+      "loss": 0.2692,
+      "step": 11658
+    },
+    {
+      "epoch": 1.7840856924254016,
+      "grad_norm": 2.4130816506654935,
+      "learning_rate": 6.051473992498158e-07,
+      "loss": 0.3215,
+      "step": 11659
+    },
+    {
+      "epoch": 1.784238714613619,
+      "grad_norm": 1.8151835358364201,
+      "learning_rate": 6.042986886854063e-07,
+      "loss": 0.252,
+      "step": 11660
+    },
+    {
+      "epoch": 1.7843917368018363,
+      "grad_norm": 1.9735591729644815,
+      "learning_rate": 6.03450555147771e-07,
+      "loss": 0.2327,
+      "step": 11661
+    },
+    {
+      "epoch": 1.7845447589900536,
+      "grad_norm": 2.292921280603086,
+      "learning_rate": 6.026029986889959e-07,
+      "loss": 0.3124,
+      "step": 11662
+    },
+    {
+      "epoch": 1.784697781178271,
+      "grad_norm": 2.143060264993622,
+      "learning_rate": 6.017560193611338e-07,
+      "loss": 0.2716,
+      "step": 11663
+    },
+    {
+      "epoch": 1.784850803366488,
+      "grad_norm": 2.1064055632602052,
+      "learning_rate": 6.00909617216201e-07,
+      "loss": 0.324,
+      "step": 11664
+    },
+    {
+      "epoch": 1.7850038255547054,
+      "grad_norm": 2.491721456754776,
+      "learning_rate": 6.000637923061769e-07,
+      "loss": 0.2778,
+      "step": 11665
+    },
+    {
+      "epoch": 1.7851568477429227,
+      "grad_norm": 1.9397667295022485,
+      "learning_rate": 5.99218544683009e-07,
+      "loss": 0.2748,
+      "step": 11666
+    },
+    {
+      "epoch": 1.78530986993114,
+      "grad_norm": 2.0315429084869323,
+      "learning_rate": 5.983738743986068e-07,
+      "loss": 0.2424,
+      "step": 11667
+    },
+    {
+      "epoch": 1.7854628921193574,
+      "grad_norm": 2.3518604760030946,
+      "learning_rate": 5.975297815048442e-07,
+      "loss": 0.3663,
+      "step": 11668
+    },
+    {
+      "epoch": 1.7856159143075745,
+      "grad_norm": 2.260235840291961,
+      "learning_rate": 5.9668626605356e-07,
+      "loss": 0.3246,
+      "step": 11669
+    },
+    {
+      "epoch": 1.785768936495792,
+      "grad_norm": 2.2893769844413283,
+      "learning_rate": 5.958433280965581e-07,
+      "loss": 0.2896,
+      "step": 11670
+    },
+    {
+      "epoch": 1.7859219586840092,
+      "grad_norm": 2.2230273719541893,
+      "learning_rate": 5.950009676856061e-07,
+      "loss": 0.36,
+      "step": 11671
+    },
+    {
+      "epoch": 1.7860749808722265,
+      "grad_norm": 2.3200719885128858,
+      "learning_rate": 5.941591848724381e-07,
+      "loss": 0.3401,
+      "step": 11672
+    },
+    {
+      "epoch": 1.7862280030604438,
+      "grad_norm": 2.1974597281976593,
+      "learning_rate": 5.933179797087507e-07,
+      "loss": 0.3361,
+      "step": 11673
+    },
+    {
+      "epoch": 1.786381025248661,
+      "grad_norm": 2.5283285291331223,
+      "learning_rate": 5.924773522462024e-07,
+      "loss": 0.2941,
+      "step": 11674
+    },
+    {
+      "epoch": 1.7865340474368785,
+      "grad_norm": 2.112766463182009,
+      "learning_rate": 5.916373025364264e-07,
+      "loss": 0.2504,
+      "step": 11675
+    },
+    {
+      "epoch": 1.7866870696250956,
+      "grad_norm": 2.1968271904468812,
+      "learning_rate": 5.907978306310081e-07,
+      "loss": 0.311,
+      "step": 11676
+    },
+    {
+      "epoch": 1.786840091813313,
+      "grad_norm": 2.21326844605421,
+      "learning_rate": 5.899589365815018e-07,
+      "loss": 0.346,
+      "step": 11677
+    },
+    {
+      "epoch": 1.7869931140015303,
+      "grad_norm": 2.249810646236595,
+      "learning_rate": 5.891206204394329e-07,
+      "loss": 0.2816,
+      "step": 11678
+    },
+    {
+      "epoch": 1.7871461361897474,
+      "grad_norm": 1.8485888703819569,
+      "learning_rate": 5.882828822562814e-07,
+      "loss": 0.2307,
+      "step": 11679
+    },
+    {
+      "epoch": 1.787299158377965,
+      "grad_norm": 2.0267399285945404,
+      "learning_rate": 5.87445722083494e-07,
+      "loss": 0.2479,
+      "step": 11680
+    },
+    {
+      "epoch": 1.787452180566182,
+      "grad_norm": 2.0911961020778653,
+      "learning_rate": 5.866091399724916e-07,
+      "loss": 0.3178,
+      "step": 11681
+    },
+    {
+      "epoch": 1.7876052027543994,
+      "grad_norm": 1.9366386472362302,
+      "learning_rate": 5.857731359746433e-07,
+      "loss": 0.2728,
+      "step": 11682
+    },
+    {
+      "epoch": 1.7877582249426167,
+      "grad_norm": 2.0375420240316817,
+      "learning_rate": 5.849377101412967e-07,
+      "loss": 0.3075,
+      "step": 11683
+    },
+    {
+      "epoch": 1.7879112471308338,
+      "grad_norm": 1.7988385396357176,
+      "learning_rate": 5.841028625237589e-07,
+      "loss": 0.2719,
+      "step": 11684
+    },
+    {
+      "epoch": 1.7880642693190514,
+      "grad_norm": 2.5279457807415464,
+      "learning_rate": 5.832685931732963e-07,
+      "loss": 0.3454,
+      "step": 11685
+    },
+    {
+      "epoch": 1.7882172915072685,
+      "grad_norm": 2.313295462058856,
+      "learning_rate": 5.824349021411502e-07,
+      "loss": 0.256,
+      "step": 11686
+    },
+    {
+      "epoch": 1.7883703136954858,
+      "grad_norm": 2.1886780113954254,
+      "learning_rate": 5.816017894785197e-07,
+      "loss": 0.2694,
+      "step": 11687
+    },
+    {
+      "epoch": 1.7885233358837032,
+      "grad_norm": 1.8681046262205732,
+      "learning_rate": 5.807692552365652e-07,
+      "loss": 0.2614,
+      "step": 11688
+    },
+    {
+      "epoch": 1.7886763580719203,
+      "grad_norm": 2.0027057223970317,
+      "learning_rate": 5.799372994664199e-07,
+      "loss": 0.259,
+      "step": 11689
+    },
+    {
+      "epoch": 1.7888293802601378,
+      "grad_norm": 2.337302769224805,
+      "learning_rate": 5.791059222191764e-07,
+      "loss": 0.2759,
+      "step": 11690
+    },
+    {
+      "epoch": 1.788982402448355,
+      "grad_norm": 2.121766483543593,
+      "learning_rate": 5.782751235458939e-07,
+      "loss": 0.323,
+      "step": 11691
+    },
+    {
+      "epoch": 1.7891354246365723,
+      "grad_norm": 1.9704435983259665,
+      "learning_rate": 5.774449034975938e-07,
+      "loss": 0.3234,
+      "step": 11692
+    },
+    {
+      "epoch": 1.7892884468247896,
+      "grad_norm": 2.0983690476734305,
+      "learning_rate": 5.766152621252629e-07,
+      "loss": 0.2725,
+      "step": 11693
+    },
+    {
+      "epoch": 1.789441469013007,
+      "grad_norm": 2.1429735780924744,
+      "learning_rate": 5.757861994798542e-07,
+      "loss": 0.2911,
+      "step": 11694
+    },
+    {
+      "epoch": 1.7895944912012243,
+      "grad_norm": 1.9899945550379567,
+      "learning_rate": 5.749577156122821e-07,
+      "loss": 0.2584,
+      "step": 11695
+    },
+    {
+      "epoch": 1.7897475133894414,
+      "grad_norm": 2.1339230085642242,
+      "learning_rate": 5.741298105734272e-07,
+      "loss": 0.2873,
+      "step": 11696
+    },
+    {
+      "epoch": 1.7899005355776587,
+      "grad_norm": 1.9769567723769148,
+      "learning_rate": 5.733024844141355e-07,
+      "loss": 0.2111,
+      "step": 11697
+    },
+    {
+      "epoch": 1.790053557765876,
+      "grad_norm": 1.8199371530775728,
+      "learning_rate": 5.724757371852141e-07,
+      "loss": 0.2833,
+      "step": 11698
+    },
+    {
+      "epoch": 1.7902065799540934,
+      "grad_norm": 2.204375861271401,
+      "learning_rate": 5.716495689374413e-07,
+      "loss": 0.3179,
+      "step": 11699
+    },
+    {
+      "epoch": 1.7903596021423107,
+      "grad_norm": 2.392939646494985,
+      "learning_rate": 5.708239797215509e-07,
+      "loss": 0.2879,
+      "step": 11700
+    },
+    {
+      "epoch": 1.7905126243305278,
+      "grad_norm": 2.262298036101752,
+      "learning_rate": 5.699989695882458e-07,
+      "loss": 0.3352,
+      "step": 11701
+    },
+    {
+      "epoch": 1.7906656465187454,
+      "grad_norm": 1.9027329378061686,
+      "learning_rate": 5.691745385881964e-07,
+      "loss": 0.1847,
+      "step": 11702
+    },
+    {
+      "epoch": 1.7908186687069625,
+      "grad_norm": 2.150646438874415,
+      "learning_rate": 5.683506867720301e-07,
+      "loss": 0.2933,
+      "step": 11703
+    },
+    {
+      "epoch": 1.7909716908951798,
+      "grad_norm": 2.2417842603845526,
+      "learning_rate": 5.675274141903464e-07,
+      "loss": 0.2996,
+      "step": 11704
+    },
+    {
+      "epoch": 1.7911247130833972,
+      "grad_norm": 2.346125193672816,
+      "learning_rate": 5.66704720893706e-07,
+      "loss": 0.2515,
+      "step": 11705
+    },
+    {
+      "epoch": 1.7912777352716143,
+      "grad_norm": 1.9921160410033858,
+      "learning_rate": 5.658826069326284e-07,
+      "loss": 0.2964,
+      "step": 11706
+    },
+    {
+      "epoch": 1.7914307574598318,
+      "grad_norm": 2.1445987832520053,
+      "learning_rate": 5.650610723576078e-07,
+      "loss": 0.2592,
+      "step": 11707
+    },
+    {
+      "epoch": 1.791583779648049,
+      "grad_norm": 2.0694519109933727,
+      "learning_rate": 5.642401172190981e-07,
+      "loss": 0.2388,
+      "step": 11708
+    },
+    {
+      "epoch": 1.7917368018362663,
+      "grad_norm": 2.238611258670583,
+      "learning_rate": 5.634197415675124e-07,
+      "loss": 0.2617,
+      "step": 11709
+    },
+    {
+      "epoch": 1.7918898240244836,
+      "grad_norm": 2.151535537763062,
+      "learning_rate": 5.625999454532383e-07,
+      "loss": 0.2704,
+      "step": 11710
+    },
+    {
+      "epoch": 1.7920428462127007,
+      "grad_norm": 2.059430799019778,
+      "learning_rate": 5.617807289266208e-07,
+      "loss": 0.3045,
+      "step": 11711
+    },
+    {
+      "epoch": 1.7921958684009183,
+      "grad_norm": 2.119929153330778,
+      "learning_rate": 5.609620920379721e-07,
+      "loss": 0.3171,
+      "step": 11712
+    },
+    {
+      "epoch": 1.7923488905891354,
+      "grad_norm": 1.9604155333595272,
+      "learning_rate": 5.601440348375653e-07,
+      "loss": 0.2426,
+      "step": 11713
+    },
+    {
+      "epoch": 1.7925019127773527,
+      "grad_norm": 1.9606199783978195,
+      "learning_rate": 5.593265573756434e-07,
+      "loss": 0.2463,
+      "step": 11714
+    },
+    {
+      "epoch": 1.79265493496557,
+      "grad_norm": 2.0879423430699737,
+      "learning_rate": 5.585096597024098e-07,
+      "loss": 0.2984,
+      "step": 11715
+    },
+    {
+      "epoch": 1.7928079571537872,
+      "grad_norm": 2.1097963203212102,
+      "learning_rate": 5.576933418680331e-07,
+      "loss": 0.3156,
+      "step": 11716
+    },
+    {
+      "epoch": 1.7929609793420047,
+      "grad_norm": 1.9424617217380995,
+      "learning_rate": 5.568776039226465e-07,
+      "loss": 0.2702,
+      "step": 11717
+    },
+    {
+      "epoch": 1.7931140015302218,
+      "grad_norm": 2.0457163778091196,
+      "learning_rate": 5.560624459163488e-07,
+      "loss": 0.2134,
+      "step": 11718
+    },
+    {
+      "epoch": 1.7932670237184392,
+      "grad_norm": 2.2486400242745916,
+      "learning_rate": 5.552478678992001e-07,
+      "loss": 0.244,
+      "step": 11719
+    },
+    {
+      "epoch": 1.7934200459066565,
+      "grad_norm": 2.1376451976669166,
+      "learning_rate": 5.544338699212281e-07,
+      "loss": 0.216,
+      "step": 11720
+    },
+    {
+      "epoch": 1.7935730680948736,
+      "grad_norm": 2.388209980608498,
+      "learning_rate": 5.536204520324239e-07,
+      "loss": 0.2641,
+      "step": 11721
+    },
+    {
+      "epoch": 1.7937260902830912,
+      "grad_norm": 2.1460112607255053,
+      "learning_rate": 5.528076142827422e-07,
+      "loss": 0.2468,
+      "step": 11722
+    },
+    {
+      "epoch": 1.7938791124713083,
+      "grad_norm": 2.0739753388599462,
+      "learning_rate": 5.519953567221025e-07,
+      "loss": 0.2704,
+      "step": 11723
+    },
+    {
+      "epoch": 1.7940321346595256,
+      "grad_norm": 2.0422144203755295,
+      "learning_rate": 5.511836794003889e-07,
+      "loss": 0.2772,
+      "step": 11724
+    },
+    {
+      "epoch": 1.794185156847743,
+      "grad_norm": 2.0214617735698037,
+      "learning_rate": 5.503725823674488e-07,
+      "loss": 0.2763,
+      "step": 11725
+    },
+    {
+      "epoch": 1.7943381790359603,
+      "grad_norm": 1.7177450706562944,
+      "learning_rate": 5.49562065673096e-07,
+      "loss": 0.2216,
+      "step": 11726
+    },
+    {
+      "epoch": 1.7944912012241776,
+      "grad_norm": 2.0555841425573567,
+      "learning_rate": 5.48752129367105e-07,
+      "loss": 0.2227,
+      "step": 11727
+    },
+    {
+      "epoch": 1.7946442234123947,
+      "grad_norm": 2.080887556455304,
+      "learning_rate": 5.479427734992226e-07,
+      "loss": 0.2297,
+      "step": 11728
+    },
+    {
+      "epoch": 1.794797245600612,
+      "grad_norm": 2.1233295078008037,
+      "learning_rate": 5.471339981191503e-07,
+      "loss": 0.3034,
+      "step": 11729
+    },
+    {
+      "epoch": 1.7949502677888294,
+      "grad_norm": 2.3167717650903046,
+      "learning_rate": 5.463258032765573e-07,
+      "loss": 0.2657,
+      "step": 11730
+    },
+    {
+      "epoch": 1.7951032899770467,
+      "grad_norm": 2.319827835372855,
+      "learning_rate": 5.455181890210814e-07,
+      "loss": 0.2852,
+      "step": 11731
+    },
+    {
+      "epoch": 1.795256312165264,
+      "grad_norm": 1.9427112608034016,
+      "learning_rate": 5.447111554023199e-07,
+      "loss": 0.2446,
+      "step": 11732
+    },
+    {
+      "epoch": 1.7954093343534812,
+      "grad_norm": 2.0329376925996603,
+      "learning_rate": 5.43904702469833e-07,
+      "loss": 0.2396,
+      "step": 11733
+    },
+    {
+      "epoch": 1.7955623565416987,
+      "grad_norm": 1.7308500793905726,
+      "learning_rate": 5.430988302731544e-07,
+      "loss": 0.2522,
+      "step": 11734
+    },
+    {
+      "epoch": 1.7957153787299158,
+      "grad_norm": 2.1219125389826554,
+      "learning_rate": 5.422935388617689e-07,
+      "loss": 0.2545,
+      "step": 11735
+    },
+    {
+      "epoch": 1.7958684009181332,
+      "grad_norm": 1.972590600455119,
+      "learning_rate": 5.414888282851372e-07,
+      "loss": 0.2451,
+      "step": 11736
+    },
+    {
+      "epoch": 1.7960214231063505,
+      "grad_norm": 1.9834371644107958,
+      "learning_rate": 5.406846985926805e-07,
+      "loss": 0.2387,
+      "step": 11737
+    },
+    {
+      "epoch": 1.7961744452945676,
+      "grad_norm": 1.9935226037345846,
+      "learning_rate": 5.398811498337786e-07,
+      "loss": 0.2365,
+      "step": 11738
+    },
+    {
+      "epoch": 1.7963274674827852,
+      "grad_norm": 3.541082073396699,
+      "learning_rate": 5.39078182057785e-07,
+      "loss": 0.3011,
+      "step": 11739
+    },
+    {
+      "epoch": 1.7964804896710023,
+      "grad_norm": 2.2449387681269695,
+      "learning_rate": 5.382757953140105e-07,
+      "loss": 0.2998,
+      "step": 11740
+    },
+    {
+      "epoch": 1.7966335118592196,
+      "grad_norm": 2.373753123628044,
+      "learning_rate": 5.374739896517345e-07,
+      "loss": 0.3223,
+      "step": 11741
+    },
+    {
+      "epoch": 1.796786534047437,
+      "grad_norm": 2.3114047848746972,
+      "learning_rate": 5.366727651201986e-07,
+      "loss": 0.2996,
+      "step": 11742
+    },
+    {
+      "epoch": 1.796939556235654,
+      "grad_norm": 2.0196626112232257,
+      "learning_rate": 5.358721217686091e-07,
+      "loss": 0.2923,
+      "step": 11743
+    },
+    {
+      "epoch": 1.7970925784238716,
+      "grad_norm": 2.035219870357041,
+      "learning_rate": 5.350720596461367e-07,
+      "loss": 0.2943,
+      "step": 11744
+    },
+    {
+      "epoch": 1.7972456006120887,
+      "grad_norm": 2.2123252681687724,
+      "learning_rate": 5.342725788019154e-07,
+      "loss": 0.2645,
+      "step": 11745
+    },
+    {
+      "epoch": 1.797398622800306,
+      "grad_norm": 2.1569997981072664,
+      "learning_rate": 5.334736792850448e-07,
+      "loss": 0.2261,
+      "step": 11746
+    },
+    {
+      "epoch": 1.7975516449885234,
+      "grad_norm": 2.420526338846463,
+      "learning_rate": 5.326753611445901e-07,
+      "loss": 0.2685,
+      "step": 11747
+    },
+    {
+      "epoch": 1.7977046671767405,
+      "grad_norm": 2.333183084239707,
+      "learning_rate": 5.318776244295776e-07,
+      "loss": 0.2612,
+      "step": 11748
+    },
+    {
+      "epoch": 1.797857689364958,
+      "grad_norm": 2.0257483589647833,
+      "learning_rate": 5.310804691889993e-07,
+      "loss": 0.2629,
+      "step": 11749
+    },
+    {
+      "epoch": 1.7980107115531752,
+      "grad_norm": 2.0788523474030485,
+      "learning_rate": 5.302838954718115e-07,
+      "loss": 0.3047,
+      "step": 11750
+    },
+    {
+      "epoch": 1.7981637337413925,
+      "grad_norm": 2.1004294540442534,
+      "learning_rate": 5.294879033269362e-07,
+      "loss": 0.2874,
+      "step": 11751
+    },
+    {
+      "epoch": 1.7983167559296098,
+      "grad_norm": 2.1730252469221836,
+      "learning_rate": 5.286924928032566e-07,
+      "loss": 0.2628,
+      "step": 11752
+    },
+    {
+      "epoch": 1.798469778117827,
+      "grad_norm": 2.1484332701793125,
+      "learning_rate": 5.278976639496236e-07,
+      "loss": 0.2705,
+      "step": 11753
+    },
+    {
+      "epoch": 1.7986228003060445,
+      "grad_norm": 2.0795707539418697,
+      "learning_rate": 5.271034168148504e-07,
+      "loss": 0.2895,
+      "step": 11754
+    },
+    {
+      "epoch": 1.7987758224942616,
+      "grad_norm": 2.0650753087535003,
+      "learning_rate": 5.263097514477145e-07,
+      "loss": 0.2599,
+      "step": 11755
+    },
+    {
+      "epoch": 1.798928844682479,
+      "grad_norm": 2.2929545293487643,
+      "learning_rate": 5.25516667896957e-07,
+      "loss": 0.3368,
+      "step": 11756
+    },
+    {
+      "epoch": 1.7990818668706963,
+      "grad_norm": 2.171294433545801,
+      "learning_rate": 5.247241662112878e-07,
+      "loss": 0.3341,
+      "step": 11757
+    },
+    {
+      "epoch": 1.7992348890589134,
+      "grad_norm": 2.510967028431956,
+      "learning_rate": 5.239322464393726e-07,
+      "loss": 0.3286,
+      "step": 11758
+    },
+    {
+      "epoch": 1.799387911247131,
+      "grad_norm": 2.5373624082576667,
+      "learning_rate": 5.231409086298489e-07,
+      "loss": 0.3282,
+      "step": 11759
+    },
+    {
+      "epoch": 1.799540933435348,
+      "grad_norm": 2.3404830830993664,
+      "learning_rate": 5.223501528313179e-07,
+      "loss": 0.2916,
+      "step": 11760
+    },
+    {
+      "epoch": 1.7996939556235654,
+      "grad_norm": 2.4516704197186145,
+      "learning_rate": 5.215599790923387e-07,
+      "loss": 0.3127,
+      "step": 11761
+    },
+    {
+      "epoch": 1.7998469778117827,
+      "grad_norm": 1.8951882844356727,
+      "learning_rate": 5.2077038746144e-07,
+      "loss": 0.2737,
+      "step": 11762
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.1140050714971848,
+      "learning_rate": 5.199813779871187e-07,
+      "loss": 0.3123,
+      "step": 11763
+    },
+    {
+      "epoch": 1.8001530221882174,
+      "grad_norm": 2.3629890157765683,
+      "learning_rate": 5.191929507178228e-07,
+      "loss": 0.2923,
+      "step": 11764
+    },
+    {
+      "epoch": 1.8003060443764345,
+      "grad_norm": 2.2207657842982704,
+      "learning_rate": 5.18405105701979e-07,
+      "loss": 0.2737,
+      "step": 11765
+    },
+    {
+      "epoch": 1.8004590665646518,
+      "grad_norm": 2.2328795744392687,
+      "learning_rate": 5.176178429879708e-07,
+      "loss": 0.2769,
+      "step": 11766
+    },
+    {
+      "epoch": 1.8006120887528692,
+      "grad_norm": 2.2494304651674804,
+      "learning_rate": 5.16831162624144e-07,
+      "loss": 0.305,
+      "step": 11767
+    },
+    {
+      "epoch": 1.8007651109410865,
+      "grad_norm": 2.097178530222374,
+      "learning_rate": 5.160450646588144e-07,
+      "loss": 0.2687,
+      "step": 11768
+    },
+    {
+      "epoch": 1.8009181331293038,
+      "grad_norm": 1.949079996975221,
+      "learning_rate": 5.152595491402612e-07,
+      "loss": 0.2349,
+      "step": 11769
+    },
+    {
+      "epoch": 1.801071155317521,
+      "grad_norm": 1.9194856467463366,
+      "learning_rate": 5.144746161167202e-07,
+      "loss": 0.2499,
+      "step": 11770
+    },
+    {
+      "epoch": 1.8012241775057385,
+      "grad_norm": 1.9094906109398266,
+      "learning_rate": 5.136902656364018e-07,
+      "loss": 0.263,
+      "step": 11771
+    },
+    {
+      "epoch": 1.8013771996939556,
+      "grad_norm": 2.0543659422279172,
+      "learning_rate": 5.129064977474752e-07,
+      "loss": 0.2677,
+      "step": 11772
+    },
+    {
+      "epoch": 1.801530221882173,
+      "grad_norm": 2.037500846294362,
+      "learning_rate": 5.121233124980751e-07,
+      "loss": 0.277,
+      "step": 11773
+    },
+    {
+      "epoch": 1.8016832440703903,
+      "grad_norm": 2.5064908670823125,
+      "learning_rate": 5.113407099362977e-07,
+      "loss": 0.3037,
+      "step": 11774
+    },
+    {
+      "epoch": 1.8018362662586074,
+      "grad_norm": 2.3767799879674065,
+      "learning_rate": 5.105586901102078e-07,
+      "loss": 0.3141,
+      "step": 11775
+    },
+    {
+      "epoch": 1.801989288446825,
+      "grad_norm": 2.1263927705629877,
+      "learning_rate": 5.097772530678325e-07,
+      "loss": 0.2535,
+      "step": 11776
+    },
+    {
+      "epoch": 1.802142310635042,
+      "grad_norm": 2.101381107662464,
+      "learning_rate": 5.089963988571589e-07,
+      "loss": 0.2888,
+      "step": 11777
+    },
+    {
+      "epoch": 1.8022953328232594,
+      "grad_norm": 2.1648239103094844,
+      "learning_rate": 5.082161275261499e-07,
+      "loss": 0.2802,
+      "step": 11778
+    },
+    {
+      "epoch": 1.8024483550114767,
+      "grad_norm": 2.4059903176752617,
+      "learning_rate": 5.074364391227182e-07,
+      "loss": 0.2784,
+      "step": 11779
+    },
+    {
+      "epoch": 1.8026013771996938,
+      "grad_norm": 2.0832771851068754,
+      "learning_rate": 5.066573336947489e-07,
+      "loss": 0.3067,
+      "step": 11780
+    },
+    {
+      "epoch": 1.8027543993879114,
+      "grad_norm": 1.8546774228621359,
+      "learning_rate": 5.058788112900925e-07,
+      "loss": 0.2336,
+      "step": 11781
+    },
+    {
+      "epoch": 1.8029074215761285,
+      "grad_norm": 1.9287318570324383,
+      "learning_rate": 5.051008719565597e-07,
+      "loss": 0.2113,
+      "step": 11782
+    },
+    {
+      "epoch": 1.8030604437643458,
+      "grad_norm": 2.329499263317703,
+      "learning_rate": 5.043235157419246e-07,
+      "loss": 0.2933,
+      "step": 11783
+    },
+    {
+      "epoch": 1.8032134659525632,
+      "grad_norm": 2.0964211634689702,
+      "learning_rate": 5.035467426939322e-07,
+      "loss": 0.3175,
+      "step": 11784
+    },
+    {
+      "epoch": 1.8033664881407803,
+      "grad_norm": 1.8844374172641056,
+      "learning_rate": 5.027705528602822e-07,
+      "loss": 0.2303,
+      "step": 11785
+    },
+    {
+      "epoch": 1.8035195103289978,
+      "grad_norm": 1.9323244223377913,
+      "learning_rate": 5.019949462886476e-07,
+      "loss": 0.2753,
+      "step": 11786
+    },
+    {
+      "epoch": 1.803672532517215,
+      "grad_norm": 2.3979087439781517,
+      "learning_rate": 5.012199230266612e-07,
+      "loss": 0.2833,
+      "step": 11787
+    },
+    {
+      "epoch": 1.8038255547054323,
+      "grad_norm": 2.066289773156986,
+      "learning_rate": 5.004454831219152e-07,
+      "loss": 0.2935,
+      "step": 11788
+    },
+    {
+      "epoch": 1.8039785768936496,
+      "grad_norm": 2.037769789992835,
+      "learning_rate": 4.99671626621977e-07,
+      "loss": 0.2397,
+      "step": 11789
+    },
+    {
+      "epoch": 1.8041315990818667,
+      "grad_norm": 2.0152504446404653,
+      "learning_rate": 4.988983535743697e-07,
+      "loss": 0.2613,
+      "step": 11790
+    },
+    {
+      "epoch": 1.8042846212700843,
+      "grad_norm": 1.7574012569673563,
+      "learning_rate": 4.981256640265808e-07,
+      "loss": 0.2328,
+      "step": 11791
+    },
+    {
+      "epoch": 1.8044376434583014,
+      "grad_norm": 2.189575784945479,
+      "learning_rate": 4.97353558026068e-07,
+      "loss": 0.3055,
+      "step": 11792
+    },
+    {
+      "epoch": 1.8045906656465187,
+      "grad_norm": 2.131528056103428,
+      "learning_rate": 4.965820356202478e-07,
+      "loss": 0.2894,
+      "step": 11793
+    },
+    {
+      "epoch": 1.804743687834736,
+      "grad_norm": 2.1817272506538496,
+      "learning_rate": 4.958110968565022e-07,
+      "loss": 0.3182,
+      "step": 11794
+    },
+    {
+      "epoch": 1.8048967100229534,
+      "grad_norm": 2.547644929513796,
+      "learning_rate": 4.950407417821779e-07,
+      "loss": 0.2626,
+      "step": 11795
+    },
+    {
+      "epoch": 1.8050497322111707,
+      "grad_norm": 2.012545629577886,
+      "learning_rate": 4.942709704445858e-07,
+      "loss": 0.298,
+      "step": 11796
+    },
+    {
+      "epoch": 1.8052027543993878,
+      "grad_norm": 2.108012508898975,
+      "learning_rate": 4.935017828909994e-07,
+      "loss": 0.2551,
+      "step": 11797
+    },
+    {
+      "epoch": 1.8053557765876052,
+      "grad_norm": 2.2952470925887303,
+      "learning_rate": 4.927331791686585e-07,
+      "loss": 0.26,
+      "step": 11798
+    },
+    {
+      "epoch": 1.8055087987758225,
+      "grad_norm": 2.075299206713513,
+      "learning_rate": 4.919651593247654e-07,
+      "loss": 0.2831,
+      "step": 11799
+    },
+    {
+      "epoch": 1.8056618209640398,
+      "grad_norm": 2.3021425499492483,
+      "learning_rate": 4.911977234064868e-07,
+      "loss": 0.2906,
+      "step": 11800
+    },
+    {
+      "epoch": 1.8058148431522572,
+      "grad_norm": 1.9590477977426395,
+      "learning_rate": 4.904308714609562e-07,
+      "loss": 0.2897,
+      "step": 11801
+    },
+    {
+      "epoch": 1.8059678653404743,
+      "grad_norm": 1.9965251043147465,
+      "learning_rate": 4.896646035352659e-07,
+      "loss": 0.3132,
+      "step": 11802
+    },
+    {
+      "epoch": 1.8061208875286918,
+      "grad_norm": 2.373562135530326,
+      "learning_rate": 4.888989196764782e-07,
+      "loss": 0.3023,
+      "step": 11803
+    },
+    {
+      "epoch": 1.806273909716909,
+      "grad_norm": 2.1986562880820486,
+      "learning_rate": 4.881338199316144e-07,
+      "loss": 0.2941,
+      "step": 11804
+    },
+    {
+      "epoch": 1.8064269319051263,
+      "grad_norm": 2.040180272914321,
+      "learning_rate": 4.873693043476646e-07,
+      "loss": 0.2436,
+      "step": 11805
+    },
+    {
+      "epoch": 1.8065799540933436,
+      "grad_norm": 1.918033241126225,
+      "learning_rate": 4.86605372971577e-07,
+      "loss": 0.2649,
+      "step": 11806
+    },
+    {
+      "epoch": 1.8067329762815607,
+      "grad_norm": 2.5039834522606426,
+      "learning_rate": 4.858420258502727e-07,
+      "loss": 0.459,
+      "step": 11807
+    },
+    {
+      "epoch": 1.8068859984697783,
+      "grad_norm": 2.1500788017787653,
+      "learning_rate": 4.850792630306289e-07,
+      "loss": 0.2669,
+      "step": 11808
+    },
+    {
+      "epoch": 1.8070390206579954,
+      "grad_norm": 1.7573674117294718,
+      "learning_rate": 4.84317084559488e-07,
+      "loss": 0.2311,
+      "step": 11809
+    },
+    {
+      "epoch": 1.8071920428462127,
+      "grad_norm": 2.116328054003031,
+      "learning_rate": 4.835554904836626e-07,
+      "loss": 0.2959,
+      "step": 11810
+    },
+    {
+      "epoch": 1.80734506503443,
+      "grad_norm": 1.9941371350603454,
+      "learning_rate": 4.82794480849923e-07,
+      "loss": 0.2193,
+      "step": 11811
+    },
+    {
+      "epoch": 1.8074980872226472,
+      "grad_norm": 2.4572585198173367,
+      "learning_rate": 4.82034055705004e-07,
+      "loss": 0.3226,
+      "step": 11812
+    },
+    {
+      "epoch": 1.8076511094108647,
+      "grad_norm": 2.18333804741628,
+      "learning_rate": 4.812742150956107e-07,
+      "loss": 0.2796,
+      "step": 11813
+    },
+    {
+      "epoch": 1.8078041315990818,
+      "grad_norm": 2.015761932746089,
+      "learning_rate": 4.805149590684022e-07,
+      "loss": 0.2538,
+      "step": 11814
+    },
+    {
+      "epoch": 1.8079571537872992,
+      "grad_norm": 2.2635890211246497,
+      "learning_rate": 4.797562876700124e-07,
+      "loss": 0.2918,
+      "step": 11815
+    },
+    {
+      "epoch": 1.8081101759755165,
+      "grad_norm": 2.0019627008557994,
+      "learning_rate": 4.78998200947034e-07,
+      "loss": 0.2271,
+      "step": 11816
+    },
+    {
+      "epoch": 1.8082631981637336,
+      "grad_norm": 1.9257373438476535,
+      "learning_rate": 4.782406989460197e-07,
+      "loss": 0.2475,
+      "step": 11817
+    },
+    {
+      "epoch": 1.8084162203519512,
+      "grad_norm": 2.1176835046895177,
+      "learning_rate": 4.774837817134937e-07,
+      "loss": 0.2816,
+      "step": 11818
+    },
+    {
+      "epoch": 1.8085692425401683,
+      "grad_norm": 2.15777176903622,
+      "learning_rate": 4.7672744929594396e-07,
+      "loss": 0.3033,
+      "step": 11819
+    },
+    {
+      "epoch": 1.8087222647283856,
+      "grad_norm": 1.9143295838287768,
+      "learning_rate": 4.759717017398124e-07,
+      "loss": 0.2498,
+      "step": 11820
+    },
+    {
+      "epoch": 1.808875286916603,
+      "grad_norm": 2.0434931578114623,
+      "learning_rate": 4.752165390915198e-07,
+      "loss": 0.2489,
+      "step": 11821
+    },
+    {
+      "epoch": 1.80902830910482,
+      "grad_norm": 1.6784767679091706,
+      "learning_rate": 4.744619613974399e-07,
+      "loss": 0.2,
+      "step": 11822
+    },
+    {
+      "epoch": 1.8091813312930376,
+      "grad_norm": 2.313772731023759,
+      "learning_rate": 4.737079687039148e-07,
+      "loss": 0.3154,
+      "step": 11823
+    },
+    {
+      "epoch": 1.8093343534812547,
+      "grad_norm": 2.30936650960722,
+      "learning_rate": 4.7295456105725057e-07,
+      "loss": 0.2594,
+      "step": 11824
+    },
+    {
+      "epoch": 1.809487375669472,
+      "grad_norm": 2.0716315153696208,
+      "learning_rate": 4.7220173850371695e-07,
+      "loss": 0.2652,
+      "step": 11825
+    },
+    {
+      "epoch": 1.8096403978576894,
+      "grad_norm": 1.9868270558385652,
+      "learning_rate": 4.71449501089547e-07,
+      "loss": 0.2478,
+      "step": 11826
+    },
+    {
+      "epoch": 1.8097934200459067,
+      "grad_norm": 2.018269885599993,
+      "learning_rate": 4.706978488609393e-07,
+      "loss": 0.2329,
+      "step": 11827
+    },
+    {
+      "epoch": 1.809946442234124,
+      "grad_norm": 1.9684741598276265,
+      "learning_rate": 4.6994678186405685e-07,
+      "loss": 0.2501,
+      "step": 11828
+    },
+    {
+      "epoch": 1.8100994644223412,
+      "grad_norm": 1.933993539998423,
+      "learning_rate": 4.691963001450228e-07,
+      "loss": 0.2868,
+      "step": 11829
+    },
+    {
+      "epoch": 1.8102524866105585,
+      "grad_norm": 2.3696993136051736,
+      "learning_rate": 4.684464037499292e-07,
+      "loss": 0.3,
+      "step": 11830
+    },
+    {
+      "epoch": 1.8104055087987758,
+      "grad_norm": 2.2717448328519603,
+      "learning_rate": 4.676970927248292e-07,
+      "loss": 0.2647,
+      "step": 11831
+    },
+    {
+      "epoch": 1.8105585309869932,
+      "grad_norm": 2.005296412078862,
+      "learning_rate": 4.6694836711574153e-07,
+      "loss": 0.2521,
+      "step": 11832
+    },
+    {
+      "epoch": 1.8107115531752105,
+      "grad_norm": 1.752049954837269,
+      "learning_rate": 4.662002269686483e-07,
+      "loss": 0.2542,
+      "step": 11833
+    },
+    {
+      "epoch": 1.8108645753634276,
+      "grad_norm": 2.015368287875028,
+      "learning_rate": 4.654526723294961e-07,
+      "loss": 0.2845,
+      "step": 11834
+    },
+    {
+      "epoch": 1.8110175975516452,
+      "grad_norm": 2.137987633418499,
+      "learning_rate": 4.6470570324419374e-07,
+      "loss": 0.3099,
+      "step": 11835
+    },
+    {
+      "epoch": 1.8111706197398623,
+      "grad_norm": 2.2337190632103794,
+      "learning_rate": 4.6395931975861673e-07,
+      "loss": 0.268,
+      "step": 11836
+    },
+    {
+      "epoch": 1.8113236419280796,
+      "grad_norm": 2.0287393644206366,
+      "learning_rate": 4.6321352191860293e-07,
+      "loss": 0.2506,
+      "step": 11837
+    },
+    {
+      "epoch": 1.811476664116297,
+      "grad_norm": 2.0219857424282845,
+      "learning_rate": 4.6246830976995336e-07,
+      "loss": 0.4011,
+      "step": 11838
+    },
+    {
+      "epoch": 1.811629686304514,
+      "grad_norm": 2.4153059420724565,
+      "learning_rate": 4.617236833584393e-07,
+      "loss": 0.2586,
+      "step": 11839
+    },
+    {
+      "epoch": 1.8117827084927316,
+      "grad_norm": 2.1473162421467507,
+      "learning_rate": 4.6097964272978634e-07,
+      "loss": 0.2317,
+      "step": 11840
+    },
+    {
+      "epoch": 1.8119357306809487,
+      "grad_norm": 2.129796298413445,
+      "learning_rate": 4.6023618792968794e-07,
+      "loss": 0.2965,
+      "step": 11841
+    },
+    {
+      "epoch": 1.812088752869166,
+      "grad_norm": 2.186068024749278,
+      "learning_rate": 4.5949331900380865e-07,
+      "loss": 0.2816,
+      "step": 11842
+    },
+    {
+      "epoch": 1.8122417750573834,
+      "grad_norm": 2.0460758575279896,
+      "learning_rate": 4.5875103599776426e-07,
+      "loss": 0.2681,
+      "step": 11843
+    },
+    {
+      "epoch": 1.8123947972456005,
+      "grad_norm": 2.3159308326608423,
+      "learning_rate": 4.5800933895714606e-07,
+      "loss": 0.2898,
+      "step": 11844
+    },
+    {
+      "epoch": 1.812547819433818,
+      "grad_norm": 2.3681240793421754,
+      "learning_rate": 4.5726822792750425e-07,
+      "loss": 0.2783,
+      "step": 11845
+    },
+    {
+      "epoch": 1.8127008416220352,
+      "grad_norm": 2.2807501390604608,
+      "learning_rate": 4.565277029543491e-07,
+      "loss": 0.2452,
+      "step": 11846
+    },
+    {
+      "epoch": 1.8128538638102525,
+      "grad_norm": 2.0219388437967574,
+      "learning_rate": 4.5578776408316315e-07,
+      "loss": 0.233,
+      "step": 11847
+    },
+    {
+      "epoch": 1.8130068859984698,
+      "grad_norm": 1.944753820724033,
+      "learning_rate": 4.5504841135938893e-07,
+      "loss": 0.2404,
+      "step": 11848
+    },
+    {
+      "epoch": 1.813159908186687,
+      "grad_norm": 1.7793328730279039,
+      "learning_rate": 4.543096448284301e-07,
+      "loss": 0.2104,
+      "step": 11849
+    },
+    {
+      "epoch": 1.8133129303749045,
+      "grad_norm": 2.0493902412848954,
+      "learning_rate": 4.5357146453565926e-07,
+      "loss": 0.2623,
+      "step": 11850
+    },
+    {
+      "epoch": 1.8134659525631216,
+      "grad_norm": 1.8090947119690544,
+      "learning_rate": 4.5283387052641125e-07,
+      "loss": 0.1942,
+      "step": 11851
+    },
+    {
+      "epoch": 1.813618974751339,
+      "grad_norm": 2.0275436226010153,
+      "learning_rate": 4.520968628459832e-07,
+      "loss": 0.2955,
+      "step": 11852
+    },
+    {
+      "epoch": 1.8137719969395563,
+      "grad_norm": 2.4728132882804537,
+      "learning_rate": 4.5136044153963887e-07,
+      "loss": 0.3628,
+      "step": 11853
+    },
+    {
+      "epoch": 1.8139250191277734,
+      "grad_norm": 2.2678757394918745,
+      "learning_rate": 4.5062460665260434e-07,
+      "loss": 0.302,
+      "step": 11854
+    },
+    {
+      "epoch": 1.814078041315991,
+      "grad_norm": 1.9115212445776606,
+      "learning_rate": 4.498893582300701e-07,
+      "loss": 0.24,
+      "step": 11855
+    },
+    {
+      "epoch": 1.814231063504208,
+      "grad_norm": 1.9723840834983504,
+      "learning_rate": 4.491546963171911e-07,
+      "loss": 0.2545,
+      "step": 11856
+    },
+    {
+      "epoch": 1.8143840856924254,
+      "grad_norm": 1.9544300904005711,
+      "learning_rate": 4.4842062095908467e-07,
+      "loss": 0.2047,
+      "step": 11857
+    },
+    {
+      "epoch": 1.8145371078806427,
+      "grad_norm": 2.0727551080590167,
+      "learning_rate": 4.476871322008336e-07,
+      "loss": 0.2165,
+      "step": 11858
+    },
+    {
+      "epoch": 1.8146901300688598,
+      "grad_norm": 2.13104978350104,
+      "learning_rate": 4.4695423008748406e-07,
+      "loss": 0.3208,
+      "step": 11859
+    },
+    {
+      "epoch": 1.8148431522570774,
+      "grad_norm": 2.242696903489749,
+      "learning_rate": 4.4622191466404894e-07,
+      "loss": 0.3112,
+      "step": 11860
+    },
+    {
+      "epoch": 1.8149961744452945,
+      "grad_norm": 2.2981606041884333,
+      "learning_rate": 4.454901859755001e-07,
+      "loss": 0.3035,
+      "step": 11861
+    },
+    {
+      "epoch": 1.8151491966335118,
+      "grad_norm": 2.1319246519937507,
+      "learning_rate": 4.4475904406677483e-07,
+      "loss": 0.2548,
+      "step": 11862
+    },
+    {
+      "epoch": 1.8153022188217292,
+      "grad_norm": 2.152578096052732,
+      "learning_rate": 4.440284889827795e-07,
+      "loss": 0.2502,
+      "step": 11863
+    },
+    {
+      "epoch": 1.8154552410099465,
+      "grad_norm": 2.2940645041026757,
+      "learning_rate": 4.4329852076837597e-07,
+      "loss": 0.3296,
+      "step": 11864
+    },
+    {
+      "epoch": 1.8156082631981638,
+      "grad_norm": 2.199915066734886,
+      "learning_rate": 4.4256913946839506e-07,
+      "loss": 0.3083,
+      "step": 11865
+    },
+    {
+      "epoch": 1.815761285386381,
+      "grad_norm": 2.1312699700865285,
+      "learning_rate": 4.418403451276354e-07,
+      "loss": 0.343,
+      "step": 11866
+    },
+    {
+      "epoch": 1.8159143075745983,
+      "grad_norm": 2.1295718756882573,
+      "learning_rate": 4.411121377908489e-07,
+      "loss": 0.2527,
+      "step": 11867
+    },
+    {
+      "epoch": 1.8160673297628156,
+      "grad_norm": 2.0468010619606978,
+      "learning_rate": 4.4038451750276213e-07,
+      "loss": 0.2323,
+      "step": 11868
+    },
+    {
+      "epoch": 1.816220351951033,
+      "grad_norm": 2.037380646764395,
+      "learning_rate": 4.396574843080603e-07,
+      "loss": 0.3027,
+      "step": 11869
+    },
+    {
+      "epoch": 1.8163733741392503,
+      "grad_norm": 2.4229862444006147,
+      "learning_rate": 4.3893103825139115e-07,
+      "loss": 0.3327,
+      "step": 11870
+    },
+    {
+      "epoch": 1.8165263963274674,
+      "grad_norm": 2.2136068015881913,
+      "learning_rate": 4.3820517937737005e-07,
+      "loss": 0.3462,
+      "step": 11871
+    },
+    {
+      "epoch": 1.816679418515685,
+      "grad_norm": 2.2981031674405994,
+      "learning_rate": 4.3747990773057693e-07,
+      "loss": 0.3054,
+      "step": 11872
+    },
+    {
+      "epoch": 1.816832440703902,
+      "grad_norm": 2.050569164217128,
+      "learning_rate": 4.367552233555494e-07,
+      "loss": 0.2547,
+      "step": 11873
+    },
+    {
+      "epoch": 1.8169854628921194,
+      "grad_norm": 2.1447671798370855,
+      "learning_rate": 4.3603112629679534e-07,
+      "loss": 0.2512,
+      "step": 11874
+    },
+    {
+      "epoch": 1.8171384850803367,
+      "grad_norm": 2.229682016061022,
+      "learning_rate": 4.353076165987846e-07,
+      "loss": 0.2623,
+      "step": 11875
+    },
+    {
+      "epoch": 1.8172915072685538,
+      "grad_norm": 2.3664239681453663,
+      "learning_rate": 4.3458469430595063e-07,
+      "loss": 0.2964,
+      "step": 11876
+    },
+    {
+      "epoch": 1.8174445294567714,
+      "grad_norm": 2.0532090128786655,
+      "learning_rate": 4.338623594626912e-07,
+      "loss": 0.3125,
+      "step": 11877
+    },
+    {
+      "epoch": 1.8175975516449885,
+      "grad_norm": 1.9329806232041002,
+      "learning_rate": 4.3314061211336633e-07,
+      "loss": 0.2434,
+      "step": 11878
+    },
+    {
+      "epoch": 1.8177505738332058,
+      "grad_norm": 2.076618181262312,
+      "learning_rate": 4.3241945230230286e-07,
+      "loss": 0.2888,
+      "step": 11879
+    },
+    {
+      "epoch": 1.8179035960214232,
+      "grad_norm": 2.740737723939962,
+      "learning_rate": 4.316988800737887e-07,
+      "loss": 0.3424,
+      "step": 11880
+    },
+    {
+      "epoch": 1.8180566182096403,
+      "grad_norm": 2.0235334902556255,
+      "learning_rate": 4.309788954720784e-07,
+      "loss": 0.2841,
+      "step": 11881
+    },
+    {
+      "epoch": 1.8182096403978578,
+      "grad_norm": 1.8685472019677516,
+      "learning_rate": 4.302594985413877e-07,
+      "loss": 0.2695,
+      "step": 11882
+    },
+    {
+      "epoch": 1.818362662586075,
+      "grad_norm": 2.088752310739199,
+      "learning_rate": 4.295406893258991e-07,
+      "loss": 0.2887,
+      "step": 11883
+    },
+    {
+      "epoch": 1.8185156847742923,
+      "grad_norm": 2.1044895242632147,
+      "learning_rate": 4.2882246786975614e-07,
+      "loss": 0.285,
+      "step": 11884
+    },
+    {
+      "epoch": 1.8186687069625096,
+      "grad_norm": 1.986025076615623,
+      "learning_rate": 4.2810483421706796e-07,
+      "loss": 0.2584,
+      "step": 11885
+    },
+    {
+      "epoch": 1.8188217291507267,
+      "grad_norm": 2.493855607868984,
+      "learning_rate": 4.273877884119071e-07,
+      "loss": 0.3174,
+      "step": 11886
+    },
+    {
+      "epoch": 1.8189747513389443,
+      "grad_norm": 2.3424713449133083,
+      "learning_rate": 4.2667133049831166e-07,
+      "loss": 0.2856,
+      "step": 11887
+    },
+    {
+      "epoch": 1.8191277735271614,
+      "grad_norm": 2.159188268341222,
+      "learning_rate": 4.259554605202787e-07,
+      "loss": 0.295,
+      "step": 11888
+    },
+    {
+      "epoch": 1.8192807957153787,
+      "grad_norm": 1.9226041677235508,
+      "learning_rate": 4.2524017852177746e-07,
+      "loss": 0.2502,
+      "step": 11889
+    },
+    {
+      "epoch": 1.819433817903596,
+      "grad_norm": 2.0557886809068306,
+      "learning_rate": 4.245254845467317e-07,
+      "loss": 0.2254,
+      "step": 11890
+    },
+    {
+      "epoch": 1.8195868400918132,
+      "grad_norm": 2.2139733495822735,
+      "learning_rate": 4.238113786390352e-07,
+      "loss": 0.3166,
+      "step": 11891
+    },
+    {
+      "epoch": 1.8197398622800307,
+      "grad_norm": 1.8869752641971287,
+      "learning_rate": 4.230978608425462e-07,
+      "loss": 0.2612,
+      "step": 11892
+    },
+    {
+      "epoch": 1.8198928844682478,
+      "grad_norm": 2.361534050142558,
+      "learning_rate": 4.223849312010808e-07,
+      "loss": 0.2905,
+      "step": 11893
+    },
+    {
+      "epoch": 1.8200459066564652,
+      "grad_norm": 2.473559879765051,
+      "learning_rate": 4.2167258975842394e-07,
+      "loss": 0.2726,
+      "step": 11894
+    },
+    {
+      "epoch": 1.8201989288446825,
+      "grad_norm": 1.8092190590853954,
+      "learning_rate": 4.209608365583262e-07,
+      "loss": 0.2333,
+      "step": 11895
+    },
+    {
+      "epoch": 1.8203519510328998,
+      "grad_norm": 1.9546857994825566,
+      "learning_rate": 4.202496716444948e-07,
+      "loss": 0.221,
+      "step": 11896
+    },
+    {
+      "epoch": 1.8205049732211172,
+      "grad_norm": 2.0141397192525066,
+      "learning_rate": 4.195390950606082e-07,
+      "loss": 0.2491,
+      "step": 11897
+    },
+    {
+      "epoch": 1.8206579954093343,
+      "grad_norm": 2.1127973177494597,
+      "learning_rate": 4.1882910685030587e-07,
+      "loss": 0.2451,
+      "step": 11898
+    },
+    {
+      "epoch": 1.8208110175975516,
+      "grad_norm": 1.9673277852549478,
+      "learning_rate": 4.181197070571874e-07,
+      "loss": 0.2684,
+      "step": 11899
+    },
+    {
+      "epoch": 1.820964039785769,
+      "grad_norm": 2.358968570632875,
+      "learning_rate": 4.174108957248246e-07,
+      "loss": 0.2537,
+      "step": 11900
+    },
+    {
+      "epoch": 1.8211170619739863,
+      "grad_norm": 2.1309862940664765,
+      "learning_rate": 4.16702672896746e-07,
+      "loss": 0.3338,
+      "step": 11901
+    },
+    {
+      "epoch": 1.8212700841622036,
+      "grad_norm": 2.170521550489901,
+      "learning_rate": 4.1599503861644355e-07,
+      "loss": 0.2589,
+      "step": 11902
+    },
+    {
+      "epoch": 1.8214231063504207,
+      "grad_norm": 1.9231878873342532,
+      "learning_rate": 4.152879929273812e-07,
+      "loss": 0.1625,
+      "step": 11903
+    },
+    {
+      "epoch": 1.8215761285386383,
+      "grad_norm": 2.136045983086775,
+      "learning_rate": 4.145815358729777e-07,
+      "loss": 0.2598,
+      "step": 11904
+    },
+    {
+      "epoch": 1.8217291507268554,
+      "grad_norm": 2.0576412417232857,
+      "learning_rate": 4.1387566749662045e-07,
+      "loss": 0.2408,
+      "step": 11905
+    },
+    {
+      "epoch": 1.8218821729150727,
+      "grad_norm": 1.9763125602432094,
+      "learning_rate": 4.131703878416604e-07,
+      "loss": 0.2477,
+      "step": 11906
+    },
+    {
+      "epoch": 1.82203519510329,
+      "grad_norm": 2.3306832460828155,
+      "learning_rate": 4.124656969514107e-07,
+      "loss": 0.2894,
+      "step": 11907
+    },
+    {
+      "epoch": 1.8221882172915072,
+      "grad_norm": 2.175829010880975,
+      "learning_rate": 4.117615948691489e-07,
+      "loss": 0.2145,
+      "step": 11908
+    },
+    {
+      "epoch": 1.8223412394797247,
+      "grad_norm": 1.8907633407981779,
+      "learning_rate": 4.11058081638116e-07,
+      "loss": 0.2225,
+      "step": 11909
+    },
+    {
+      "epoch": 1.8224942616679418,
+      "grad_norm": 2.2137627125533275,
+      "learning_rate": 4.103551573015196e-07,
+      "loss": 0.2826,
+      "step": 11910
+    },
+    {
+      "epoch": 1.8226472838561592,
+      "grad_norm": 2.6138023580709238,
+      "learning_rate": 4.096528219025275e-07,
+      "loss": 0.3311,
+      "step": 11911
+    },
+    {
+      "epoch": 1.8228003060443765,
+      "grad_norm": 2.483862774064997,
+      "learning_rate": 4.0895107548427293e-07,
+      "loss": 0.3195,
+      "step": 11912
+    },
+    {
+      "epoch": 1.8229533282325936,
+      "grad_norm": 2.487626032754856,
+      "learning_rate": 4.082499180898536e-07,
+      "loss": 0.2799,
+      "step": 11913
+    },
+    {
+      "epoch": 1.8231063504208112,
+      "grad_norm": 2.1376354412056844,
+      "learning_rate": 4.0754934976233065e-07,
+      "loss": 0.2945,
+      "step": 11914
+    },
+    {
+      "epoch": 1.8232593726090283,
+      "grad_norm": 1.574613938962143,
+      "learning_rate": 4.068493705447274e-07,
+      "loss": 0.1875,
+      "step": 11915
+    },
+    {
+      "epoch": 1.8234123947972456,
+      "grad_norm": 2.1105301064666726,
+      "learning_rate": 4.0614998048003284e-07,
+      "loss": 0.2669,
+      "step": 11916
+    },
+    {
+      "epoch": 1.823565416985463,
+      "grad_norm": 2.268105261960538,
+      "learning_rate": 4.054511796111982e-07,
+      "loss": 0.2442,
+      "step": 11917
+    },
+    {
+      "epoch": 1.82371843917368,
+      "grad_norm": 2.281379829918933,
+      "learning_rate": 4.047529679811424e-07,
+      "loss": 0.2729,
+      "step": 11918
+    },
+    {
+      "epoch": 1.8238714613618976,
+      "grad_norm": 1.9162797641433544,
+      "learning_rate": 4.040553456327434e-07,
+      "loss": 0.2425,
+      "step": 11919
+    },
+    {
+      "epoch": 1.8240244835501147,
+      "grad_norm": 2.123847206220364,
+      "learning_rate": 4.0335831260884363e-07,
+      "loss": 0.256,
+      "step": 11920
+    },
+    {
+      "epoch": 1.824177505738332,
+      "grad_norm": 1.9678421561811497,
+      "learning_rate": 4.026618689522543e-07,
+      "loss": 0.255,
+      "step": 11921
+    },
+    {
+      "epoch": 1.8243305279265494,
+      "grad_norm": 2.1499690708656307,
+      "learning_rate": 4.019660147057436e-07,
+      "loss": 0.2687,
+      "step": 11922
+    },
+    {
+      "epoch": 1.8244835501147665,
+      "grad_norm": 1.839216684375747,
+      "learning_rate": 4.012707499120472e-07,
+      "loss": 0.2095,
+      "step": 11923
+    },
+    {
+      "epoch": 1.824636572302984,
+      "grad_norm": 1.8671693115789987,
+      "learning_rate": 4.005760746138654e-07,
+      "loss": 0.2566,
+      "step": 11924
+    },
+    {
+      "epoch": 1.8247895944912012,
+      "grad_norm": 2.007621689391616,
+      "learning_rate": 3.998819888538585e-07,
+      "loss": 0.2492,
+      "step": 11925
+    },
+    {
+      "epoch": 1.8249426166794185,
+      "grad_norm": 2.33315783039703,
+      "learning_rate": 3.9918849267465587e-07,
+      "loss": 0.2988,
+      "step": 11926
+    },
+    {
+      "epoch": 1.8250956388676358,
+      "grad_norm": 2.2086776410871973,
+      "learning_rate": 3.984955861188478e-07,
+      "loss": 0.2698,
+      "step": 11927
+    },
+    {
+      "epoch": 1.8252486610558531,
+      "grad_norm": 2.0325254374514365,
+      "learning_rate": 3.978032692289846e-07,
+      "loss": 0.2833,
+      "step": 11928
+    },
+    {
+      "epoch": 1.8254016832440705,
+      "grad_norm": 1.9777584449620282,
+      "learning_rate": 3.9711154204758686e-07,
+      "loss": 0.2598,
+      "step": 11929
+    },
+    {
+      "epoch": 1.8255547054322876,
+      "grad_norm": 2.0637922928930887,
+      "learning_rate": 3.964204046171383e-07,
+      "loss": 0.2631,
+      "step": 11930
+    },
+    {
+      "epoch": 1.825707727620505,
+      "grad_norm": 2.3317485527455073,
+      "learning_rate": 3.957298569800794e-07,
+      "loss": 0.2423,
+      "step": 11931
+    },
+    {
+      "epoch": 1.8258607498087223,
+      "grad_norm": 2.252331720437728,
+      "learning_rate": 3.950398991788229e-07,
+      "loss": 0.2706,
+      "step": 11932
+    },
+    {
+      "epoch": 1.8260137719969396,
+      "grad_norm": 2.163044989421106,
+      "learning_rate": 3.9435053125574164e-07,
+      "loss": 0.2988,
+      "step": 11933
+    },
+    {
+      "epoch": 1.826166794185157,
+      "grad_norm": 2.129347941639684,
+      "learning_rate": 3.936617532531717e-07,
+      "loss": 0.297,
+      "step": 11934
+    },
+    {
+      "epoch": 1.826319816373374,
+      "grad_norm": 2.0345026352262123,
+      "learning_rate": 3.929735652134137e-07,
+      "loss": 0.2762,
+      "step": 11935
+    },
+    {
+      "epoch": 1.8264728385615916,
+      "grad_norm": 2.0560842105913837,
+      "learning_rate": 3.922859671787316e-07,
+      "loss": 0.3026,
+      "step": 11936
+    },
+    {
+      "epoch": 1.8266258607498087,
+      "grad_norm": 1.9701130038274002,
+      "learning_rate": 3.915989591913538e-07,
+      "loss": 0.2759,
+      "step": 11937
+    },
+    {
+      "epoch": 1.826778882938026,
+      "grad_norm": 1.8164126230552349,
+      "learning_rate": 3.9091254129347225e-07,
+      "loss": 0.2115,
+      "step": 11938
+    },
+    {
+      "epoch": 1.8269319051262434,
+      "grad_norm": 2.095226323648143,
+      "learning_rate": 3.902267135272431e-07,
+      "loss": 0.2918,
+      "step": 11939
+    },
+    {
+      "epoch": 1.8270849273144605,
+      "grad_norm": 1.8632956629908928,
+      "learning_rate": 3.8954147593478486e-07,
+      "loss": 0.2635,
+      "step": 11940
+    },
+    {
+      "epoch": 1.827237949502678,
+      "grad_norm": 2.125554599686935,
+      "learning_rate": 3.888568285581795e-07,
+      "loss": 0.2541,
+      "step": 11941
+    },
+    {
+      "epoch": 1.8273909716908951,
+      "grad_norm": 1.8686548914535643,
+      "learning_rate": 3.8817277143947786e-07,
+      "loss": 0.2687,
+      "step": 11942
+    },
+    {
+      "epoch": 1.8275439938791125,
+      "grad_norm": 2.2724859585571004,
+      "learning_rate": 3.874893046206862e-07,
+      "loss": 0.2784,
+      "step": 11943
+    },
+    {
+      "epoch": 1.8276970160673298,
+      "grad_norm": 1.9277596501627852,
+      "learning_rate": 3.8680642814378e-07,
+      "loss": 0.2747,
+      "step": 11944
+    },
+    {
+      "epoch": 1.827850038255547,
+      "grad_norm": 1.8385682044969542,
+      "learning_rate": 3.861241420507e-07,
+      "loss": 0.2125,
+      "step": 11945
+    },
+    {
+      "epoch": 1.8280030604437645,
+      "grad_norm": 2.273734952165846,
+      "learning_rate": 3.8544244638334617e-07,
+      "loss": 0.2406,
+      "step": 11946
+    },
+    {
+      "epoch": 1.8281560826319816,
+      "grad_norm": 1.6741109276874935,
+      "learning_rate": 3.8476134118358153e-07,
+      "loss": 0.1967,
+      "step": 11947
+    },
+    {
+      "epoch": 1.828309104820199,
+      "grad_norm": 2.0544841920395363,
+      "learning_rate": 3.8408082649324165e-07,
+      "loss": 0.2309,
+      "step": 11948
+    },
+    {
+      "epoch": 1.8284621270084163,
+      "grad_norm": 2.120247464040846,
+      "learning_rate": 3.8340090235411186e-07,
+      "loss": 0.2576,
+      "step": 11949
+    },
+    {
+      "epoch": 1.8286151491966334,
+      "grad_norm": 2.19249778357952,
+      "learning_rate": 3.827215688079555e-07,
+      "loss": 0.2595,
+      "step": 11950
+    },
+    {
+      "epoch": 1.828768171384851,
+      "grad_norm": 2.0334839445427293,
+      "learning_rate": 3.8204282589649144e-07,
+      "loss": 0.3497,
+      "step": 11951
+    },
+    {
+      "epoch": 1.828921193573068,
+      "grad_norm": 2.414610411001209,
+      "learning_rate": 3.8136467366140073e-07,
+      "loss": 0.3199,
+      "step": 11952
+    },
+    {
+      "epoch": 1.8290742157612854,
+      "grad_norm": 2.293890111032897,
+      "learning_rate": 3.8068711214433453e-07,
+      "loss": 0.3205,
+      "step": 11953
+    },
+    {
+      "epoch": 1.8292272379495027,
+      "grad_norm": 2.141584926134872,
+      "learning_rate": 3.8001014138690396e-07,
+      "loss": 0.2612,
+      "step": 11954
+    },
+    {
+      "epoch": 1.8293802601377198,
+      "grad_norm": 1.9667031981212337,
+      "learning_rate": 3.793337614306847e-07,
+      "loss": 0.3036,
+      "step": 11955
+    },
+    {
+      "epoch": 1.8295332823259374,
+      "grad_norm": 1.8036232157888064,
+      "learning_rate": 3.7865797231721456e-07,
+      "loss": 0.2635,
+      "step": 11956
+    },
+    {
+      "epoch": 1.8296863045141545,
+      "grad_norm": 2.0793528801360943,
+      "learning_rate": 3.779827740879982e-07,
+      "loss": 0.2694,
+      "step": 11957
+    },
+    {
+      "epoch": 1.8298393267023718,
+      "grad_norm": 2.3880010286275724,
+      "learning_rate": 3.773081667845002e-07,
+      "loss": 0.26,
+      "step": 11958
+    },
+    {
+      "epoch": 1.8299923488905891,
+      "grad_norm": 1.922707375441064,
+      "learning_rate": 3.76634150448153e-07,
+      "loss": 0.2438,
+      "step": 11959
+    },
+    {
+      "epoch": 1.8301453710788065,
+      "grad_norm": 1.9779623539916547,
+      "learning_rate": 3.75960725120349e-07,
+      "loss": 0.2355,
+      "step": 11960
+    },
+    {
+      "epoch": 1.8302983932670238,
+      "grad_norm": 2.116293706736384,
+      "learning_rate": 3.752878908424462e-07,
+      "loss": 0.2952,
+      "step": 11961
+    },
+    {
+      "epoch": 1.830451415455241,
+      "grad_norm": 2.0257737983103787,
+      "learning_rate": 3.7461564765576607e-07,
+      "loss": 0.2281,
+      "step": 11962
+    },
+    {
+      "epoch": 1.8306044376434583,
+      "grad_norm": 1.8736255320167776,
+      "learning_rate": 3.7394399560159336e-07,
+      "loss": 0.2413,
+      "step": 11963
+    },
+    {
+      "epoch": 1.8307574598316756,
+      "grad_norm": 2.3071687308411466,
+      "learning_rate": 3.732729347211772e-07,
+      "loss": 0.2546,
+      "step": 11964
+    },
+    {
+      "epoch": 1.830910482019893,
+      "grad_norm": 1.873675215731886,
+      "learning_rate": 3.726024650557303e-07,
+      "loss": 0.2861,
+      "step": 11965
+    },
+    {
+      "epoch": 1.8310635042081103,
+      "grad_norm": 2.7959098918112555,
+      "learning_rate": 3.7193258664642964e-07,
+      "loss": 0.276,
+      "step": 11966
+    },
+    {
+      "epoch": 1.8312165263963274,
+      "grad_norm": 2.3147884301052275,
+      "learning_rate": 3.712632995344123e-07,
+      "loss": 0.2683,
+      "step": 11967
+    },
+    {
+      "epoch": 1.8313695485845447,
+      "grad_norm": 2.1285332905357786,
+      "learning_rate": 3.7059460376078547e-07,
+      "loss": 0.2911,
+      "step": 11968
+    },
+    {
+      "epoch": 1.831522570772762,
+      "grad_norm": 1.9785370093599328,
+      "learning_rate": 3.6992649936661294e-07,
+      "loss": 0.2751,
+      "step": 11969
+    },
+    {
+      "epoch": 1.8316755929609794,
+      "grad_norm": 1.9944491857282887,
+      "learning_rate": 3.6925898639292634e-07,
+      "loss": 0.2614,
+      "step": 11970
+    },
+    {
+      "epoch": 1.8318286151491967,
+      "grad_norm": 2.1796578318532,
+      "learning_rate": 3.6859206488072396e-07,
+      "loss": 0.3357,
+      "step": 11971
+    },
+    {
+      "epoch": 1.8319816373374138,
+      "grad_norm": 2.266227460375382,
+      "learning_rate": 3.6792573487095975e-07,
+      "loss": 0.2783,
+      "step": 11972
+    },
+    {
+      "epoch": 1.8321346595256314,
+      "grad_norm": 2.113008303916984,
+      "learning_rate": 3.6725999640455533e-07,
+      "loss": 0.3254,
+      "step": 11973
+    },
+    {
+      "epoch": 1.8322876817138485,
+      "grad_norm": 1.8368245644367691,
+      "learning_rate": 3.6659484952240033e-07,
+      "loss": 0.2533,
+      "step": 11974
+    },
+    {
+      "epoch": 1.8324407039020658,
+      "grad_norm": 1.9882890052049291,
+      "learning_rate": 3.6593029426534086e-07,
+      "loss": 0.2611,
+      "step": 11975
+    },
+    {
+      "epoch": 1.8325937260902831,
+      "grad_norm": 1.8442393407352757,
+      "learning_rate": 3.652663306741899e-07,
+      "loss": 0.2888,
+      "step": 11976
+    },
+    {
+      "epoch": 1.8327467482785003,
+      "grad_norm": 2.227356989444283,
+      "learning_rate": 3.6460295878972704e-07,
+      "loss": 0.3018,
+      "step": 11977
+    },
+    {
+      "epoch": 1.8328997704667178,
+      "grad_norm": 2.6299960205316566,
+      "learning_rate": 3.639401786526875e-07,
+      "loss": 0.2764,
+      "step": 11978
+    },
+    {
+      "epoch": 1.833052792654935,
+      "grad_norm": 2.3886068477380373,
+      "learning_rate": 3.6327799030377976e-07,
+      "loss": 0.282,
+      "step": 11979
+    },
+    {
+      "epoch": 1.8332058148431523,
+      "grad_norm": 2.3571824622687125,
+      "learning_rate": 3.6261639378367133e-07,
+      "loss": 0.2807,
+      "step": 11980
+    },
+    {
+      "epoch": 1.8333588370313696,
+      "grad_norm": 2.2983280032511173,
+      "learning_rate": 3.6195538913298856e-07,
+      "loss": 0.2866,
+      "step": 11981
+    },
+    {
+      "epoch": 1.8335118592195867,
+      "grad_norm": 2.280510188770605,
+      "learning_rate": 3.6129497639233123e-07,
+      "loss": 0.2887,
+      "step": 11982
+    },
+    {
+      "epoch": 1.8336648814078043,
+      "grad_norm": 2.2769950664737117,
+      "learning_rate": 3.606351556022558e-07,
+      "loss": 0.3011,
+      "step": 11983
+    },
+    {
+      "epoch": 1.8338179035960214,
+      "grad_norm": 2.183026594822403,
+      "learning_rate": 3.599759268032854e-07,
+      "loss": 0.2637,
+      "step": 11984
+    },
+    {
+      "epoch": 1.8339709257842387,
+      "grad_norm": 2.033810868967627,
+      "learning_rate": 3.5931729003590544e-07,
+      "loss": 0.2677,
+      "step": 11985
+    },
+    {
+      "epoch": 1.834123947972456,
+      "grad_norm": 2.0557534905143453,
+      "learning_rate": 3.586592453405646e-07,
+      "loss": 0.2727,
+      "step": 11986
+    },
+    {
+      "epoch": 1.8342769701606731,
+      "grad_norm": 2.005984349800885,
+      "learning_rate": 3.580017927576773e-07,
+      "loss": 0.3202,
+      "step": 11987
+    },
+    {
+      "epoch": 1.8344299923488907,
+      "grad_norm": 2.070018685158,
+      "learning_rate": 3.57344932327619e-07,
+      "loss": 0.2542,
+      "step": 11988
+    },
+    {
+      "epoch": 1.8345830145371078,
+      "grad_norm": 2.3376594691999077,
+      "learning_rate": 3.5668866409073075e-07,
+      "loss": 0.2996,
+      "step": 11989
+    },
+    {
+      "epoch": 1.8347360367253251,
+      "grad_norm": 2.441617675543389,
+      "learning_rate": 3.56032988087317e-07,
+      "loss": 0.3233,
+      "step": 11990
+    },
+    {
+      "epoch": 1.8348890589135425,
+      "grad_norm": 2.4082276906172493,
+      "learning_rate": 3.5537790435764443e-07,
+      "loss": 0.3318,
+      "step": 11991
+    },
+    {
+      "epoch": 1.8350420811017596,
+      "grad_norm": 2.585091922064814,
+      "learning_rate": 3.5472341294194413e-07,
+      "loss": 0.2845,
+      "step": 11992
+    },
+    {
+      "epoch": 1.8351951032899771,
+      "grad_norm": 2.2536038309694173,
+      "learning_rate": 3.540695138804129e-07,
+      "loss": 0.303,
+      "step": 11993
+    },
+    {
+      "epoch": 1.8353481254781943,
+      "grad_norm": 2.255993235887367,
+      "learning_rate": 3.5341620721320746e-07,
+      "loss": 0.2994,
+      "step": 11994
+    },
+    {
+      "epoch": 1.8355011476664116,
+      "grad_norm": 1.9197326776399797,
+      "learning_rate": 3.527634929804502e-07,
+      "loss": 0.2682,
+      "step": 11995
+    },
+    {
+      "epoch": 1.835654169854629,
+      "grad_norm": 2.0008368864449113,
+      "learning_rate": 3.521113712222268e-07,
+      "loss": 0.2804,
+      "step": 11996
+    },
+    {
+      "epoch": 1.8358071920428463,
+      "grad_norm": 1.7878197925898167,
+      "learning_rate": 3.514598419785875e-07,
+      "loss": 0.2248,
+      "step": 11997
+    },
+    {
+      "epoch": 1.8359602142310636,
+      "grad_norm": 2.0309007271436763,
+      "learning_rate": 3.508089052895436e-07,
+      "loss": 0.2969,
+      "step": 11998
+    },
+    {
+      "epoch": 1.8361132364192807,
+      "grad_norm": 1.9657444590479858,
+      "learning_rate": 3.50158561195072e-07,
+      "loss": 0.2682,
+      "step": 11999
+    },
+    {
+      "epoch": 1.836266258607498,
+      "grad_norm": 2.395003585976222,
+      "learning_rate": 3.4950880973511626e-07,
+      "loss": 0.3577,
+      "step": 12000
+    },
+    {
+      "epoch": 1.8364192807957154,
+      "grad_norm": 2.4444206059912723,
+      "learning_rate": 3.488596509495756e-07,
+      "loss": 0.5389,
+      "step": 12001
+    },
+    {
+      "epoch": 1.8365723029839327,
+      "grad_norm": 2.045241268445169,
+      "learning_rate": 3.4821108487831936e-07,
+      "loss": 0.2474,
+      "step": 12002
+    },
+    {
+      "epoch": 1.83672532517215,
+      "grad_norm": 2.0052203274872338,
+      "learning_rate": 3.4756311156117995e-07,
+      "loss": 0.2466,
+      "step": 12003
+    },
+    {
+      "epoch": 1.8368783473603671,
+      "grad_norm": 2.0755297021165817,
+      "learning_rate": 3.4691573103794894e-07,
+      "loss": 0.2683,
+      "step": 12004
+    },
+    {
+      "epoch": 1.8370313695485847,
+      "grad_norm": 2.2371449837594946,
+      "learning_rate": 3.4626894334838457e-07,
+      "loss": 0.3022,
+      "step": 12005
+    },
+    {
+      "epoch": 1.8371843917368018,
+      "grad_norm": 2.4836021577730865,
+      "learning_rate": 3.456227485322128e-07,
+      "loss": 0.3733,
+      "step": 12006
+    },
+    {
+      "epoch": 1.8373374139250191,
+      "grad_norm": 2.1804822352614504,
+      "learning_rate": 3.4497714662911406e-07,
+      "loss": 0.2799,
+      "step": 12007
+    },
+    {
+      "epoch": 1.8374904361132365,
+      "grad_norm": 2.1739341705368256,
+      "learning_rate": 3.443321376787401e-07,
+      "loss": 0.2481,
+      "step": 12008
+    },
+    {
+      "epoch": 1.8376434583014536,
+      "grad_norm": 1.9035207841071022,
+      "learning_rate": 3.436877217207046e-07,
+      "loss": 0.2046,
+      "step": 12009
+    },
+    {
+      "epoch": 1.8377964804896711,
+      "grad_norm": 2.5314565993420812,
+      "learning_rate": 3.4304389879457835e-07,
+      "loss": 0.2207,
+      "step": 12010
+    },
+    {
+      "epoch": 1.8379495026778883,
+      "grad_norm": 2.4787411745654127,
+      "learning_rate": 3.424006689399073e-07,
+      "loss": 0.3614,
+      "step": 12011
+    },
+    {
+      "epoch": 1.8381025248661056,
+      "grad_norm": 1.8064961409465559,
+      "learning_rate": 3.4175803219619217e-07,
+      "loss": 0.24,
+      "step": 12012
+    },
+    {
+      "epoch": 1.838255547054323,
+      "grad_norm": 2.2357013035541717,
+      "learning_rate": 3.4111598860289696e-07,
+      "loss": 0.2798,
+      "step": 12013
+    },
+    {
+      "epoch": 1.83840856924254,
+      "grad_norm": 2.1526836237095313,
+      "learning_rate": 3.4047453819945564e-07,
+      "loss": 0.2627,
+      "step": 12014
+    },
+    {
+      "epoch": 1.8385615914307576,
+      "grad_norm": 2.2397589758269048,
+      "learning_rate": 3.398336810252623e-07,
+      "loss": 0.3025,
+      "step": 12015
+    },
+    {
+      "epoch": 1.8387146136189747,
+      "grad_norm": 2.284066487090659,
+      "learning_rate": 3.39193417119672e-07,
+      "loss": 0.3526,
+      "step": 12016
+    },
+    {
+      "epoch": 1.838867635807192,
+      "grad_norm": 2.1539273397928618,
+      "learning_rate": 3.385537465220079e-07,
+      "loss": 0.3264,
+      "step": 12017
+    },
+    {
+      "epoch": 1.8390206579954094,
+      "grad_norm": 2.0446058854531945,
+      "learning_rate": 3.3791466927155403e-07,
+      "loss": 0.2517,
+      "step": 12018
+    },
+    {
+      "epoch": 1.8391736801836265,
+      "grad_norm": 1.8921978306564606,
+      "learning_rate": 3.37276185407559e-07,
+      "loss": 0.2678,
+      "step": 12019
+    },
+    {
+      "epoch": 1.839326702371844,
+      "grad_norm": 1.9626988958729294,
+      "learning_rate": 3.366382949692326e-07,
+      "loss": 0.2971,
+      "step": 12020
+    },
+    {
+      "epoch": 1.8394797245600611,
+      "grad_norm": 2.0565364790390217,
+      "learning_rate": 3.360009979957546e-07,
+      "loss": 0.2869,
+      "step": 12021
+    },
+    {
+      "epoch": 1.8396327467482785,
+      "grad_norm": 2.1341795864028397,
+      "learning_rate": 3.353642945262592e-07,
+      "loss": 0.2688,
+      "step": 12022
+    },
+    {
+      "epoch": 1.8397857689364958,
+      "grad_norm": 2.4168517975060135,
+      "learning_rate": 3.347281845998496e-07,
+      "loss": 0.3241,
+      "step": 12023
+    },
+    {
+      "epoch": 1.839938791124713,
+      "grad_norm": 2.3182262108339344,
+      "learning_rate": 3.340926682555956e-07,
+      "loss": 0.2966,
+      "step": 12024
+    },
+    {
+      "epoch": 1.8400918133129305,
+      "grad_norm": 2.116133961225955,
+      "learning_rate": 3.3345774553252273e-07,
+      "loss": 0.3218,
+      "step": 12025
+    },
+    {
+      "epoch": 1.8402448355011476,
+      "grad_norm": 2.1324931497746125,
+      "learning_rate": 3.328234164696242e-07,
+      "loss": 0.3341,
+      "step": 12026
+    },
+    {
+      "epoch": 1.840397857689365,
+      "grad_norm": 1.9944972529438794,
+      "learning_rate": 3.3218968110586114e-07,
+      "loss": 0.2519,
+      "step": 12027
+    },
+    {
+      "epoch": 1.8405508798775823,
+      "grad_norm": 1.9909357917518529,
+      "learning_rate": 3.3155653948014674e-07,
+      "loss": 0.2383,
+      "step": 12028
+    },
+    {
+      "epoch": 1.8407039020657996,
+      "grad_norm": 2.515119551859431,
+      "learning_rate": 3.3092399163137113e-07,
+      "loss": 0.3151,
+      "step": 12029
+    },
+    {
+      "epoch": 1.840856924254017,
+      "grad_norm": 2.3275887638623285,
+      "learning_rate": 3.302920375983787e-07,
+      "loss": 0.2848,
+      "step": 12030
+    },
+    {
+      "epoch": 1.841009946442234,
+      "grad_norm": 1.9495815422530196,
+      "learning_rate": 3.2966067741997844e-07,
+      "loss": 0.2069,
+      "step": 12031
+    },
+    {
+      "epoch": 1.8411629686304514,
+      "grad_norm": 1.9592391943120646,
+      "learning_rate": 3.290299111349471e-07,
+      "loss": 0.2031,
+      "step": 12032
+    },
+    {
+      "epoch": 1.8413159908186687,
+      "grad_norm": 2.4296933720566023,
+      "learning_rate": 3.283997387820248e-07,
+      "loss": 0.3,
+      "step": 12033
+    },
+    {
+      "epoch": 1.841469013006886,
+      "grad_norm": 2.056443705826707,
+      "learning_rate": 3.2777016039990615e-07,
+      "loss": 0.2354,
+      "step": 12034
+    },
+    {
+      "epoch": 1.8416220351951034,
+      "grad_norm": 2.4922123003136005,
+      "learning_rate": 3.2714117602726137e-07,
+      "loss": 0.2721,
+      "step": 12035
+    },
+    {
+      "epoch": 1.8417750573833205,
+      "grad_norm": 2.2539525516934087,
+      "learning_rate": 3.265127857027184e-07,
+      "loss": 0.2819,
+      "step": 12036
+    },
+    {
+      "epoch": 1.841928079571538,
+      "grad_norm": 2.3067629986694316,
+      "learning_rate": 3.2588498946486634e-07,
+      "loss": 0.2877,
+      "step": 12037
+    },
+    {
+      "epoch": 1.8420811017597551,
+      "grad_norm": 2.1901125523396123,
+      "learning_rate": 3.2525778735226444e-07,
+      "loss": 0.2798,
+      "step": 12038
+    },
+    {
+      "epoch": 1.8422341239479725,
+      "grad_norm": 2.3560154126153017,
+      "learning_rate": 3.2463117940342846e-07,
+      "loss": 0.3309,
+      "step": 12039
+    },
+    {
+      "epoch": 1.8423871461361898,
+      "grad_norm": 2.0695678919381315,
+      "learning_rate": 3.2400516565684217e-07,
+      "loss": 0.2819,
+      "step": 12040
+    },
+    {
+      "epoch": 1.842540168324407,
+      "grad_norm": 2.0646589283114603,
+      "learning_rate": 3.2337974615095134e-07,
+      "loss": 0.247,
+      "step": 12041
+    },
+    {
+      "epoch": 1.8426931905126245,
+      "grad_norm": 2.1398428401944214,
+      "learning_rate": 3.2275492092416536e-07,
+      "loss": 0.2873,
+      "step": 12042
+    },
+    {
+      "epoch": 1.8428462127008416,
+      "grad_norm": 2.289253406530182,
+      "learning_rate": 3.2213069001485797e-07,
+      "loss": 0.3365,
+      "step": 12043
+    },
+    {
+      "epoch": 1.842999234889059,
+      "grad_norm": 1.9642071977337394,
+      "learning_rate": 3.21507053461364e-07,
+      "loss": 0.2847,
+      "step": 12044
+    },
+    {
+      "epoch": 1.8431522570772763,
+      "grad_norm": 2.3167926844831674,
+      "learning_rate": 3.208840113019851e-07,
+      "loss": 0.2929,
+      "step": 12045
+    },
+    {
+      "epoch": 1.8433052792654934,
+      "grad_norm": 2.1338331581061585,
+      "learning_rate": 3.2026156357498284e-07,
+      "loss": 0.2996,
+      "step": 12046
+    },
+    {
+      "epoch": 1.843458301453711,
+      "grad_norm": 2.1716582602915993,
+      "learning_rate": 3.1963971031858664e-07,
+      "loss": 0.2735,
+      "step": 12047
+    },
+    {
+      "epoch": 1.843611323641928,
+      "grad_norm": 2.07267232853433,
+      "learning_rate": 3.1901845157098486e-07,
+      "loss": 0.2846,
+      "step": 12048
+    },
+    {
+      "epoch": 1.8437643458301454,
+      "grad_norm": 2.0413863077886303,
+      "learning_rate": 3.1839778737033256e-07,
+      "loss": 0.2788,
+      "step": 12049
+    },
+    {
+      "epoch": 1.8439173680183627,
+      "grad_norm": 2.132499243615788,
+      "learning_rate": 3.1777771775474697e-07,
+      "loss": 0.2613,
+      "step": 12050
+    },
+    {
+      "epoch": 1.8440703902065798,
+      "grad_norm": 2.012435426584495,
+      "learning_rate": 3.171582427623077e-07,
+      "loss": 0.2077,
+      "step": 12051
+    },
+    {
+      "epoch": 1.8442234123947974,
+      "grad_norm": 2.3338924545977537,
+      "learning_rate": 3.1653936243105974e-07,
+      "loss": 0.2969,
+      "step": 12052
+    },
+    {
+      "epoch": 1.8443764345830145,
+      "grad_norm": 2.439473361933731,
+      "learning_rate": 3.1592107679901393e-07,
+      "loss": 0.3477,
+      "step": 12053
+    },
+    {
+      "epoch": 1.8445294567712318,
+      "grad_norm": 1.9503200571517338,
+      "learning_rate": 3.1530338590413877e-07,
+      "loss": 0.247,
+      "step": 12054
+    },
+    {
+      "epoch": 1.8446824789594491,
+      "grad_norm": 1.9160994751283733,
+      "learning_rate": 3.1468628978436723e-07,
+      "loss": 0.2587,
+      "step": 12055
+    },
+    {
+      "epoch": 1.8448355011476663,
+      "grad_norm": 2.2006778123960355,
+      "learning_rate": 3.1406978847760226e-07,
+      "loss": 0.2572,
+      "step": 12056
+    },
+    {
+      "epoch": 1.8449885233358838,
+      "grad_norm": 2.298106954995458,
+      "learning_rate": 3.1345388202170034e-07,
+      "loss": 0.339,
+      "step": 12057
+    },
+    {
+      "epoch": 1.845141545524101,
+      "grad_norm": 2.2870522685072068,
+      "learning_rate": 3.128385704544912e-07,
+      "loss": 0.2774,
+      "step": 12058
+    },
+    {
+      "epoch": 1.8452945677123183,
+      "grad_norm": 2.0815855675565844,
+      "learning_rate": 3.122238538137634e-07,
+      "loss": 0.2914,
+      "step": 12059
+    },
+    {
+      "epoch": 1.8454475899005356,
+      "grad_norm": 2.0949484272761567,
+      "learning_rate": 3.1160973213726466e-07,
+      "loss": 0.2838,
+      "step": 12060
+    },
+    {
+      "epoch": 1.845600612088753,
+      "grad_norm": 2.1367734350362744,
+      "learning_rate": 3.109962054627147e-07,
+      "loss": 0.2684,
+      "step": 12061
+    },
+    {
+      "epoch": 1.8457536342769703,
+      "grad_norm": 1.9203254480433196,
+      "learning_rate": 3.103832738277923e-07,
+      "loss": 0.2774,
+      "step": 12062
+    },
+    {
+      "epoch": 1.8459066564651874,
+      "grad_norm": 2.2174394172339484,
+      "learning_rate": 3.097709372701374e-07,
+      "loss": 0.3001,
+      "step": 12063
+    },
+    {
+      "epoch": 1.8460596786534047,
+      "grad_norm": 2.07364457958336,
+      "learning_rate": 3.0915919582735763e-07,
+      "loss": 0.2439,
+      "step": 12064
+    },
+    {
+      "epoch": 1.846212700841622,
+      "grad_norm": 2.4434845521963062,
+      "learning_rate": 3.085480495370241e-07,
+      "loss": 0.2569,
+      "step": 12065
+    },
+    {
+      "epoch": 1.8463657230298394,
+      "grad_norm": 2.155863352925435,
+      "learning_rate": 3.079374984366668e-07,
+      "loss": 0.3176,
+      "step": 12066
+    },
+    {
+      "epoch": 1.8465187452180567,
+      "grad_norm": 1.9860984079286215,
+      "learning_rate": 3.073275425637834e-07,
+      "loss": 0.2281,
+      "step": 12067
+    },
+    {
+      "epoch": 1.8466717674062738,
+      "grad_norm": 2.4004615498420803,
+      "learning_rate": 3.06718181955834e-07,
+      "loss": 0.2703,
+      "step": 12068
+    },
+    {
+      "epoch": 1.8468247895944914,
+      "grad_norm": 2.2840856637864255,
+      "learning_rate": 3.0610941665024097e-07,
+      "loss": 0.272,
+      "step": 12069
+    },
+    {
+      "epoch": 1.8469778117827085,
+      "grad_norm": 2.089748940200058,
+      "learning_rate": 3.0550124668439097e-07,
+      "loss": 0.2438,
+      "step": 12070
+    },
+    {
+      "epoch": 1.8471308339709258,
+      "grad_norm": 2.219418314806367,
+      "learning_rate": 3.0489367209563527e-07,
+      "loss": 0.2898,
+      "step": 12071
+    },
+    {
+      "epoch": 1.8472838561591431,
+      "grad_norm": 2.060183624027669,
+      "learning_rate": 3.042866929212862e-07,
+      "loss": 0.2405,
+      "step": 12072
+    },
+    {
+      "epoch": 1.8474368783473603,
+      "grad_norm": 2.010860439464339,
+      "learning_rate": 3.036803091986218e-07,
+      "loss": 0.2258,
+      "step": 12073
+    },
+    {
+      "epoch": 1.8475899005355778,
+      "grad_norm": 1.9942119712715967,
+      "learning_rate": 3.030745209648811e-07,
+      "loss": 0.3044,
+      "step": 12074
+    },
+    {
+      "epoch": 1.847742922723795,
+      "grad_norm": 2.232052662763698,
+      "learning_rate": 3.024693282572688e-07,
+      "loss": 0.285,
+      "step": 12075
+    },
+    {
+      "epoch": 1.8478959449120123,
+      "grad_norm": 2.0825006725065585,
+      "learning_rate": 3.01864731112953e-07,
+      "loss": 0.2628,
+      "step": 12076
+    },
+    {
+      "epoch": 1.8480489671002296,
+      "grad_norm": 1.9978220340953112,
+      "learning_rate": 3.012607295690617e-07,
+      "loss": 0.1955,
+      "step": 12077
+    },
+    {
+      "epoch": 1.8482019892884467,
+      "grad_norm": 2.0714850256582724,
+      "learning_rate": 3.0065732366269197e-07,
+      "loss": 0.2598,
+      "step": 12078
+    },
+    {
+      "epoch": 1.8483550114766643,
+      "grad_norm": 1.9595160794383788,
+      "learning_rate": 3.0005451343089964e-07,
+      "loss": 0.213,
+      "step": 12079
+    },
+    {
+      "epoch": 1.8485080336648814,
+      "grad_norm": 2.179653007063922,
+      "learning_rate": 2.9945229891070624e-07,
+      "loss": 0.2505,
+      "step": 12080
+    },
+    {
+      "epoch": 1.8486610558530987,
+      "grad_norm": 2.0967176533438967,
+      "learning_rate": 2.9885068013909444e-07,
+      "loss": 0.2826,
+      "step": 12081
+    },
+    {
+      "epoch": 1.848814078041316,
+      "grad_norm": 1.8612577442029008,
+      "learning_rate": 2.982496571530158e-07,
+      "loss": 0.2543,
+      "step": 12082
+    },
+    {
+      "epoch": 1.8489671002295331,
+      "grad_norm": 2.3138064490032746,
+      "learning_rate": 2.976492299893774e-07,
+      "loss": 0.2842,
+      "step": 12083
+    },
+    {
+      "epoch": 1.8491201224177507,
+      "grad_norm": 2.426532374935647,
+      "learning_rate": 2.9704939868505533e-07,
+      "loss": 0.3227,
+      "step": 12084
+    },
+    {
+      "epoch": 1.8492731446059678,
+      "grad_norm": 1.9737808547398248,
+      "learning_rate": 2.9645016327689014e-07,
+      "loss": 0.3011,
+      "step": 12085
+    },
+    {
+      "epoch": 1.8494261667941851,
+      "grad_norm": 1.9982381311792963,
+      "learning_rate": 2.9585152380167906e-07,
+      "loss": 0.2686,
+      "step": 12086
+    },
+    {
+      "epoch": 1.8495791889824025,
+      "grad_norm": 1.985073678732461,
+      "learning_rate": 2.952534802961882e-07,
+      "loss": 0.2569,
+      "step": 12087
+    },
+    {
+      "epoch": 1.8497322111706196,
+      "grad_norm": 1.9793562344285673,
+      "learning_rate": 2.946560327971493e-07,
+      "loss": 0.2316,
+      "step": 12088
+    },
+    {
+      "epoch": 1.8498852333588371,
+      "grad_norm": 2.424299187829567,
+      "learning_rate": 2.9405918134124747e-07,
+      "loss": 0.2653,
+      "step": 12089
+    },
+    {
+      "epoch": 1.8500382555470543,
+      "grad_norm": 2.119611836762099,
+      "learning_rate": 2.9346292596514227e-07,
+      "loss": 0.2722,
+      "step": 12090
+    },
+    {
+      "epoch": 1.8501912777352716,
+      "grad_norm": 2.386035985833903,
+      "learning_rate": 2.928672667054533e-07,
+      "loss": 0.3155,
+      "step": 12091
+    },
+    {
+      "epoch": 1.850344299923489,
+      "grad_norm": 1.9398662606503192,
+      "learning_rate": 2.9227220359875685e-07,
+      "loss": 0.279,
+      "step": 12092
+    },
+    {
+      "epoch": 1.850497322111706,
+      "grad_norm": 2.569255627334306,
+      "learning_rate": 2.916777366816026e-07,
+      "loss": 0.2903,
+      "step": 12093
+    },
+    {
+      "epoch": 1.8506503442999236,
+      "grad_norm": 2.268604135723616,
+      "learning_rate": 2.9108386599049685e-07,
+      "loss": 0.2867,
+      "step": 12094
+    },
+    {
+      "epoch": 1.8508033664881407,
+      "grad_norm": 2.3828483785992374,
+      "learning_rate": 2.9049059156191385e-07,
+      "loss": 0.3234,
+      "step": 12095
+    },
+    {
+      "epoch": 1.850956388676358,
+      "grad_norm": 1.988877704957895,
+      "learning_rate": 2.8989791343228657e-07,
+      "loss": 0.2306,
+      "step": 12096
+    },
+    {
+      "epoch": 1.8511094108645754,
+      "grad_norm": 2.2265877027141383,
+      "learning_rate": 2.8930583163801487e-07,
+      "loss": 0.2704,
+      "step": 12097
+    },
+    {
+      "epoch": 1.8512624330527927,
+      "grad_norm": 2.0913282906056505,
+      "learning_rate": 2.887143462154596e-07,
+      "loss": 0.271,
+      "step": 12098
+    },
+    {
+      "epoch": 1.85141545524101,
+      "grad_norm": 2.172619994500396,
+      "learning_rate": 2.8812345720094836e-07,
+      "loss": 0.3294,
+      "step": 12099
+    },
+    {
+      "epoch": 1.8515684774292271,
+      "grad_norm": 2.136696908463675,
+      "learning_rate": 2.875331646307677e-07,
+      "loss": 0.2473,
+      "step": 12100
+    },
+    {
+      "epoch": 1.8517214996174445,
+      "grad_norm": 2.074684483541853,
+      "learning_rate": 2.869434685411709e-07,
+      "loss": 0.2527,
+      "step": 12101
+    },
+    {
+      "epoch": 1.8518745218056618,
+      "grad_norm": 2.133340003665217,
+      "learning_rate": 2.863543689683734e-07,
+      "loss": 0.2935,
+      "step": 12102
+    },
+    {
+      "epoch": 1.8520275439938791,
+      "grad_norm": 1.951720876586633,
+      "learning_rate": 2.8576586594855517e-07,
+      "loss": 0.2645,
+      "step": 12103
+    },
+    {
+      "epoch": 1.8521805661820965,
+      "grad_norm": 2.147938107651279,
+      "learning_rate": 2.851779595178561e-07,
+      "loss": 0.2582,
+      "step": 12104
+    },
+    {
+      "epoch": 1.8523335883703136,
+      "grad_norm": 2.226493399132335,
+      "learning_rate": 2.8459064971238293e-07,
+      "loss": 0.2827,
+      "step": 12105
+    },
+    {
+      "epoch": 1.8524866105585311,
+      "grad_norm": 2.2685272163662655,
+      "learning_rate": 2.840039365682068e-07,
+      "loss": 0.3398,
+      "step": 12106
+    },
+    {
+      "epoch": 1.8526396327467483,
+      "grad_norm": 2.313428395457457,
+      "learning_rate": 2.8341782012135665e-07,
+      "loss": 0.2826,
+      "step": 12107
+    },
+    {
+      "epoch": 1.8527926549349656,
+      "grad_norm": 2.3340885234134943,
+      "learning_rate": 2.828323004078293e-07,
+      "loss": 0.2838,
+      "step": 12108
+    },
+    {
+      "epoch": 1.852945677123183,
+      "grad_norm": 1.784707937660604,
+      "learning_rate": 2.822473774635859e-07,
+      "loss": 0.2615,
+      "step": 12109
+    },
+    {
+      "epoch": 1.8530986993114,
+      "grad_norm": 1.9472104360411793,
+      "learning_rate": 2.8166305132454443e-07,
+      "loss": 0.2362,
+      "step": 12110
+    },
+    {
+      "epoch": 1.8532517214996176,
+      "grad_norm": 1.9800530360384525,
+      "learning_rate": 2.8107932202659504e-07,
+      "loss": 0.298,
+      "step": 12111
+    },
+    {
+      "epoch": 1.8534047436878347,
+      "grad_norm": 2.2978407370753695,
+      "learning_rate": 2.804961896055858e-07,
+      "loss": 0.325,
+      "step": 12112
+    },
+    {
+      "epoch": 1.853557765876052,
+      "grad_norm": 2.141324858232483,
+      "learning_rate": 2.7991365409732683e-07,
+      "loss": 0.2936,
+      "step": 12113
+    },
+    {
+      "epoch": 1.8537107880642694,
+      "grad_norm": 2.481993377344709,
+      "learning_rate": 2.7933171553759633e-07,
+      "loss": 0.3219,
+      "step": 12114
+    },
+    {
+      "epoch": 1.8538638102524865,
+      "grad_norm": 1.9074014250506492,
+      "learning_rate": 2.7875037396213335e-07,
+      "loss": 0.2674,
+      "step": 12115
+    },
+    {
+      "epoch": 1.854016832440704,
+      "grad_norm": 2.168713968336492,
+      "learning_rate": 2.7816962940663827e-07,
+      "loss": 0.2711,
+      "step": 12116
+    },
+    {
+      "epoch": 1.8541698546289211,
+      "grad_norm": 2.495180968740349,
+      "learning_rate": 2.7758948190677927e-07,
+      "loss": 0.3058,
+      "step": 12117
+    },
+    {
+      "epoch": 1.8543228768171385,
+      "grad_norm": 2.281648475117201,
+      "learning_rate": 2.7700993149818336e-07,
+      "loss": 0.2417,
+      "step": 12118
+    },
+    {
+      "epoch": 1.8544758990053558,
+      "grad_norm": 2.151961070143828,
+      "learning_rate": 2.764309782164454e-07,
+      "loss": 0.2624,
+      "step": 12119
+    },
+    {
+      "epoch": 1.854628921193573,
+      "grad_norm": 1.9512571981380695,
+      "learning_rate": 2.7585262209711916e-07,
+      "loss": 0.2198,
+      "step": 12120
+    },
+    {
+      "epoch": 1.8547819433817905,
+      "grad_norm": 2.072196470323757,
+      "learning_rate": 2.7527486317572515e-07,
+      "loss": 0.281,
+      "step": 12121
+    },
+    {
+      "epoch": 1.8549349655700076,
+      "grad_norm": 2.157528977725661,
+      "learning_rate": 2.746977014877439e-07,
+      "loss": 0.3241,
+      "step": 12122
+    },
+    {
+      "epoch": 1.855087987758225,
+      "grad_norm": 2.1224216420455,
+      "learning_rate": 2.7412113706862366e-07,
+      "loss": 0.2503,
+      "step": 12123
+    },
+    {
+      "epoch": 1.8552410099464423,
+      "grad_norm": 1.9186758271968076,
+      "learning_rate": 2.7354516995377165e-07,
+      "loss": 0.228,
+      "step": 12124
+    },
+    {
+      "epoch": 1.8553940321346594,
+      "grad_norm": 2.2800359932130942,
+      "learning_rate": 2.729698001785608e-07,
+      "loss": 0.2943,
+      "step": 12125
+    },
+    {
+      "epoch": 1.855547054322877,
+      "grad_norm": 2.138629020001359,
+      "learning_rate": 2.723950277783272e-07,
+      "loss": 0.2682,
+      "step": 12126
+    },
+    {
+      "epoch": 1.855700076511094,
+      "grad_norm": 2.059609498153823,
+      "learning_rate": 2.718208527883692e-07,
+      "loss": 0.2549,
+      "step": 12127
+    },
+    {
+      "epoch": 1.8558530986993114,
+      "grad_norm": 2.0336964526176806,
+      "learning_rate": 2.712472752439499e-07,
+      "loss": 0.2152,
+      "step": 12128
+    },
+    {
+      "epoch": 1.8560061208875287,
+      "grad_norm": 2.575196197121381,
+      "learning_rate": 2.706742951802943e-07,
+      "loss": 0.2578,
+      "step": 12129
+    },
+    {
+      "epoch": 1.856159143075746,
+      "grad_norm": 2.0950964294914676,
+      "learning_rate": 2.701019126325921e-07,
+      "loss": 0.2927,
+      "step": 12130
+    },
+    {
+      "epoch": 1.8563121652639634,
+      "grad_norm": 1.9792281674828598,
+      "learning_rate": 2.6953012763599293e-07,
+      "loss": 0.2385,
+      "step": 12131
+    },
+    {
+      "epoch": 1.8564651874521805,
+      "grad_norm": 2.45643230103461,
+      "learning_rate": 2.689589402256165e-07,
+      "loss": 0.3375,
+      "step": 12132
+    },
+    {
+      "epoch": 1.8566182096403978,
+      "grad_norm": 2.1003291220472655,
+      "learning_rate": 2.683883504365392e-07,
+      "loss": 0.2646,
+      "step": 12133
+    },
+    {
+      "epoch": 1.8567712318286151,
+      "grad_norm": 2.4439436285141083,
+      "learning_rate": 2.67818358303803e-07,
+      "loss": 0.3014,
+      "step": 12134
+    },
+    {
+      "epoch": 1.8569242540168325,
+      "grad_norm": 2.0784054999384582,
+      "learning_rate": 2.672489638624154e-07,
+      "loss": 0.2691,
+      "step": 12135
+    },
+    {
+      "epoch": 1.8570772762050498,
+      "grad_norm": 2.3457672844778803,
+      "learning_rate": 2.666801671473429e-07,
+      "loss": 0.2712,
+      "step": 12136
+    },
+    {
+      "epoch": 1.857230298393267,
+      "grad_norm": 2.2535684260502205,
+      "learning_rate": 2.6611196819351647e-07,
+      "loss": 0.2782,
+      "step": 12137
+    },
+    {
+      "epoch": 1.8573833205814845,
+      "grad_norm": 2.4045532560845486,
+      "learning_rate": 2.6554436703583595e-07,
+      "loss": 0.2972,
+      "step": 12138
+    },
+    {
+      "epoch": 1.8575363427697016,
+      "grad_norm": 2.6298059779139744,
+      "learning_rate": 2.6497736370915573e-07,
+      "loss": 0.2809,
+      "step": 12139
+    },
+    {
+      "epoch": 1.857689364957919,
+      "grad_norm": 2.332091100831672,
+      "learning_rate": 2.64410958248299e-07,
+      "loss": 0.2753,
+      "step": 12140
+    },
+    {
+      "epoch": 1.8578423871461363,
+      "grad_norm": 1.7920366458419108,
+      "learning_rate": 2.638451506880524e-07,
+      "loss": 0.2298,
+      "step": 12141
+    },
+    {
+      "epoch": 1.8579954093343534,
+      "grad_norm": 1.9857680513176472,
+      "learning_rate": 2.6327994106316146e-07,
+      "loss": 0.2223,
+      "step": 12142
+    },
+    {
+      "epoch": 1.858148431522571,
+      "grad_norm": 2.1848014681855763,
+      "learning_rate": 2.627153294083418e-07,
+      "loss": 0.2969,
+      "step": 12143
+    },
+    {
+      "epoch": 1.858301453710788,
+      "grad_norm": 1.9778201877077595,
+      "learning_rate": 2.621513157582656e-07,
+      "loss": 0.2965,
+      "step": 12144
+    },
+    {
+      "epoch": 1.8584544758990054,
+      "grad_norm": 2.229623695748292,
+      "learning_rate": 2.615879001475707e-07,
+      "loss": 0.2964,
+      "step": 12145
+    },
+    {
+      "epoch": 1.8586074980872227,
+      "grad_norm": 2.183425999238914,
+      "learning_rate": 2.6102508261085957e-07,
+      "loss": 0.3343,
+      "step": 12146
+    },
+    {
+      "epoch": 1.8587605202754398,
+      "grad_norm": 2.206930477766042,
+      "learning_rate": 2.6046286318269775e-07,
+      "loss": 0.2877,
+      "step": 12147
+    },
+    {
+      "epoch": 1.8589135424636574,
+      "grad_norm": 3.0938588466275148,
+      "learning_rate": 2.5990124189761325e-07,
+      "loss": 0.3017,
+      "step": 12148
+    },
+    {
+      "epoch": 1.8590665646518745,
+      "grad_norm": 2.059332216618004,
+      "learning_rate": 2.593402187900973e-07,
+      "loss": 0.2688,
+      "step": 12149
+    },
+    {
+      "epoch": 1.8592195868400918,
+      "grad_norm": 2.1161232443504594,
+      "learning_rate": 2.587797938946035e-07,
+      "loss": 0.3114,
+      "step": 12150
+    },
+    {
+      "epoch": 1.8593726090283091,
+      "grad_norm": 2.0546549960609592,
+      "learning_rate": 2.5821996724554985e-07,
+      "loss": 0.2921,
+      "step": 12151
+    },
+    {
+      "epoch": 1.8595256312165263,
+      "grad_norm": 2.1037161550199492,
+      "learning_rate": 2.576607388773189e-07,
+      "loss": 0.2607,
+      "step": 12152
+    },
+    {
+      "epoch": 1.8596786534047438,
+      "grad_norm": 2.0870667813048684,
+      "learning_rate": 2.5710210882425424e-07,
+      "loss": 0.2594,
+      "step": 12153
+    },
+    {
+      "epoch": 1.859831675592961,
+      "grad_norm": 2.2012957148422396,
+      "learning_rate": 2.5654407712066287e-07,
+      "loss": 0.2806,
+      "step": 12154
+    },
+    {
+      "epoch": 1.8599846977811783,
+      "grad_norm": 2.1918892229646465,
+      "learning_rate": 2.559866438008174e-07,
+      "loss": 0.3092,
+      "step": 12155
+    },
+    {
+      "epoch": 1.8601377199693956,
+      "grad_norm": 2.2243548563257796,
+      "learning_rate": 2.554298088989504e-07,
+      "loss": 0.3184,
+      "step": 12156
+    },
+    {
+      "epoch": 1.8602907421576127,
+      "grad_norm": 2.3448776203557977,
+      "learning_rate": 2.54873572449259e-07,
+      "loss": 0.3585,
+      "step": 12157
+    },
+    {
+      "epoch": 1.8604437643458303,
+      "grad_norm": 2.063362643976679,
+      "learning_rate": 2.543179344859059e-07,
+      "loss": 0.2759,
+      "step": 12158
+    },
+    {
+      "epoch": 1.8605967865340474,
+      "grad_norm": 2.336058681945544,
+      "learning_rate": 2.5376289504301375e-07,
+      "loss": 0.3135,
+      "step": 12159
+    },
+    {
+      "epoch": 1.8607498087222647,
+      "grad_norm": 2.312197431308048,
+      "learning_rate": 2.5320845415466756e-07,
+      "loss": 0.2921,
+      "step": 12160
+    },
+    {
+      "epoch": 1.860902830910482,
+      "grad_norm": 2.0693817907130034,
+      "learning_rate": 2.5265461185492224e-07,
+      "loss": 0.2645,
+      "step": 12161
+    },
+    {
+      "epoch": 1.8610558530986994,
+      "grad_norm": 2.0482766219818296,
+      "learning_rate": 2.5210136817778843e-07,
+      "loss": 0.2503,
+      "step": 12162
+    },
+    {
+      "epoch": 1.8612088752869167,
+      "grad_norm": 2.391935771524626,
+      "learning_rate": 2.5154872315724333e-07,
+      "loss": 0.3501,
+      "step": 12163
+    },
+    {
+      "epoch": 1.8613618974751338,
+      "grad_norm": 2.236266464810737,
+      "learning_rate": 2.509966768272276e-07,
+      "loss": 0.2932,
+      "step": 12164
+    },
+    {
+      "epoch": 1.8615149196633511,
+      "grad_norm": 2.310119778373762,
+      "learning_rate": 2.5044522922164414e-07,
+      "loss": 0.3187,
+      "step": 12165
+    },
+    {
+      "epoch": 1.8616679418515685,
+      "grad_norm": 2.126907185113331,
+      "learning_rate": 2.498943803743592e-07,
+      "loss": 0.2129,
+      "step": 12166
+    },
+    {
+      "epoch": 1.8618209640397858,
+      "grad_norm": 2.048088304206397,
+      "learning_rate": 2.4934413031920456e-07,
+      "loss": 0.2373,
+      "step": 12167
+    },
+    {
+      "epoch": 1.8619739862280031,
+      "grad_norm": 1.9929259125015373,
+      "learning_rate": 2.487944790899688e-07,
+      "loss": 0.2448,
+      "step": 12168
+    },
+    {
+      "epoch": 1.8621270084162203,
+      "grad_norm": 1.934843136968899,
+      "learning_rate": 2.4824542672041263e-07,
+      "loss": 0.25,
+      "step": 12169
+    },
+    {
+      "epoch": 1.8622800306044378,
+      "grad_norm": 1.8995690929106952,
+      "learning_rate": 2.4769697324425577e-07,
+      "loss": 0.2292,
+      "step": 12170
+    },
+    {
+      "epoch": 1.862433052792655,
+      "grad_norm": 2.1530208013880423,
+      "learning_rate": 2.471491186951758e-07,
+      "loss": 0.2673,
+      "step": 12171
+    },
+    {
+      "epoch": 1.8625860749808723,
+      "grad_norm": 1.8712526951486876,
+      "learning_rate": 2.4660186310682343e-07,
+      "loss": 0.2554,
+      "step": 12172
+    },
+    {
+      "epoch": 1.8627390971690896,
+      "grad_norm": 2.2666491985656885,
+      "learning_rate": 2.4605520651280634e-07,
+      "loss": 0.314,
+      "step": 12173
+    },
+    {
+      "epoch": 1.8628921193573067,
+      "grad_norm": 1.7644271265516238,
+      "learning_rate": 2.4550914894669544e-07,
+      "loss": 0.2522,
+      "step": 12174
+    },
+    {
+      "epoch": 1.8630451415455243,
+      "grad_norm": 2.1397863046954284,
+      "learning_rate": 2.4496369044202826e-07,
+      "loss": 0.3065,
+      "step": 12175
+    },
+    {
+      "epoch": 1.8631981637337414,
+      "grad_norm": 2.2356898789015585,
+      "learning_rate": 2.4441883103230256e-07,
+      "loss": 0.2685,
+      "step": 12176
+    },
+    {
+      "epoch": 1.8633511859219587,
+      "grad_norm": 2.227907810018491,
+      "learning_rate": 2.438745707509815e-07,
+      "loss": 0.3018,
+      "step": 12177
+    },
+    {
+      "epoch": 1.863504208110176,
+      "grad_norm": 2.307269133388442,
+      "learning_rate": 2.4333090963148843e-07,
+      "loss": 0.3108,
+      "step": 12178
+    },
+    {
+      "epoch": 1.8636572302983931,
+      "grad_norm": 2.2926041689812564,
+      "learning_rate": 2.427878477072121e-07,
+      "loss": 0.2948,
+      "step": 12179
+    },
+    {
+      "epoch": 1.8638102524866107,
+      "grad_norm": 2.2200065269776497,
+      "learning_rate": 2.4224538501150475e-07,
+      "loss": 0.2883,
+      "step": 12180
+    },
+    {
+      "epoch": 1.8639632746748278,
+      "grad_norm": 2.0355521766689733,
+      "learning_rate": 2.417035215776808e-07,
+      "loss": 0.2665,
+      "step": 12181
+    },
+    {
+      "epoch": 1.8641162968630451,
+      "grad_norm": 2.0094946627575867,
+      "learning_rate": 2.4116225743901935e-07,
+      "loss": 0.2702,
+      "step": 12182
+    },
+    {
+      "epoch": 1.8642693190512625,
+      "grad_norm": 2.172054629406562,
+      "learning_rate": 2.4062159262875917e-07,
+      "loss": 0.3103,
+      "step": 12183
+    },
+    {
+      "epoch": 1.8644223412394796,
+      "grad_norm": 2.2125272166267274,
+      "learning_rate": 2.4008152718010493e-07,
+      "loss": 0.2704,
+      "step": 12184
+    },
+    {
+      "epoch": 1.8645753634276971,
+      "grad_norm": 1.9717482284027534,
+      "learning_rate": 2.395420611262278e-07,
+      "loss": 0.2178,
+      "step": 12185
+    },
+    {
+      "epoch": 1.8647283856159143,
+      "grad_norm": 2.256922690173856,
+      "learning_rate": 2.3900319450025464e-07,
+      "loss": 0.2931,
+      "step": 12186
+    },
+    {
+      "epoch": 1.8648814078041316,
+      "grad_norm": 2.739839348840143,
+      "learning_rate": 2.3846492733527905e-07,
+      "loss": 0.2836,
+      "step": 12187
+    },
+    {
+      "epoch": 1.865034429992349,
+      "grad_norm": 1.700510993453522,
+      "learning_rate": 2.3792725966436224e-07,
+      "loss": 0.2278,
+      "step": 12188
+    },
+    {
+      "epoch": 1.865187452180566,
+      "grad_norm": 1.9545979397503868,
+      "learning_rate": 2.3739019152052013e-07,
+      "loss": 0.2552,
+      "step": 12189
+    },
+    {
+      "epoch": 1.8653404743687836,
+      "grad_norm": 2.195267570630192,
+      "learning_rate": 2.3685372293673737e-07,
+      "loss": 0.2948,
+      "step": 12190
+    },
+    {
+      "epoch": 1.8654934965570007,
+      "grad_norm": 1.8382213012672446,
+      "learning_rate": 2.3631785394596317e-07,
+      "loss": 0.2203,
+      "step": 12191
+    },
+    {
+      "epoch": 1.865646518745218,
+      "grad_norm": 1.8451009051414737,
+      "learning_rate": 2.3578258458110347e-07,
+      "loss": 0.2316,
+      "step": 12192
+    },
+    {
+      "epoch": 1.8657995409334354,
+      "grad_norm": 2.3210095540011624,
+      "learning_rate": 2.3524791487503418e-07,
+      "loss": 0.2854,
+      "step": 12193
+    },
+    {
+      "epoch": 1.8659525631216525,
+      "grad_norm": 2.146736780352666,
+      "learning_rate": 2.3471384486059123e-07,
+      "loss": 0.2932,
+      "step": 12194
+    },
+    {
+      "epoch": 1.86610558530987,
+      "grad_norm": 1.9464980080620138,
+      "learning_rate": 2.3418037457057063e-07,
+      "loss": 0.278,
+      "step": 12195
+    },
+    {
+      "epoch": 1.8662586074980871,
+      "grad_norm": 2.066963030143947,
+      "learning_rate": 2.336475040377384e-07,
+      "loss": 0.2585,
+      "step": 12196
+    },
+    {
+      "epoch": 1.8664116296863045,
+      "grad_norm": 2.311400102076601,
+      "learning_rate": 2.3311523329481943e-07,
+      "loss": 0.2815,
+      "step": 12197
+    },
+    {
+      "epoch": 1.8665646518745218,
+      "grad_norm": 2.067821743921238,
+      "learning_rate": 2.3258356237450208e-07,
+      "loss": 0.2692,
+      "step": 12198
+    },
+    {
+      "epoch": 1.8667176740627391,
+      "grad_norm": 3.1778578650717884,
+      "learning_rate": 2.3205249130943908e-07,
+      "loss": 0.3116,
+      "step": 12199
+    },
+    {
+      "epoch": 1.8668706962509565,
+      "grad_norm": 2.2301021426943506,
+      "learning_rate": 2.3152202013224434e-07,
+      "loss": 0.2504,
+      "step": 12200
+    },
+    {
+      "epoch": 1.8670237184391736,
+      "grad_norm": 1.783192454316723,
+      "learning_rate": 2.3099214887549626e-07,
+      "loss": 0.1713,
+      "step": 12201
+    },
+    {
+      "epoch": 1.867176740627391,
+      "grad_norm": 2.0357835170166374,
+      "learning_rate": 2.3046287757173768e-07,
+      "loss": 0.2838,
+      "step": 12202
+    },
+    {
+      "epoch": 1.8673297628156083,
+      "grad_norm": 1.9008329193246671,
+      "learning_rate": 2.2993420625347263e-07,
+      "loss": 0.2363,
+      "step": 12203
+    },
+    {
+      "epoch": 1.8674827850038256,
+      "grad_norm": 2.164894992849952,
+      "learning_rate": 2.2940613495316844e-07,
+      "loss": 0.2903,
+      "step": 12204
+    },
+    {
+      "epoch": 1.867635807192043,
+      "grad_norm": 1.818748508477644,
+      "learning_rate": 2.2887866370325696e-07,
+      "loss": 0.1966,
+      "step": 12205
+    },
+    {
+      "epoch": 1.86778882938026,
+      "grad_norm": 2.5998851775941416,
+      "learning_rate": 2.2835179253613005e-07,
+      "loss": 0.2729,
+      "step": 12206
+    },
+    {
+      "epoch": 1.8679418515684776,
+      "grad_norm": 2.2297654340062523,
+      "learning_rate": 2.2782552148414738e-07,
+      "loss": 0.3441,
+      "step": 12207
+    },
+    {
+      "epoch": 1.8680948737566947,
+      "grad_norm": 2.147215993679389,
+      "learning_rate": 2.2729985057962866e-07,
+      "loss": 0.2219,
+      "step": 12208
+    },
+    {
+      "epoch": 1.868247895944912,
+      "grad_norm": 2.015334317509236,
+      "learning_rate": 2.2677477985485697e-07,
+      "loss": 0.2214,
+      "step": 12209
+    },
+    {
+      "epoch": 1.8684009181331294,
+      "grad_norm": 2.0696416363691137,
+      "learning_rate": 2.262503093420787e-07,
+      "loss": 0.2825,
+      "step": 12210
+    },
+    {
+      "epoch": 1.8685539403213465,
+      "grad_norm": 2.0538848905826192,
+      "learning_rate": 2.2572643907350477e-07,
+      "loss": 0.2514,
+      "step": 12211
+    },
+    {
+      "epoch": 1.868706962509564,
+      "grad_norm": 2.0587824354921422,
+      "learning_rate": 2.2520316908130612e-07,
+      "loss": 0.2369,
+      "step": 12212
+    },
+    {
+      "epoch": 1.8688599846977811,
+      "grad_norm": 2.2401753137094067,
+      "learning_rate": 2.2468049939762038e-07,
+      "loss": 0.3055,
+      "step": 12213
+    },
+    {
+      "epoch": 1.8690130068859985,
+      "grad_norm": 1.9765528689673304,
+      "learning_rate": 2.241584300545485e-07,
+      "loss": 0.2514,
+      "step": 12214
+    },
+    {
+      "epoch": 1.8691660290742158,
+      "grad_norm": 2.16610698984525,
+      "learning_rate": 2.2363696108414822e-07,
+      "loss": 0.2618,
+      "step": 12215
+    },
+    {
+      "epoch": 1.869319051262433,
+      "grad_norm": 2.272639274412879,
+      "learning_rate": 2.2311609251844834e-07,
+      "loss": 0.3037,
+      "step": 12216
+    },
+    {
+      "epoch": 1.8694720734506505,
+      "grad_norm": 2.2152036642185946,
+      "learning_rate": 2.2259582438943773e-07,
+      "loss": 0.3251,
+      "step": 12217
+    },
+    {
+      "epoch": 1.8696250956388676,
+      "grad_norm": 2.0843874546884247,
+      "learning_rate": 2.2207615672906523e-07,
+      "loss": 0.2705,
+      "step": 12218
+    },
+    {
+      "epoch": 1.869778117827085,
+      "grad_norm": 2.192065193696456,
+      "learning_rate": 2.215570895692476e-07,
+      "loss": 0.274,
+      "step": 12219
+    },
+    {
+      "epoch": 1.8699311400153023,
+      "grad_norm": 2.252514725359488,
+      "learning_rate": 2.2103862294186374e-07,
+      "loss": 0.3029,
+      "step": 12220
+    },
+    {
+      "epoch": 1.8700841622035194,
+      "grad_norm": 1.7304474836249897,
+      "learning_rate": 2.2052075687875262e-07,
+      "loss": 0.2527,
+      "step": 12221
+    },
+    {
+      "epoch": 1.870237184391737,
+      "grad_norm": 2.319453397662655,
+      "learning_rate": 2.2000349141171995e-07,
+      "loss": 0.3471,
+      "step": 12222
+    },
+    {
+      "epoch": 1.870390206579954,
+      "grad_norm": 2.096261818076652,
+      "learning_rate": 2.194868265725325e-07,
+      "loss": 0.2351,
+      "step": 12223
+    },
+    {
+      "epoch": 1.8705432287681714,
+      "grad_norm": 1.7582270248063583,
+      "learning_rate": 2.1897076239291936e-07,
+      "loss": 0.3032,
+      "step": 12224
+    },
+    {
+      "epoch": 1.8706962509563887,
+      "grad_norm": 2.144383125659378,
+      "learning_rate": 2.184552989045763e-07,
+      "loss": 0.2726,
+      "step": 12225
+    },
+    {
+      "epoch": 1.8708492731446058,
+      "grad_norm": 4.532884435368886,
+      "learning_rate": 2.1794043613916015e-07,
+      "loss": 0.3022,
+      "step": 12226
+    },
+    {
+      "epoch": 1.8710022953328234,
+      "grad_norm": 2.559387386062656,
+      "learning_rate": 2.1742617412828682e-07,
+      "loss": 0.3715,
+      "step": 12227
+    },
+    {
+      "epoch": 1.8711553175210405,
+      "grad_norm": 2.162713822623037,
+      "learning_rate": 2.169125129035432e-07,
+      "loss": 0.2815,
+      "step": 12228
+    },
+    {
+      "epoch": 1.8713083397092578,
+      "grad_norm": 2.2515483930461526,
+      "learning_rate": 2.163994524964741e-07,
+      "loss": 0.3593,
+      "step": 12229
+    },
+    {
+      "epoch": 1.8714613618974751,
+      "grad_norm": 1.8777623529142353,
+      "learning_rate": 2.158869929385876e-07,
+      "loss": 0.21,
+      "step": 12230
+    },
+    {
+      "epoch": 1.8716143840856925,
+      "grad_norm": 2.0469775737816946,
+      "learning_rate": 2.1537513426135637e-07,
+      "loss": 0.2926,
+      "step": 12231
+    },
+    {
+      "epoch": 1.8717674062739098,
+      "grad_norm": 2.303687292526856,
+      "learning_rate": 2.1486387649621632e-07,
+      "loss": 0.324,
+      "step": 12232
+    },
+    {
+      "epoch": 1.871920428462127,
+      "grad_norm": 2.3834650957610792,
+      "learning_rate": 2.1435321967456567e-07,
+      "loss": 0.3276,
+      "step": 12233
+    },
+    {
+      "epoch": 1.8720734506503443,
+      "grad_norm": 2.2343221661881434,
+      "learning_rate": 2.1384316382776493e-07,
+      "loss": 0.3089,
+      "step": 12234
+    },
+    {
+      "epoch": 1.8722264728385616,
+      "grad_norm": 1.8172686458735308,
+      "learning_rate": 2.1333370898713902e-07,
+      "loss": 0.2393,
+      "step": 12235
+    },
+    {
+      "epoch": 1.872379495026779,
+      "grad_norm": 2.608854617407368,
+      "learning_rate": 2.1282485518397622e-07,
+      "loss": 0.3307,
+      "step": 12236
+    },
+    {
+      "epoch": 1.8725325172149963,
+      "grad_norm": 2.50025253282384,
+      "learning_rate": 2.1231660244952713e-07,
+      "loss": 0.2851,
+      "step": 12237
+    },
+    {
+      "epoch": 1.8726855394032134,
+      "grad_norm": 2.0332244215536095,
+      "learning_rate": 2.1180895081500453e-07,
+      "loss": 0.2816,
+      "step": 12238
+    },
+    {
+      "epoch": 1.872838561591431,
+      "grad_norm": 1.8344196972742979,
+      "learning_rate": 2.113019003115857e-07,
+      "loss": 0.2178,
+      "step": 12239
+    },
+    {
+      "epoch": 1.872991583779648,
+      "grad_norm": 2.1325275388576395,
+      "learning_rate": 2.107954509704102e-07,
+      "loss": 0.2919,
+      "step": 12240
+    },
+    {
+      "epoch": 1.8731446059678654,
+      "grad_norm": 2.03238396828965,
+      "learning_rate": 2.1028960282258204e-07,
+      "loss": 0.312,
+      "step": 12241
+    },
+    {
+      "epoch": 1.8732976281560827,
+      "grad_norm": 2.134605895045237,
+      "learning_rate": 2.0978435589916635e-07,
+      "loss": 0.2619,
+      "step": 12242
+    },
+    {
+      "epoch": 1.8734506503442998,
+      "grad_norm": 2.0790319221111764,
+      "learning_rate": 2.0927971023119498e-07,
+      "loss": 0.2806,
+      "step": 12243
+    },
+    {
+      "epoch": 1.8736036725325174,
+      "grad_norm": 2.1151547837451323,
+      "learning_rate": 2.0877566584965646e-07,
+      "loss": 0.2672,
+      "step": 12244
+    },
+    {
+      "epoch": 1.8737566947207345,
+      "grad_norm": 1.8800490838528072,
+      "learning_rate": 2.082722227855083e-07,
+      "loss": 0.2575,
+      "step": 12245
+    },
+    {
+      "epoch": 1.8739097169089518,
+      "grad_norm": 1.9156558849897791,
+      "learning_rate": 2.0776938106966903e-07,
+      "loss": 0.2848,
+      "step": 12246
+    },
+    {
+      "epoch": 1.8740627390971691,
+      "grad_norm": 2.1518764006308797,
+      "learning_rate": 2.0726714073301845e-07,
+      "loss": 0.278,
+      "step": 12247
+    },
+    {
+      "epoch": 1.8742157612853863,
+      "grad_norm": 2.115861447241316,
+      "learning_rate": 2.0676550180640187e-07,
+      "loss": 0.3185,
+      "step": 12248
+    },
+    {
+      "epoch": 1.8743687834736038,
+      "grad_norm": 2.426652595668358,
+      "learning_rate": 2.0626446432062798e-07,
+      "loss": 0.2452,
+      "step": 12249
+    },
+    {
+      "epoch": 1.874521805661821,
+      "grad_norm": 2.011114310109736,
+      "learning_rate": 2.0576402830646548e-07,
+      "loss": 0.2554,
+      "step": 12250
+    },
+    {
+      "epoch": 1.8746748278500383,
+      "grad_norm": 2.441039647615812,
+      "learning_rate": 2.052641937946509e-07,
+      "loss": 0.3226,
+      "step": 12251
+    },
+    {
+      "epoch": 1.8748278500382556,
+      "grad_norm": 2.2657634566140827,
+      "learning_rate": 2.0476496081587972e-07,
+      "loss": 0.2532,
+      "step": 12252
+    },
+    {
+      "epoch": 1.8749808722264727,
+      "grad_norm": 2.2793594074333727,
+      "learning_rate": 2.042663294008096e-07,
+      "loss": 0.3178,
+      "step": 12253
+    },
+    {
+      "epoch": 1.8751338944146902,
+      "grad_norm": 2.3583753493859168,
+      "learning_rate": 2.0376829958006606e-07,
+      "loss": 0.3027,
+      "step": 12254
+    },
+    {
+      "epoch": 1.8752869166029074,
+      "grad_norm": 2.3976898464771317,
+      "learning_rate": 2.0327087138423464e-07,
+      "loss": 0.3205,
+      "step": 12255
+    },
+    {
+      "epoch": 1.8754399387911247,
+      "grad_norm": 2.3833629536154044,
+      "learning_rate": 2.027740448438631e-07,
+      "loss": 0.2334,
+      "step": 12256
+    },
+    {
+      "epoch": 1.875592960979342,
+      "grad_norm": 2.5591541641816455,
+      "learning_rate": 2.0227781998946483e-07,
+      "loss": 0.2747,
+      "step": 12257
+    },
+    {
+      "epoch": 1.8757459831675591,
+      "grad_norm": 1.9460841714107107,
+      "learning_rate": 2.0178219685151544e-07,
+      "loss": 0.2954,
+      "step": 12258
+    },
+    {
+      "epoch": 1.8758990053557767,
+      "grad_norm": 2.130399857557917,
+      "learning_rate": 2.0128717546045174e-07,
+      "loss": 0.3298,
+      "step": 12259
+    },
+    {
+      "epoch": 1.8760520275439938,
+      "grad_norm": 2.054983131339847,
+      "learning_rate": 2.007927558466749e-07,
+      "loss": 0.2231,
+      "step": 12260
+    },
+    {
+      "epoch": 1.8762050497322111,
+      "grad_norm": 1.9846915653480426,
+      "learning_rate": 2.0029893804054956e-07,
+      "loss": 0.2199,
+      "step": 12261
+    },
+    {
+      "epoch": 1.8763580719204285,
+      "grad_norm": 2.1665049130134575,
+      "learning_rate": 1.9980572207240367e-07,
+      "loss": 0.3055,
+      "step": 12262
+    },
+    {
+      "epoch": 1.8765110941086458,
+      "grad_norm": 2.0578045313911857,
+      "learning_rate": 1.9931310797252635e-07,
+      "loss": 0.2991,
+      "step": 12263
+    },
+    {
+      "epoch": 1.8766641162968631,
+      "grad_norm": 1.7469439315722723,
+      "learning_rate": 1.9882109577117337e-07,
+      "loss": 0.2242,
+      "step": 12264
+    },
+    {
+      "epoch": 1.8768171384850802,
+      "grad_norm": 2.174488307923911,
+      "learning_rate": 1.9832968549855724e-07,
+      "loss": 0.3399,
+      "step": 12265
+    },
+    {
+      "epoch": 1.8769701606732976,
+      "grad_norm": 1.947263660138171,
+      "learning_rate": 1.978388771848594e-07,
+      "loss": 0.238,
+      "step": 12266
+    },
+    {
+      "epoch": 1.877123182861515,
+      "grad_norm": 2.155219460753691,
+      "learning_rate": 1.9734867086022458e-07,
+      "loss": 0.2746,
+      "step": 12267
+    },
+    {
+      "epoch": 1.8772762050497322,
+      "grad_norm": 2.042073767668411,
+      "learning_rate": 1.9685906655475428e-07,
+      "loss": 0.2658,
+      "step": 12268
+    },
+    {
+      "epoch": 1.8774292272379496,
+      "grad_norm": 1.8416151846022213,
+      "learning_rate": 1.9637006429851778e-07,
+      "loss": 0.214,
+      "step": 12269
+    },
+    {
+      "epoch": 1.8775822494261667,
+      "grad_norm": 2.3539053420151492,
+      "learning_rate": 1.9588166412154997e-07,
+      "loss": 0.3265,
+      "step": 12270
+    },
+    {
+      "epoch": 1.8777352716143842,
+      "grad_norm": 2.4376515005557637,
+      "learning_rate": 1.9539386605384125e-07,
+      "loss": 0.2822,
+      "step": 12271
+    },
+    {
+      "epoch": 1.8778882938026014,
+      "grad_norm": 1.8886652401346042,
+      "learning_rate": 1.9490667012535215e-07,
+      "loss": 0.2249,
+      "step": 12272
+    },
+    {
+      "epoch": 1.8780413159908187,
+      "grad_norm": 2.329875267873853,
+      "learning_rate": 1.9442007636600312e-07,
+      "loss": 0.2711,
+      "step": 12273
+    },
+    {
+      "epoch": 1.878194338179036,
+      "grad_norm": 2.189273001021051,
+      "learning_rate": 1.9393408480567477e-07,
+      "loss": 0.2499,
+      "step": 12274
+    },
+    {
+      "epoch": 1.8783473603672531,
+      "grad_norm": 2.439968634578106,
+      "learning_rate": 1.934486954742165e-07,
+      "loss": 0.3041,
+      "step": 12275
+    },
+    {
+      "epoch": 1.8785003825554707,
+      "grad_norm": 1.9757854141283482,
+      "learning_rate": 1.9296390840143897e-07,
+      "loss": 0.2711,
+      "step": 12276
+    },
+    {
+      "epoch": 1.8786534047436878,
+      "grad_norm": 2.0689238149257476,
+      "learning_rate": 1.9247972361711053e-07,
+      "loss": 0.2598,
+      "step": 12277
+    },
+    {
+      "epoch": 1.8788064269319051,
+      "grad_norm": 2.109148148749415,
+      "learning_rate": 1.9199614115097076e-07,
+      "loss": 0.2545,
+      "step": 12278
+    },
+    {
+      "epoch": 1.8789594491201225,
+      "grad_norm": 2.4297663317572438,
+      "learning_rate": 1.91513161032717e-07,
+      "loss": 0.2645,
+      "step": 12279
+    },
+    {
+      "epoch": 1.8791124713083396,
+      "grad_norm": 2.1144787392422755,
+      "learning_rate": 1.9103078329201108e-07,
+      "loss": 0.2495,
+      "step": 12280
+    },
+    {
+      "epoch": 1.8792654934965571,
+      "grad_norm": 2.559522126569114,
+      "learning_rate": 1.905490079584782e-07,
+      "loss": 0.2995,
+      "step": 12281
+    },
+    {
+      "epoch": 1.8794185156847742,
+      "grad_norm": 2.186023542033774,
+      "learning_rate": 1.9006783506170467e-07,
+      "loss": 0.288,
+      "step": 12282
+    },
+    {
+      "epoch": 1.8795715378729916,
+      "grad_norm": 2.3879072385137405,
+      "learning_rate": 1.8958726463124243e-07,
+      "loss": 0.28,
+      "step": 12283
+    },
+    {
+      "epoch": 1.879724560061209,
+      "grad_norm": 2.342145125582227,
+      "learning_rate": 1.8910729669660454e-07,
+      "loss": 0.2903,
+      "step": 12284
+    },
+    {
+      "epoch": 1.879877582249426,
+      "grad_norm": 2.2675058883591985,
+      "learning_rate": 1.8862793128726853e-07,
+      "loss": 0.2908,
+      "step": 12285
+    },
+    {
+      "epoch": 1.8800306044376436,
+      "grad_norm": 2.1204252540865367,
+      "learning_rate": 1.8814916843267306e-07,
+      "loss": 0.3024,
+      "step": 12286
+    },
+    {
+      "epoch": 1.8801836266258607,
+      "grad_norm": 2.1179211787734027,
+      "learning_rate": 1.876710081622224e-07,
+      "loss": 0.2748,
+      "step": 12287
+    },
+    {
+      "epoch": 1.880336648814078,
+      "grad_norm": 2.3750767038120775,
+      "learning_rate": 1.871934505052797e-07,
+      "loss": 0.3196,
+      "step": 12288
+    },
+    {
+      "epoch": 1.8804896710022954,
+      "grad_norm": 2.156315708260614,
+      "learning_rate": 1.86716495491176e-07,
+      "loss": 0.2977,
+      "step": 12289
+    },
+    {
+      "epoch": 1.8806426931905125,
+      "grad_norm": 2.0621837306649367,
+      "learning_rate": 1.8624014314920114e-07,
+      "loss": 0.3236,
+      "step": 12290
+    },
+    {
+      "epoch": 1.88079571537873,
+      "grad_norm": 1.917989006367677,
+      "learning_rate": 1.8576439350861175e-07,
+      "loss": 0.2343,
+      "step": 12291
+    },
+    {
+      "epoch": 1.8809487375669471,
+      "grad_norm": 1.9783581142507576,
+      "learning_rate": 1.8528924659862447e-07,
+      "loss": 0.2416,
+      "step": 12292
+    },
+    {
+      "epoch": 1.8811017597551645,
+      "grad_norm": 2.0896830236966557,
+      "learning_rate": 1.8481470244841925e-07,
+      "loss": 0.2807,
+      "step": 12293
+    },
+    {
+      "epoch": 1.8812547819433818,
+      "grad_norm": 2.09622278574327,
+      "learning_rate": 1.843407610871406e-07,
+      "loss": 0.2848,
+      "step": 12294
+    },
+    {
+      "epoch": 1.881407804131599,
+      "grad_norm": 1.9323158480708773,
+      "learning_rate": 1.838674225438941e-07,
+      "loss": 0.2593,
+      "step": 12295
+    },
+    {
+      "epoch": 1.8815608263198165,
+      "grad_norm": 2.2310584238434843,
+      "learning_rate": 1.8339468684775209e-07,
+      "loss": 0.2542,
+      "step": 12296
+    },
+    {
+      "epoch": 1.8817138485080336,
+      "grad_norm": 2.0500256612721874,
+      "learning_rate": 1.8292255402774462e-07,
+      "loss": 0.2342,
+      "step": 12297
+    },
+    {
+      "epoch": 1.881866870696251,
+      "grad_norm": 2.115013908907418,
+      "learning_rate": 1.824510241128663e-07,
+      "loss": 0.2748,
+      "step": 12298
+    },
+    {
+      "epoch": 1.8820198928844682,
+      "grad_norm": 1.907028475113633,
+      "learning_rate": 1.8198009713207843e-07,
+      "loss": 0.2571,
+      "step": 12299
+    },
+    {
+      "epoch": 1.8821729150726856,
+      "grad_norm": 2.071888069560488,
+      "learning_rate": 1.8150977311430007e-07,
+      "loss": 0.3025,
+      "step": 12300
+    },
+    {
+      "epoch": 1.882325937260903,
+      "grad_norm": 1.9734825940094098,
+      "learning_rate": 1.8104005208841702e-07,
+      "loss": 0.296,
+      "step": 12301
+    },
+    {
+      "epoch": 1.88247895944912,
+      "grad_norm": 2.07524085707333,
+      "learning_rate": 1.805709340832773e-07,
+      "loss": 0.2904,
+      "step": 12302
+    },
+    {
+      "epoch": 1.8826319816373374,
+      "grad_norm": 2.006754654899638,
+      "learning_rate": 1.80102419127689e-07,
+      "loss": 0.2159,
+      "step": 12303
+    },
+    {
+      "epoch": 1.8827850038255547,
+      "grad_norm": 2.0742510517617077,
+      "learning_rate": 1.7963450725042687e-07,
+      "loss": 0.2474,
+      "step": 12304
+    },
+    {
+      "epoch": 1.882938026013772,
+      "grad_norm": 2.082631008210018,
+      "learning_rate": 1.7916719848022902e-07,
+      "loss": 0.2379,
+      "step": 12305
+    },
+    {
+      "epoch": 1.8830910482019894,
+      "grad_norm": 2.032130659886695,
+      "learning_rate": 1.787004928457903e-07,
+      "loss": 0.2687,
+      "step": 12306
+    },
+    {
+      "epoch": 1.8832440703902065,
+      "grad_norm": 2.1550406519384424,
+      "learning_rate": 1.782343903757766e-07,
+      "loss": 0.2834,
+      "step": 12307
+    },
+    {
+      "epoch": 1.883397092578424,
+      "grad_norm": 2.22732515632349,
+      "learning_rate": 1.777688910988118e-07,
+      "loss": 0.3246,
+      "step": 12308
+    },
+    {
+      "epoch": 1.8835501147666411,
+      "grad_norm": 2.4500515480082665,
+      "learning_rate": 1.7730399504348404e-07,
+      "loss": 0.3195,
+      "step": 12309
+    },
+    {
+      "epoch": 1.8837031369548585,
+      "grad_norm": 2.268660747881073,
+      "learning_rate": 1.76839702238345e-07,
+      "loss": 0.3028,
+      "step": 12310
+    },
+    {
+      "epoch": 1.8838561591430758,
+      "grad_norm": 1.6763232823881535,
+      "learning_rate": 1.7637601271190852e-07,
+      "loss": 0.213,
+      "step": 12311
+    },
+    {
+      "epoch": 1.884009181331293,
+      "grad_norm": 2.132550267954547,
+      "learning_rate": 1.7591292649265069e-07,
+      "loss": 0.2764,
+      "step": 12312
+    },
+    {
+      "epoch": 1.8841622035195105,
+      "grad_norm": 3.1774836443797185,
+      "learning_rate": 1.754504436090121e-07,
+      "loss": 0.3129,
+      "step": 12313
+    },
+    {
+      "epoch": 1.8843152257077276,
+      "grad_norm": 2.3320424100195085,
+      "learning_rate": 1.7498856408939558e-07,
+      "loss": 0.2989,
+      "step": 12314
+    },
+    {
+      "epoch": 1.884468247895945,
+      "grad_norm": 2.0604693493677138,
+      "learning_rate": 1.7452728796216732e-07,
+      "loss": 0.3438,
+      "step": 12315
+    },
+    {
+      "epoch": 1.8846212700841622,
+      "grad_norm": 2.090110244983448,
+      "learning_rate": 1.7406661525565582e-07,
+      "loss": 0.2364,
+      "step": 12316
+    },
+    {
+      "epoch": 1.8847742922723794,
+      "grad_norm": 2.235882951459802,
+      "learning_rate": 1.7360654599815285e-07,
+      "loss": 0.3374,
+      "step": 12317
+    },
+    {
+      "epoch": 1.884927314460597,
+      "grad_norm": 2.080730815047136,
+      "learning_rate": 1.731470802179125e-07,
+      "loss": 0.3256,
+      "step": 12318
+    },
+    {
+      "epoch": 1.885080336648814,
+      "grad_norm": 2.3867969064570334,
+      "learning_rate": 1.7268821794315328e-07,
+      "loss": 0.2551,
+      "step": 12319
+    },
+    {
+      "epoch": 1.8852333588370314,
+      "grad_norm": 2.2624261138843966,
+      "learning_rate": 1.7222995920205488e-07,
+      "loss": 0.3053,
+      "step": 12320
+    },
+    {
+      "epoch": 1.8853863810252487,
+      "grad_norm": 2.095280750448679,
+      "learning_rate": 1.7177230402276147e-07,
+      "loss": 0.2418,
+      "step": 12321
+    },
+    {
+      "epoch": 1.8855394032134658,
+      "grad_norm": 1.9204988322442775,
+      "learning_rate": 1.7131525243337944e-07,
+      "loss": 0.2802,
+      "step": 12322
+    },
+    {
+      "epoch": 1.8856924254016834,
+      "grad_norm": 2.3341667000082715,
+      "learning_rate": 1.7085880446197633e-07,
+      "loss": 0.2716,
+      "step": 12323
+    },
+    {
+      "epoch": 1.8858454475899005,
+      "grad_norm": 2.4746042589685717,
+      "learning_rate": 1.7040296013658643e-07,
+      "loss": 0.3252,
+      "step": 12324
+    },
+    {
+      "epoch": 1.8859984697781178,
+      "grad_norm": 2.1480370697284847,
+      "learning_rate": 1.6994771948520506e-07,
+      "loss": 0.2229,
+      "step": 12325
+    },
+    {
+      "epoch": 1.8861514919663351,
+      "grad_norm": 1.8615177634968916,
+      "learning_rate": 1.694930825357899e-07,
+      "loss": 0.2294,
+      "step": 12326
+    },
+    {
+      "epoch": 1.8863045141545522,
+      "grad_norm": 2.215497558948797,
+      "learning_rate": 1.6903904931625968e-07,
+      "loss": 0.2807,
+      "step": 12327
+    },
+    {
+      "epoch": 1.8864575363427698,
+      "grad_norm": 2.191961181905781,
+      "learning_rate": 1.6858561985450327e-07,
+      "loss": 0.2419,
+      "step": 12328
+    },
+    {
+      "epoch": 1.886610558530987,
+      "grad_norm": 1.9739018107667339,
+      "learning_rate": 1.681327941783628e-07,
+      "loss": 0.3075,
+      "step": 12329
+    },
+    {
+      "epoch": 1.8867635807192042,
+      "grad_norm": 1.7804060009453269,
+      "learning_rate": 1.6768057231564938e-07,
+      "loss": 0.2203,
+      "step": 12330
+    },
+    {
+      "epoch": 1.8869166029074216,
+      "grad_norm": 2.2020925173246333,
+      "learning_rate": 1.6722895429413856e-07,
+      "loss": 0.2131,
+      "step": 12331
+    },
+    {
+      "epoch": 1.887069625095639,
+      "grad_norm": 2.1911656545700677,
+      "learning_rate": 1.6677794014156145e-07,
+      "loss": 0.3261,
+      "step": 12332
+    },
+    {
+      "epoch": 1.8872226472838562,
+      "grad_norm": 2.007242826500715,
+      "learning_rate": 1.6632752988561929e-07,
+      "loss": 0.2273,
+      "step": 12333
+    },
+    {
+      "epoch": 1.8873756694720734,
+      "grad_norm": 1.9420679663669713,
+      "learning_rate": 1.6587772355397437e-07,
+      "loss": 0.2709,
+      "step": 12334
+    },
+    {
+      "epoch": 1.8875286916602907,
+      "grad_norm": 2.1440316856194244,
+      "learning_rate": 1.6542852117424791e-07,
+      "loss": 0.239,
+      "step": 12335
+    },
+    {
+      "epoch": 1.887681713848508,
+      "grad_norm": 2.155474547422099,
+      "learning_rate": 1.6497992277403008e-07,
+      "loss": 0.2677,
+      "step": 12336
+    },
+    {
+      "epoch": 1.8878347360367254,
+      "grad_norm": 1.772226946632601,
+      "learning_rate": 1.6453192838086996e-07,
+      "loss": 0.2311,
+      "step": 12337
+    },
+    {
+      "epoch": 1.8879877582249427,
+      "grad_norm": 2.0741420930815706,
+      "learning_rate": 1.6408453802227998e-07,
+      "loss": 0.301,
+      "step": 12338
+    },
+    {
+      "epoch": 1.8881407804131598,
+      "grad_norm": 2.3602693015766856,
+      "learning_rate": 1.6363775172573814e-07,
+      "loss": 0.3471,
+      "step": 12339
+    },
+    {
+      "epoch": 1.8882938026013774,
+      "grad_norm": 1.8958952695132465,
+      "learning_rate": 1.631915695186803e-07,
+      "loss": 0.2583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.8884468247895945,
+      "grad_norm": 2.248151990745066,
+      "learning_rate": 1.6274599142851123e-07,
+      "loss": 0.3297,
+      "step": 12341
+    },
+    {
+      "epoch": 1.8885998469778118,
+      "grad_norm": 2.240236669309958,
+      "learning_rate": 1.6230101748259453e-07,
+      "loss": 0.2999,
+      "step": 12342
+    },
+    {
+      "epoch": 1.8887528691660291,
+      "grad_norm": 2.068997854305708,
+      "learning_rate": 1.6185664770825727e-07,
+      "loss": 0.2607,
+      "step": 12343
+    },
+    {
+      "epoch": 1.8889058913542462,
+      "grad_norm": 2.298467877444234,
+      "learning_rate": 1.6141288213278982e-07,
+      "loss": 0.3697,
+      "step": 12344
+    },
+    {
+      "epoch": 1.8890589135424638,
+      "grad_norm": 2.063704879522499,
+      "learning_rate": 1.6096972078344598e-07,
+      "loss": 0.3015,
+      "step": 12345
+    },
+    {
+      "epoch": 1.889211935730681,
+      "grad_norm": 2.368876065700203,
+      "learning_rate": 1.6052716368744392e-07,
+      "loss": 0.2898,
+      "step": 12346
+    },
+    {
+      "epoch": 1.8893649579188982,
+      "grad_norm": 2.07193482877275,
+      "learning_rate": 1.600852108719597e-07,
+      "loss": 0.2883,
+      "step": 12347
+    },
+    {
+      "epoch": 1.8895179801071156,
+      "grad_norm": 2.1398240131474298,
+      "learning_rate": 1.59643862364135e-07,
+      "loss": 0.297,
+      "step": 12348
+    },
+    {
+      "epoch": 1.8896710022953327,
+      "grad_norm": 2.2600179058392778,
+      "learning_rate": 1.5920311819107915e-07,
+      "loss": 0.328,
+      "step": 12349
+    },
+    {
+      "epoch": 1.8898240244835502,
+      "grad_norm": 2.0289016769988897,
+      "learning_rate": 1.587629783798561e-07,
+      "loss": 0.2667,
+      "step": 12350
+    },
+    {
+      "epoch": 1.8899770466717674,
+      "grad_norm": 1.9507615969456944,
+      "learning_rate": 1.5832344295749758e-07,
+      "loss": 0.2921,
+      "step": 12351
+    },
+    {
+      "epoch": 1.8901300688599847,
+      "grad_norm": 2.2770940001537148,
+      "learning_rate": 1.5788451195099752e-07,
+      "loss": 0.2753,
+      "step": 12352
+    },
+    {
+      "epoch": 1.890283091048202,
+      "grad_norm": 2.434599224167367,
+      "learning_rate": 1.5744618538731106e-07,
+      "loss": 0.3081,
+      "step": 12353
+    },
+    {
+      "epoch": 1.8904361132364191,
+      "grad_norm": 2.1500543081305867,
+      "learning_rate": 1.5700846329335995e-07,
+      "loss": 0.2885,
+      "step": 12354
+    },
+    {
+      "epoch": 1.8905891354246367,
+      "grad_norm": 1.920834570258339,
+      "learning_rate": 1.5657134569602604e-07,
+      "loss": 0.2705,
+      "step": 12355
+    },
+    {
+      "epoch": 1.8907421576128538,
+      "grad_norm": 2.1694939010872174,
+      "learning_rate": 1.5613483262215125e-07,
+      "loss": 0.2975,
+      "step": 12356
+    },
+    {
+      "epoch": 1.8908951798010711,
+      "grad_norm": 2.155350044486877,
+      "learning_rate": 1.5569892409854626e-07,
+      "loss": 0.2426,
+      "step": 12357
+    },
+    {
+      "epoch": 1.8910482019892885,
+      "grad_norm": 2.3352026536074475,
+      "learning_rate": 1.5526362015198305e-07,
+      "loss": 0.3415,
+      "step": 12358
+    },
+    {
+      "epoch": 1.8912012241775056,
+      "grad_norm": 2.425425876866039,
+      "learning_rate": 1.5482892080919132e-07,
+      "loss": 0.3196,
+      "step": 12359
+    },
+    {
+      "epoch": 1.8913542463657231,
+      "grad_norm": 2.1513830506914813,
+      "learning_rate": 1.5439482609687196e-07,
+      "loss": 0.2952,
+      "step": 12360
+    },
+    {
+      "epoch": 1.8915072685539402,
+      "grad_norm": 2.0535776550069884,
+      "learning_rate": 1.539613360416825e-07,
+      "loss": 0.2172,
+      "step": 12361
+    },
+    {
+      "epoch": 1.8916602907421576,
+      "grad_norm": 2.1660181727611514,
+      "learning_rate": 1.5352845067024392e-07,
+      "loss": 0.2655,
+      "step": 12362
+    },
+    {
+      "epoch": 1.891813312930375,
+      "grad_norm": 2.0220484053824195,
+      "learning_rate": 1.530961700091438e-07,
+      "loss": 0.235,
+      "step": 12363
+    },
+    {
+      "epoch": 1.8919663351185922,
+      "grad_norm": 2.0225572484268746,
+      "learning_rate": 1.5266449408492979e-07,
+      "loss": 0.2296,
+      "step": 12364
+    },
+    {
+      "epoch": 1.8921193573068096,
+      "grad_norm": 2.5943193506917366,
+      "learning_rate": 1.522334229241107e-07,
+      "loss": 0.3002,
+      "step": 12365
+    },
+    {
+      "epoch": 1.8922723794950267,
+      "grad_norm": 2.314365201977131,
+      "learning_rate": 1.5180295655316312e-07,
+      "loss": 0.3017,
+      "step": 12366
+    },
+    {
+      "epoch": 1.892425401683244,
+      "grad_norm": 2.0155411326429147,
+      "learning_rate": 1.513730949985215e-07,
+      "loss": 0.268,
+      "step": 12367
+    },
+    {
+      "epoch": 1.8925784238714614,
+      "grad_norm": 1.936628601404376,
+      "learning_rate": 1.5094383828658687e-07,
+      "loss": 0.2173,
+      "step": 12368
+    },
+    {
+      "epoch": 1.8927314460596787,
+      "grad_norm": 2.6655173289279808,
+      "learning_rate": 1.5051518644372044e-07,
+      "loss": 0.3635,
+      "step": 12369
+    },
+    {
+      "epoch": 1.892884468247896,
+      "grad_norm": 1.9786815581966082,
+      "learning_rate": 1.5008713949624777e-07,
+      "loss": 0.3102,
+      "step": 12370
+    },
+    {
+      "epoch": 1.8930374904361131,
+      "grad_norm": 2.1595530135445213,
+      "learning_rate": 1.4965969747045673e-07,
+      "loss": 0.3011,
+      "step": 12371
+    },
+    {
+      "epoch": 1.8931905126243307,
+      "grad_norm": 2.203393747451202,
+      "learning_rate": 1.4923286039259855e-07,
+      "loss": 0.2665,
+      "step": 12372
+    },
+    {
+      "epoch": 1.8933435348125478,
+      "grad_norm": 2.1359510563270736,
+      "learning_rate": 1.488066282888878e-07,
+      "loss": 0.3154,
+      "step": 12373
+    },
+    {
+      "epoch": 1.8934965570007651,
+      "grad_norm": 1.952958864492451,
+      "learning_rate": 1.4838100118549803e-07,
+      "loss": 0.2494,
+      "step": 12374
+    },
+    {
+      "epoch": 1.8936495791889825,
+      "grad_norm": 2.7907391263373866,
+      "learning_rate": 1.4795597910857274e-07,
+      "loss": 0.3871,
+      "step": 12375
+    },
+    {
+      "epoch": 1.8938026013771996,
+      "grad_norm": 2.601807768733684,
+      "learning_rate": 1.4753156208421104e-07,
+      "loss": 0.2742,
+      "step": 12376
+    },
+    {
+      "epoch": 1.8939556235654171,
+      "grad_norm": 2.0452811830427478,
+      "learning_rate": 1.4710775013847879e-07,
+      "loss": 0.2629,
+      "step": 12377
+    },
+    {
+      "epoch": 1.8941086457536342,
+      "grad_norm": 2.1949730369156284,
+      "learning_rate": 1.4668454329740621e-07,
+      "loss": 0.3188,
+      "step": 12378
+    },
+    {
+      "epoch": 1.8942616679418516,
+      "grad_norm": 2.1266212476482753,
+      "learning_rate": 1.4626194158698149e-07,
+      "loss": 0.293,
+      "step": 12379
+    },
+    {
+      "epoch": 1.894414690130069,
+      "grad_norm": 2.3910652112091997,
+      "learning_rate": 1.4583994503315713e-07,
+      "loss": 0.3309,
+      "step": 12380
+    },
+    {
+      "epoch": 1.894567712318286,
+      "grad_norm": 2.195165911640832,
+      "learning_rate": 1.4541855366185464e-07,
+      "loss": 0.261,
+      "step": 12381
+    },
+    {
+      "epoch": 1.8947207345065036,
+      "grad_norm": 2.2227897094612015,
+      "learning_rate": 1.4499776749894668e-07,
+      "loss": 0.2892,
+      "step": 12382
+    },
+    {
+      "epoch": 1.8948737566947207,
+      "grad_norm": 1.983371860106799,
+      "learning_rate": 1.4457758657028142e-07,
+      "loss": 0.2333,
+      "step": 12383
+    },
+    {
+      "epoch": 1.895026778882938,
+      "grad_norm": 2.166258470216104,
+      "learning_rate": 1.4415801090166048e-07,
+      "loss": 0.2943,
+      "step": 12384
+    },
+    {
+      "epoch": 1.8951798010711554,
+      "grad_norm": 2.421166045256063,
+      "learning_rate": 1.43739040518851e-07,
+      "loss": 0.3261,
+      "step": 12385
+    },
+    {
+      "epoch": 1.8953328232593725,
+      "grad_norm": 2.5040245258127927,
+      "learning_rate": 1.4332067544758688e-07,
+      "loss": 0.3423,
+      "step": 12386
+    },
+    {
+      "epoch": 1.89548584544759,
+      "grad_norm": 2.0306426946320775,
+      "learning_rate": 1.4290291571355975e-07,
+      "loss": 0.2614,
+      "step": 12387
+    },
+    {
+      "epoch": 1.8956388676358071,
+      "grad_norm": 1.902404848644786,
+      "learning_rate": 1.4248576134242353e-07,
+      "loss": 0.235,
+      "step": 12388
+    },
+    {
+      "epoch": 1.8957918898240245,
+      "grad_norm": 2.030986303913293,
+      "learning_rate": 1.420692123598011e-07,
+      "loss": 0.2736,
+      "step": 12389
+    },
+    {
+      "epoch": 1.8959449120122418,
+      "grad_norm": 2.281148057551701,
+      "learning_rate": 1.4165326879127195e-07,
+      "loss": 0.357,
+      "step": 12390
+    },
+    {
+      "epoch": 1.896097934200459,
+      "grad_norm": 2.1204540996922163,
+      "learning_rate": 1.4123793066238233e-07,
+      "loss": 0.2949,
+      "step": 12391
+    },
+    {
+      "epoch": 1.8962509563886765,
+      "grad_norm": 2.023496130387062,
+      "learning_rate": 1.4082319799863963e-07,
+      "loss": 0.276,
+      "step": 12392
+    },
+    {
+      "epoch": 1.8964039785768936,
+      "grad_norm": 2.1823663321503886,
+      "learning_rate": 1.4040907082551237e-07,
+      "loss": 0.3208,
+      "step": 12393
+    },
+    {
+      "epoch": 1.896557000765111,
+      "grad_norm": 2.002947903538194,
+      "learning_rate": 1.3999554916843571e-07,
+      "loss": 0.2331,
+      "step": 12394
+    },
+    {
+      "epoch": 1.8967100229533282,
+      "grad_norm": 2.046209375144362,
+      "learning_rate": 1.3958263305280494e-07,
+      "loss": 0.2501,
+      "step": 12395
+    },
+    {
+      "epoch": 1.8968630451415456,
+      "grad_norm": 1.9425497013925002,
+      "learning_rate": 1.3917032250397867e-07,
+      "loss": 0.2359,
+      "step": 12396
+    },
+    {
+      "epoch": 1.897016067329763,
+      "grad_norm": 2.0608664471035745,
+      "learning_rate": 1.387586175472788e-07,
+      "loss": 0.2626,
+      "step": 12397
+    },
+    {
+      "epoch": 1.89716908951798,
+      "grad_norm": 2.1798385264217734,
+      "learning_rate": 1.3834751820798964e-07,
+      "loss": 0.2987,
+      "step": 12398
+    },
+    {
+      "epoch": 1.8973221117061974,
+      "grad_norm": 1.976167049765422,
+      "learning_rate": 1.3793702451135872e-07,
+      "loss": 0.2521,
+      "step": 12399
+    },
+    {
+      "epoch": 1.8974751338944147,
+      "grad_norm": 2.1064510108566052,
+      "learning_rate": 1.3752713648259475e-07,
+      "loss": 0.3585,
+      "step": 12400
+    },
+    {
+      "epoch": 1.897628156082632,
+      "grad_norm": 1.8467627262346928,
+      "learning_rate": 1.3711785414687207e-07,
+      "loss": 0.2776,
+      "step": 12401
+    },
+    {
+      "epoch": 1.8977811782708494,
+      "grad_norm": 2.0021520489916513,
+      "learning_rate": 1.367091775293261e-07,
+      "loss": 0.2253,
+      "step": 12402
+    },
+    {
+      "epoch": 1.8979342004590665,
+      "grad_norm": 2.004962053610836,
+      "learning_rate": 1.3630110665505347e-07,
+      "loss": 0.2494,
+      "step": 12403
+    },
+    {
+      "epoch": 1.898087222647284,
+      "grad_norm": 2.1024242844915,
+      "learning_rate": 1.3589364154911855e-07,
+      "loss": 0.2501,
+      "step": 12404
+    },
+    {
+      "epoch": 1.8982402448355011,
+      "grad_norm": 1.974976585706706,
+      "learning_rate": 1.3548678223654354e-07,
+      "loss": 0.2945,
+      "step": 12405
+    },
+    {
+      "epoch": 1.8983932670237185,
+      "grad_norm": 1.822674609909908,
+      "learning_rate": 1.3508052874231514e-07,
+      "loss": 0.2078,
+      "step": 12406
+    },
+    {
+      "epoch": 1.8985462892119358,
+      "grad_norm": 2.409596533644061,
+      "learning_rate": 1.346748810913834e-07,
+      "loss": 0.2808,
+      "step": 12407
+    },
+    {
+      "epoch": 1.898699311400153,
+      "grad_norm": 2.1866218250648752,
+      "learning_rate": 1.3426983930866055e-07,
+      "loss": 0.2917,
+      "step": 12408
+    },
+    {
+      "epoch": 1.8988523335883705,
+      "grad_norm": 2.2318137725230778,
+      "learning_rate": 1.338654034190212e-07,
+      "loss": 0.3389,
+      "step": 12409
+    },
+    {
+      "epoch": 1.8990053557765876,
+      "grad_norm": 2.1864107341649834,
+      "learning_rate": 1.3346157344730438e-07,
+      "loss": 0.2668,
+      "step": 12410
+    },
+    {
+      "epoch": 1.899158377964805,
+      "grad_norm": 2.0640684514509036,
+      "learning_rate": 1.330583494183102e-07,
+      "loss": 0.3253,
+      "step": 12411
+    },
+    {
+      "epoch": 1.8993114001530222,
+      "grad_norm": 2.061874070757145,
+      "learning_rate": 1.326557313568033e-07,
+      "loss": 0.2789,
+      "step": 12412
+    },
+    {
+      "epoch": 1.8994644223412394,
+      "grad_norm": 2.262601025711515,
+      "learning_rate": 1.3225371928750842e-07,
+      "loss": 0.3414,
+      "step": 12413
+    },
+    {
+      "epoch": 1.899617444529457,
+      "grad_norm": 2.071223047182192,
+      "learning_rate": 1.318523132351157e-07,
+      "loss": 0.2732,
+      "step": 12414
+    },
+    {
+      "epoch": 1.899770466717674,
+      "grad_norm": 2.2037935773831743,
+      "learning_rate": 1.3145151322427663e-07,
+      "loss": 0.2175,
+      "step": 12415
+    },
+    {
+      "epoch": 1.8999234889058914,
+      "grad_norm": 1.9905111743598205,
+      "learning_rate": 1.3105131927960702e-07,
+      "loss": 0.257,
+      "step": 12416
+    },
+    {
+      "epoch": 1.9000765110941087,
+      "grad_norm": 2.0918754035359166,
+      "learning_rate": 1.3065173142568276e-07,
+      "loss": 0.2877,
+      "step": 12417
+    },
+    {
+      "epoch": 1.9002295332823258,
+      "grad_norm": 2.1951012882049237,
+      "learning_rate": 1.3025274968704426e-07,
+      "loss": 0.3017,
+      "step": 12418
+    },
+    {
+      "epoch": 1.9003825554705434,
+      "grad_norm": 2.047502157807223,
+      "learning_rate": 1.2985437408819524e-07,
+      "loss": 0.2675,
+      "step": 12419
+    },
+    {
+      "epoch": 1.9005355776587605,
+      "grad_norm": 2.345449255427472,
+      "learning_rate": 1.2945660465360166e-07,
+      "loss": 0.2968,
+      "step": 12420
+    },
+    {
+      "epoch": 1.9006885998469778,
+      "grad_norm": 2.2405813389873455,
+      "learning_rate": 1.2905944140769178e-07,
+      "loss": 0.2276,
+      "step": 12421
+    },
+    {
+      "epoch": 1.9008416220351951,
+      "grad_norm": 1.8624021115253389,
+      "learning_rate": 1.2866288437485718e-07,
+      "loss": 0.252,
+      "step": 12422
+    },
+    {
+      "epoch": 1.9009946442234122,
+      "grad_norm": 2.0359610897944154,
+      "learning_rate": 1.2826693357945176e-07,
+      "loss": 0.2366,
+      "step": 12423
+    },
+    {
+      "epoch": 1.9011476664116298,
+      "grad_norm": 2.2615454945887556,
+      "learning_rate": 1.2787158904579268e-07,
+      "loss": 0.2951,
+      "step": 12424
+    },
+    {
+      "epoch": 1.901300688599847,
+      "grad_norm": 2.738627055525505,
+      "learning_rate": 1.274768507981583e-07,
+      "loss": 0.3292,
+      "step": 12425
+    },
+    {
+      "epoch": 1.9014537107880642,
+      "grad_norm": 2.153882026555609,
+      "learning_rate": 1.270827188607926e-07,
+      "loss": 0.2416,
+      "step": 12426
+    },
+    {
+      "epoch": 1.9016067329762816,
+      "grad_norm": 2.000143651464614,
+      "learning_rate": 1.2668919325789953e-07,
+      "loss": 0.2669,
+      "step": 12427
+    },
+    {
+      "epoch": 1.9017597551644987,
+      "grad_norm": 2.1215493546758073,
+      "learning_rate": 1.2629627401364864e-07,
+      "loss": 0.3234,
+      "step": 12428
+    },
+    {
+      "epoch": 1.9019127773527162,
+      "grad_norm": 1.9597291534053758,
+      "learning_rate": 1.2590396115216953e-07,
+      "loss": 0.2631,
+      "step": 12429
+    },
+    {
+      "epoch": 1.9020657995409334,
+      "grad_norm": 2.429837051524621,
+      "learning_rate": 1.2551225469755512e-07,
+      "loss": 0.3115,
+      "step": 12430
+    },
+    {
+      "epoch": 1.9022188217291507,
+      "grad_norm": 2.089905167506059,
+      "learning_rate": 1.25121154673864e-07,
+      "loss": 0.2776,
+      "step": 12431
+    },
+    {
+      "epoch": 1.902371843917368,
+      "grad_norm": 1.9529949744574147,
+      "learning_rate": 1.2473066110511244e-07,
+      "loss": 0.2083,
+      "step": 12432
+    },
+    {
+      "epoch": 1.9025248661055854,
+      "grad_norm": 2.2045107277586258,
+      "learning_rate": 1.2434077401528245e-07,
+      "loss": 0.2174,
+      "step": 12433
+    },
+    {
+      "epoch": 1.9026778882938027,
+      "grad_norm": 2.1168068178123405,
+      "learning_rate": 1.2395149342832035e-07,
+      "loss": 0.2749,
+      "step": 12434
+    },
+    {
+      "epoch": 1.9028309104820198,
+      "grad_norm": 2.0035038229922337,
+      "learning_rate": 1.2356281936813152e-07,
+      "loss": 0.2764,
+      "step": 12435
+    },
+    {
+      "epoch": 1.9029839326702371,
+      "grad_norm": 1.8645531982106915,
+      "learning_rate": 1.2317475185858797e-07,
+      "loss": 0.2857,
+      "step": 12436
+    },
+    {
+      "epoch": 1.9031369548584545,
+      "grad_norm": 1.9683962909981605,
+      "learning_rate": 1.227872909235206e-07,
+      "loss": 0.2156,
+      "step": 12437
+    },
+    {
+      "epoch": 1.9032899770466718,
+      "grad_norm": 2.5372871229932756,
+      "learning_rate": 1.2240043658672485e-07,
+      "loss": 0.3423,
+      "step": 12438
+    },
+    {
+      "epoch": 1.9034429992348891,
+      "grad_norm": 2.2487446160355384,
+      "learning_rate": 1.220141888719606e-07,
+      "loss": 0.3086,
+      "step": 12439
+    },
+    {
+      "epoch": 1.9035960214231062,
+      "grad_norm": 2.4433518403609487,
+      "learning_rate": 1.2162854780294775e-07,
+      "loss": 0.3025,
+      "step": 12440
+    },
+    {
+      "epoch": 1.9037490436113238,
+      "grad_norm": 2.285883593805717,
+      "learning_rate": 1.212435134033707e-07,
+      "loss": 0.3215,
+      "step": 12441
+    },
+    {
+      "epoch": 1.903902065799541,
+      "grad_norm": 2.07019078472154,
+      "learning_rate": 1.20859085696875e-07,
+      "loss": 0.2195,
+      "step": 12442
+    },
+    {
+      "epoch": 1.9040550879877582,
+      "grad_norm": 2.1764385332904204,
+      "learning_rate": 1.204752647070706e-07,
+      "loss": 0.2439,
+      "step": 12443
+    },
+    {
+      "epoch": 1.9042081101759756,
+      "grad_norm": 2.0657575704110194,
+      "learning_rate": 1.200920504575287e-07,
+      "loss": 0.3091,
+      "step": 12444
+    },
+    {
+      "epoch": 1.9043611323641927,
+      "grad_norm": 2.2142022624529867,
+      "learning_rate": 1.1970944297178377e-07,
+      "loss": 0.3519,
+      "step": 12445
+    },
+    {
+      "epoch": 1.9045141545524102,
+      "grad_norm": 2.2201081578650896,
+      "learning_rate": 1.1932744227333481e-07,
+      "loss": 0.2401,
+      "step": 12446
+    },
+    {
+      "epoch": 1.9046671767406274,
+      "grad_norm": 2.0766398494803036,
+      "learning_rate": 1.1894604838564083e-07,
+      "loss": 0.2747,
+      "step": 12447
+    },
+    {
+      "epoch": 1.9048201989288447,
+      "grad_norm": 1.8669964143468456,
+      "learning_rate": 1.1856526133212421e-07,
+      "loss": 0.2535,
+      "step": 12448
+    },
+    {
+      "epoch": 1.904973221117062,
+      "grad_norm": 2.1864682779631717,
+      "learning_rate": 1.1818508113617288e-07,
+      "loss": 0.2969,
+      "step": 12449
+    },
+    {
+      "epoch": 1.9051262433052791,
+      "grad_norm": 2.1671985516248315,
+      "learning_rate": 1.1780550782113154e-07,
+      "loss": 0.2791,
+      "step": 12450
+    },
+    {
+      "epoch": 1.9052792654934967,
+      "grad_norm": 1.7088933065196328,
+      "learning_rate": 1.1742654141031484e-07,
+      "loss": 0.1846,
+      "step": 12451
+    },
+    {
+      "epoch": 1.9054322876817138,
+      "grad_norm": 2.148970892648732,
+      "learning_rate": 1.1704818192699419e-07,
+      "loss": 0.2711,
+      "step": 12452
+    },
+    {
+      "epoch": 1.9055853098699311,
+      "grad_norm": 2.100215150528268,
+      "learning_rate": 1.1667042939440765e-07,
+      "loss": 0.252,
+      "step": 12453
+    },
+    {
+      "epoch": 1.9057383320581485,
+      "grad_norm": 2.360942987759934,
+      "learning_rate": 1.1629328383575444e-07,
+      "loss": 0.3214,
+      "step": 12454
+    },
+    {
+      "epoch": 1.9058913542463656,
+      "grad_norm": 2.1994757368737536,
+      "learning_rate": 1.1591674527419495e-07,
+      "loss": 0.2825,
+      "step": 12455
+    },
+    {
+      "epoch": 1.9060443764345831,
+      "grad_norm": 2.1044747531659933,
+      "learning_rate": 1.1554081373285398e-07,
+      "loss": 0.3447,
+      "step": 12456
+    },
+    {
+      "epoch": 1.9061973986228002,
+      "grad_norm": 1.779675570955133,
+      "learning_rate": 1.1516548923482196e-07,
+      "loss": 0.2186,
+      "step": 12457
+    },
+    {
+      "epoch": 1.9063504208110176,
+      "grad_norm": 2.09708548162683,
+      "learning_rate": 1.1479077180314601e-07,
+      "loss": 0.2884,
+      "step": 12458
+    },
+    {
+      "epoch": 1.906503442999235,
+      "grad_norm": 1.909809773435519,
+      "learning_rate": 1.1441666146083885e-07,
+      "loss": 0.2479,
+      "step": 12459
+    },
+    {
+      "epoch": 1.906656465187452,
+      "grad_norm": 1.806316291654644,
+      "learning_rate": 1.1404315823087875e-07,
+      "loss": 0.2085,
+      "step": 12460
+    },
+    {
+      "epoch": 1.9068094873756696,
+      "grad_norm": 2.127004509384845,
+      "learning_rate": 1.1367026213620181e-07,
+      "loss": 0.2636,
+      "step": 12461
+    },
+    {
+      "epoch": 1.9069625095638867,
+      "grad_norm": 2.205530653360376,
+      "learning_rate": 1.1329797319970859e-07,
+      "loss": 0.2809,
+      "step": 12462
+    },
+    {
+      "epoch": 1.907115531752104,
+      "grad_norm": 2.41892985391259,
+      "learning_rate": 1.1292629144426637e-07,
+      "loss": 0.3434,
+      "step": 12463
+    },
+    {
+      "epoch": 1.9072685539403214,
+      "grad_norm": 2.0336010224200036,
+      "learning_rate": 1.1255521689269577e-07,
+      "loss": 0.2807,
+      "step": 12464
+    },
+    {
+      "epoch": 1.9074215761285387,
+      "grad_norm": 2.721142821132342,
+      "learning_rate": 1.1218474956779191e-07,
+      "loss": 0.3746,
+      "step": 12465
+    },
+    {
+      "epoch": 1.907574598316756,
+      "grad_norm": 1.8455778126082905,
+      "learning_rate": 1.1181488949230323e-07,
+      "loss": 0.2009,
+      "step": 12466
+    },
+    {
+      "epoch": 1.9077276205049731,
+      "grad_norm": 2.164500498682435,
+      "learning_rate": 1.1144563668894381e-07,
+      "loss": 0.3123,
+      "step": 12467
+    },
+    {
+      "epoch": 1.9078806426931905,
+      "grad_norm": 2.3175653893300705,
+      "learning_rate": 1.1107699118039328e-07,
+      "loss": 0.2817,
+      "step": 12468
+    },
+    {
+      "epoch": 1.9080336648814078,
+      "grad_norm": 2.1705603938031337,
+      "learning_rate": 1.1070895298929019e-07,
+      "loss": 0.2563,
+      "step": 12469
+    },
+    {
+      "epoch": 1.9081866870696251,
+      "grad_norm": 2.170225385518329,
+      "learning_rate": 1.1034152213823646e-07,
+      "loss": 0.3054,
+      "step": 12470
+    },
+    {
+      "epoch": 1.9083397092578425,
+      "grad_norm": 2.392108868688555,
+      "learning_rate": 1.099746986497996e-07,
+      "loss": 0.3374,
+      "step": 12471
+    },
+    {
+      "epoch": 1.9084927314460596,
+      "grad_norm": 1.943591082629844,
+      "learning_rate": 1.0960848254650603e-07,
+      "loss": 0.2415,
+      "step": 12472
+    },
+    {
+      "epoch": 1.9086457536342771,
+      "grad_norm": 2.144948863915394,
+      "learning_rate": 1.0924287385084776e-07,
+      "loss": 0.2933,
+      "step": 12473
+    },
+    {
+      "epoch": 1.9087987758224942,
+      "grad_norm": 1.9600270754442108,
+      "learning_rate": 1.0887787258527682e-07,
+      "loss": 0.27,
+      "step": 12474
+    },
+    {
+      "epoch": 1.9089517980107116,
+      "grad_norm": 2.29484241675231,
+      "learning_rate": 1.0851347877220975e-07,
+      "loss": 0.3108,
+      "step": 12475
+    },
+    {
+      "epoch": 1.909104820198929,
+      "grad_norm": 1.8828750051292165,
+      "learning_rate": 1.0814969243402529e-07,
+      "loss": 0.2352,
+      "step": 12476
+    },
+    {
+      "epoch": 1.909257842387146,
+      "grad_norm": 2.089128050877077,
+      "learning_rate": 1.0778651359306669e-07,
+      "loss": 0.2714,
+      "step": 12477
+    },
+    {
+      "epoch": 1.9094108645753636,
+      "grad_norm": 2.2924888911310894,
+      "learning_rate": 1.074239422716361e-07,
+      "loss": 0.2585,
+      "step": 12478
+    },
+    {
+      "epoch": 1.9095638867635807,
+      "grad_norm": 2.1199032370475708,
+      "learning_rate": 1.0706197849200128e-07,
+      "loss": 0.2535,
+      "step": 12479
+    },
+    {
+      "epoch": 1.909716908951798,
+      "grad_norm": 1.80239557608064,
+      "learning_rate": 1.0670062227639111e-07,
+      "loss": 0.265,
+      "step": 12480
+    },
+    {
+      "epoch": 1.9098699311400154,
+      "grad_norm": 2.147866520218618,
+      "learning_rate": 1.0633987364700004e-07,
+      "loss": 0.223,
+      "step": 12481
+    },
+    {
+      "epoch": 1.9100229533282325,
+      "grad_norm": 2.466143189763118,
+      "learning_rate": 1.0597973262598038e-07,
+      "loss": 0.3343,
+      "step": 12482
+    },
+    {
+      "epoch": 1.91017597551645,
+      "grad_norm": 2.3340145813910613,
+      "learning_rate": 1.0562019923545108e-07,
+      "loss": 0.3126,
+      "step": 12483
+    },
+    {
+      "epoch": 1.9103289977046671,
+      "grad_norm": 2.1269179837487022,
+      "learning_rate": 1.0526127349749227e-07,
+      "loss": 0.2429,
+      "step": 12484
+    },
+    {
+      "epoch": 1.9104820198928845,
+      "grad_norm": 2.248127125499551,
+      "learning_rate": 1.049029554341463e-07,
+      "loss": 0.2539,
+      "step": 12485
+    },
+    {
+      "epoch": 1.9106350420811018,
+      "grad_norm": 2.0877173544121357,
+      "learning_rate": 1.0454524506742225e-07,
+      "loss": 0.2841,
+      "step": 12486
+    },
+    {
+      "epoch": 1.910788064269319,
+      "grad_norm": 2.1014344280122867,
+      "learning_rate": 1.0418814241928366e-07,
+      "loss": 0.2673,
+      "step": 12487
+    },
+    {
+      "epoch": 1.9109410864575365,
+      "grad_norm": 2.2504178395361425,
+      "learning_rate": 1.0383164751166409e-07,
+      "loss": 0.2564,
+      "step": 12488
+    },
+    {
+      "epoch": 1.9110941086457536,
+      "grad_norm": 1.9582453306000769,
+      "learning_rate": 1.0347576036645824e-07,
+      "loss": 0.2639,
+      "step": 12489
+    },
+    {
+      "epoch": 1.911247130833971,
+      "grad_norm": 2.0337728796022656,
+      "learning_rate": 1.0312048100552085e-07,
+      "loss": 0.2854,
+      "step": 12490
+    },
+    {
+      "epoch": 1.9114001530221882,
+      "grad_norm": 2.3114295530401257,
+      "learning_rate": 1.0276580945067116e-07,
+      "loss": 0.2843,
+      "step": 12491
+    },
+    {
+      "epoch": 1.9115531752104054,
+      "grad_norm": 1.7069199598118328,
+      "learning_rate": 1.0241174572369172e-07,
+      "loss": 0.2143,
+      "step": 12492
+    },
+    {
+      "epoch": 1.911706197398623,
+      "grad_norm": 2.099298087764273,
+      "learning_rate": 1.0205828984632626e-07,
+      "loss": 0.2617,
+      "step": 12493
+    },
+    {
+      "epoch": 1.91185921958684,
+      "grad_norm": 2.2982638891837803,
+      "learning_rate": 1.0170544184028186e-07,
+      "loss": 0.3266,
+      "step": 12494
+    },
+    {
+      "epoch": 1.9120122417750574,
+      "grad_norm": 1.921762199018458,
+      "learning_rate": 1.0135320172723007e-07,
+      "loss": 0.2517,
+      "step": 12495
+    },
+    {
+      "epoch": 1.9121652639632747,
+      "grad_norm": 2.0989247222133285,
+      "learning_rate": 1.0100156952880025e-07,
+      "loss": 0.2335,
+      "step": 12496
+    },
+    {
+      "epoch": 1.912318286151492,
+      "grad_norm": 2.2323329776738263,
+      "learning_rate": 1.0065054526658957e-07,
+      "loss": 0.2551,
+      "step": 12497
+    },
+    {
+      "epoch": 1.9124713083397094,
+      "grad_norm": 2.2590344295387066,
+      "learning_rate": 1.0030012896215635e-07,
+      "loss": 0.2943,
+      "step": 12498
+    },
+    {
+      "epoch": 1.9126243305279265,
+      "grad_norm": 2.2349113812770134,
+      "learning_rate": 9.995032063701781e-08,
+      "loss": 0.2179,
+      "step": 12499
+    },
+    {
+      "epoch": 1.9127773527161438,
+      "grad_norm": 2.086848029996208,
+      "learning_rate": 9.960112031266123e-08,
+      "loss": 0.2488,
+      "step": 12500
+    },
+    {
+      "epoch": 1.9129303749043611,
+      "grad_norm": 2.5167862029364088,
+      "learning_rate": 9.925252801052943e-08,
+      "loss": 0.2748,
+      "step": 12501
+    },
+    {
+      "epoch": 1.9130833970925785,
+      "grad_norm": 1.8074348221085181,
+      "learning_rate": 9.890454375203306e-08,
+      "loss": 0.2377,
+      "step": 12502
+    },
+    {
+      "epoch": 1.9132364192807958,
+      "grad_norm": 1.9891945424069828,
+      "learning_rate": 9.855716755854062e-08,
+      "loss": 0.2427,
+      "step": 12503
+    },
+    {
+      "epoch": 1.913389441469013,
+      "grad_norm": 1.9930912025356717,
+      "learning_rate": 9.821039945138833e-08,
+      "loss": 0.2679,
+      "step": 12504
+    },
+    {
+      "epoch": 1.9135424636572305,
+      "grad_norm": 1.9143467675625112,
+      "learning_rate": 9.78642394518714e-08,
+      "loss": 0.2671,
+      "step": 12505
+    },
+    {
+      "epoch": 1.9136954858454476,
+      "grad_norm": 2.3126000901167094,
+      "learning_rate": 9.751868758124839e-08,
+      "loss": 0.3273,
+      "step": 12506
+    },
+    {
+      "epoch": 1.913848508033665,
+      "grad_norm": 2.48212979320504,
+      "learning_rate": 9.717374386074118e-08,
+      "loss": 0.3258,
+      "step": 12507
+    },
+    {
+      "epoch": 1.9140015302218822,
+      "grad_norm": 1.9366579694374044,
+      "learning_rate": 9.682940831153509e-08,
+      "loss": 0.2183,
+      "step": 12508
+    },
+    {
+      "epoch": 1.9141545524100994,
+      "grad_norm": 2.146378976268681,
+      "learning_rate": 9.648568095477539e-08,
+      "loss": 0.3257,
+      "step": 12509
+    },
+    {
+      "epoch": 1.914307574598317,
+      "grad_norm": 2.0975727968979614,
+      "learning_rate": 9.614256181157411e-08,
+      "loss": 0.2838,
+      "step": 12510
+    },
+    {
+      "epoch": 1.914460596786534,
+      "grad_norm": 2.3790386333219535,
+      "learning_rate": 9.580005090300104e-08,
+      "loss": 0.2916,
+      "step": 12511
+    },
+    {
+      "epoch": 1.9146136189747514,
+      "grad_norm": 2.063496598618312,
+      "learning_rate": 9.54581482500927e-08,
+      "loss": 0.2567,
+      "step": 12512
+    },
+    {
+      "epoch": 1.9147666411629687,
+      "grad_norm": 2.177710720193587,
+      "learning_rate": 9.511685387384673e-08,
+      "loss": 0.2677,
+      "step": 12513
+    },
+    {
+      "epoch": 1.9149196633511858,
+      "grad_norm": 2.3674245980586703,
+      "learning_rate": 9.477616779522191e-08,
+      "loss": 0.3017,
+      "step": 12514
+    },
+    {
+      "epoch": 1.9150726855394034,
+      "grad_norm": 2.3237512722434808,
+      "learning_rate": 9.443609003514376e-08,
+      "loss": 0.2724,
+      "step": 12515
+    },
+    {
+      "epoch": 1.9152257077276205,
+      "grad_norm": 2.1349857240907815,
+      "learning_rate": 9.409662061449553e-08,
+      "loss": 0.278,
+      "step": 12516
+    },
+    {
+      "epoch": 1.9153787299158378,
+      "grad_norm": 1.929247113089893,
+      "learning_rate": 9.375775955412502e-08,
+      "loss": 0.2658,
+      "step": 12517
+    },
+    {
+      "epoch": 1.9155317521040551,
+      "grad_norm": 2.078402299391299,
+      "learning_rate": 9.341950687484447e-08,
+      "loss": 0.2476,
+      "step": 12518
+    },
+    {
+      "epoch": 1.9156847742922722,
+      "grad_norm": 2.150314478366211,
+      "learning_rate": 9.308186259742724e-08,
+      "loss": 0.292,
+      "step": 12519
+    },
+    {
+      "epoch": 1.9158377964804898,
+      "grad_norm": 2.2735395559159626,
+      "learning_rate": 9.274482674260787e-08,
+      "loss": 0.295,
+      "step": 12520
+    },
+    {
+      "epoch": 1.915990818668707,
+      "grad_norm": 2.0485204164380066,
+      "learning_rate": 9.240839933108647e-08,
+      "loss": 0.2815,
+      "step": 12521
+    },
+    {
+      "epoch": 1.9161438408569242,
+      "grad_norm": 1.9152040534504782,
+      "learning_rate": 9.20725803835243e-08,
+      "loss": 0.2533,
+      "step": 12522
+    },
+    {
+      "epoch": 1.9162968630451416,
+      "grad_norm": 2.280007794782544,
+      "learning_rate": 9.173736992054483e-08,
+      "loss": 0.2678,
+      "step": 12523
+    },
+    {
+      "epoch": 1.9164498852333587,
+      "grad_norm": 2.5147731875289656,
+      "learning_rate": 9.140276796273495e-08,
+      "loss": 0.3013,
+      "step": 12524
+    },
+    {
+      "epoch": 1.9166029074215762,
+      "grad_norm": 2.1676092373292843,
+      "learning_rate": 9.106877453064267e-08,
+      "loss": 0.2655,
+      "step": 12525
+    },
+    {
+      "epoch": 1.9167559296097934,
+      "grad_norm": 2.3267045294522135,
+      "learning_rate": 9.073538964478267e-08,
+      "loss": 0.2656,
+      "step": 12526
+    },
+    {
+      "epoch": 1.9169089517980107,
+      "grad_norm": 2.089513835534899,
+      "learning_rate": 9.040261332562639e-08,
+      "loss": 0.2434,
+      "step": 12527
+    },
+    {
+      "epoch": 1.917061973986228,
+      "grad_norm": 2.163041959965656,
+      "learning_rate": 9.0070445593613e-08,
+      "loss": 0.2686,
+      "step": 12528
+    },
+    {
+      "epoch": 1.9172149961744451,
+      "grad_norm": 1.979792412059793,
+      "learning_rate": 8.973888646914064e-08,
+      "loss": 0.241,
+      "step": 12529
+    },
+    {
+      "epoch": 1.9173680183626627,
+      "grad_norm": 1.8840145007212765,
+      "learning_rate": 8.940793597257303e-08,
+      "loss": 0.2338,
+      "step": 12530
+    },
+    {
+      "epoch": 1.9175210405508798,
+      "grad_norm": 2.101688674121348,
+      "learning_rate": 8.90775941242339e-08,
+      "loss": 0.2335,
+      "step": 12531
+    },
+    {
+      "epoch": 1.9176740627390971,
+      "grad_norm": 2.200367918955295,
+      "learning_rate": 8.874786094441257e-08,
+      "loss": 0.2976,
+      "step": 12532
+    },
+    {
+      "epoch": 1.9178270849273145,
+      "grad_norm": 2.0360117662478583,
+      "learning_rate": 8.84187364533573e-08,
+      "loss": 0.2713,
+      "step": 12533
+    },
+    {
+      "epoch": 1.9179801071155318,
+      "grad_norm": 1.870087321117023,
+      "learning_rate": 8.809022067128193e-08,
+      "loss": 0.2147,
+      "step": 12534
+    },
+    {
+      "epoch": 1.9181331293037491,
+      "grad_norm": 1.9813617183509307,
+      "learning_rate": 8.776231361836141e-08,
+      "loss": 0.2583,
+      "step": 12535
+    },
+    {
+      "epoch": 1.9182861514919662,
+      "grad_norm": 2.285906135950109,
+      "learning_rate": 8.743501531473409e-08,
+      "loss": 0.2987,
+      "step": 12536
+    },
+    {
+      "epoch": 1.9184391736801836,
+      "grad_norm": 2.0323696479504765,
+      "learning_rate": 8.710832578050166e-08,
+      "loss": 0.2593,
+      "step": 12537
+    },
+    {
+      "epoch": 1.918592195868401,
+      "grad_norm": 2.1254409285361473,
+      "learning_rate": 8.678224503572474e-08,
+      "loss": 0.2387,
+      "step": 12538
+    },
+    {
+      "epoch": 1.9187452180566182,
+      "grad_norm": 2.023023101695638,
+      "learning_rate": 8.645677310043288e-08,
+      "loss": 0.2443,
+      "step": 12539
+    },
+    {
+      "epoch": 1.9188982402448356,
+      "grad_norm": 2.0102458000598773,
+      "learning_rate": 8.613190999461119e-08,
+      "loss": 0.2564,
+      "step": 12540
+    },
+    {
+      "epoch": 1.9190512624330527,
+      "grad_norm": 1.8525700822632671,
+      "learning_rate": 8.580765573821148e-08,
+      "loss": 0.2515,
+      "step": 12541
+    },
+    {
+      "epoch": 1.9192042846212702,
+      "grad_norm": 2.096632072259417,
+      "learning_rate": 8.548401035115006e-08,
+      "loss": 0.2639,
+      "step": 12542
+    },
+    {
+      "epoch": 1.9193573068094874,
+      "grad_norm": 2.1823118733419493,
+      "learning_rate": 8.516097385330102e-08,
+      "loss": 0.2719,
+      "step": 12543
+    },
+    {
+      "epoch": 1.9195103289977047,
+      "grad_norm": 2.217747952601633,
+      "learning_rate": 8.483854626450183e-08,
+      "loss": 0.2093,
+      "step": 12544
+    },
+    {
+      "epoch": 1.919663351185922,
+      "grad_norm": 2.109037847197852,
+      "learning_rate": 8.451672760455775e-08,
+      "loss": 0.3524,
+      "step": 12545
+    },
+    {
+      "epoch": 1.9198163733741391,
+      "grad_norm": 2.2662799826187023,
+      "learning_rate": 8.419551789322966e-08,
+      "loss": 0.293,
+      "step": 12546
+    },
+    {
+      "epoch": 1.9199693955623567,
+      "grad_norm": 2.123070326666157,
+      "learning_rate": 8.38749171502462e-08,
+      "loss": 0.2602,
+      "step": 12547
+    },
+    {
+      "epoch": 1.9201224177505738,
+      "grad_norm": 2.0546614358340296,
+      "learning_rate": 8.355492539529719e-08,
+      "loss": 0.305,
+      "step": 12548
+    },
+    {
+      "epoch": 1.9202754399387911,
+      "grad_norm": 2.072183050978695,
+      "learning_rate": 8.323554264803247e-08,
+      "loss": 0.2799,
+      "step": 12549
+    },
+    {
+      "epoch": 1.9204284621270085,
+      "grad_norm": 2.3442769629175904,
+      "learning_rate": 8.291676892806743e-08,
+      "loss": 0.3164,
+      "step": 12550
+    },
+    {
+      "epoch": 1.9205814843152256,
+      "grad_norm": 2.035394827803196,
+      "learning_rate": 8.25986042549809e-08,
+      "loss": 0.2435,
+      "step": 12551
+    },
+    {
+      "epoch": 1.9207345065034431,
+      "grad_norm": 2.1027386412155615,
+      "learning_rate": 8.228104864830943e-08,
+      "loss": 0.3025,
+      "step": 12552
+    },
+    {
+      "epoch": 1.9208875286916602,
+      "grad_norm": 2.094493796125041,
+      "learning_rate": 8.196410212755856e-08,
+      "loss": 0.2693,
+      "step": 12553
+    },
+    {
+      "epoch": 1.9210405508798776,
+      "grad_norm": 2.4802220142511264,
+      "learning_rate": 8.164776471219049e-08,
+      "loss": 0.3032,
+      "step": 12554
+    },
+    {
+      "epoch": 1.921193573068095,
+      "grad_norm": 1.9498766283903592,
+      "learning_rate": 8.133203642163523e-08,
+      "loss": 0.237,
+      "step": 12555
+    },
+    {
+      "epoch": 1.921346595256312,
+      "grad_norm": 2.2455347228710574,
+      "learning_rate": 8.10169172752806e-08,
+      "loss": 0.2767,
+      "step": 12556
+    },
+    {
+      "epoch": 1.9214996174445296,
+      "grad_norm": 1.957315184740743,
+      "learning_rate": 8.070240729248114e-08,
+      "loss": 0.2109,
+      "step": 12557
+    },
+    {
+      "epoch": 1.9216526396327467,
+      "grad_norm": 1.965643467306098,
+      "learning_rate": 8.038850649255137e-08,
+      "loss": 0.2379,
+      "step": 12558
+    },
+    {
+      "epoch": 1.921805661820964,
+      "grad_norm": 2.19852476658279,
+      "learning_rate": 8.00752148947681e-08,
+      "loss": 0.2581,
+      "step": 12559
+    },
+    {
+      "epoch": 1.9219586840091814,
+      "grad_norm": 2.004743748037509,
+      "learning_rate": 7.976253251837374e-08,
+      "loss": 0.301,
+      "step": 12560
+    },
+    {
+      "epoch": 1.9221117061973985,
+      "grad_norm": 2.2203783236108947,
+      "learning_rate": 7.945045938257067e-08,
+      "loss": 0.2943,
+      "step": 12561
+    },
+    {
+      "epoch": 1.922264728385616,
+      "grad_norm": 2.355138775293594,
+      "learning_rate": 7.91389955065236e-08,
+      "loss": 0.5327,
+      "step": 12562
+    },
+    {
+      "epoch": 1.9224177505738331,
+      "grad_norm": 2.2869360368009946,
+      "learning_rate": 7.882814090936164e-08,
+      "loss": 0.3156,
+      "step": 12563
+    },
+    {
+      "epoch": 1.9225707727620505,
+      "grad_norm": 2.269187238957002,
+      "learning_rate": 7.85178956101762e-08,
+      "loss": 0.3369,
+      "step": 12564
+    },
+    {
+      "epoch": 1.9227237949502678,
+      "grad_norm": 2.260139178110384,
+      "learning_rate": 7.82082596280187e-08,
+      "loss": 0.2689,
+      "step": 12565
+    },
+    {
+      "epoch": 1.9228768171384851,
+      "grad_norm": 2.307748324586521,
+      "learning_rate": 7.789923298190616e-08,
+      "loss": 0.317,
+      "step": 12566
+    },
+    {
+      "epoch": 1.9230298393267025,
+      "grad_norm": 1.878965997966544,
+      "learning_rate": 7.759081569081784e-08,
+      "loss": 0.2762,
+      "step": 12567
+    },
+    {
+      "epoch": 1.9231828615149196,
+      "grad_norm": 2.121729343823374,
+      "learning_rate": 7.728300777369412e-08,
+      "loss": 0.2658,
+      "step": 12568
+    },
+    {
+      "epoch": 1.923335883703137,
+      "grad_norm": 2.1638380361755796,
+      "learning_rate": 7.697580924943881e-08,
+      "loss": 0.2773,
+      "step": 12569
+    },
+    {
+      "epoch": 1.9234889058913542,
+      "grad_norm": 2.1895033980428797,
+      "learning_rate": 7.666922013691791e-08,
+      "loss": 0.1938,
+      "step": 12570
+    },
+    {
+      "epoch": 1.9236419280795716,
+      "grad_norm": 2.3956553016334556,
+      "learning_rate": 7.63632404549608e-08,
+      "loss": 0.3345,
+      "step": 12571
+    },
+    {
+      "epoch": 1.923794950267789,
+      "grad_norm": 1.7449932799207868,
+      "learning_rate": 7.605787022235912e-08,
+      "loss": 0.2398,
+      "step": 12572
+    },
+    {
+      "epoch": 1.923947972456006,
+      "grad_norm": 2.2376418795698387,
+      "learning_rate": 7.575310945786452e-08,
+      "loss": 0.284,
+      "step": 12573
+    },
+    {
+      "epoch": 1.9241009946442236,
+      "grad_norm": 1.9035723603217567,
+      "learning_rate": 7.54489581801976e-08,
+      "loss": 0.2773,
+      "step": 12574
+    },
+    {
+      "epoch": 1.9242540168324407,
+      "grad_norm": 2.2575554720495394,
+      "learning_rate": 7.514541640803342e-08,
+      "loss": 0.2984,
+      "step": 12575
+    },
+    {
+      "epoch": 1.924407039020658,
+      "grad_norm": 1.981363852999967,
+      "learning_rate": 7.484248416001594e-08,
+      "loss": 0.2508,
+      "step": 12576
+    },
+    {
+      "epoch": 1.9245600612088754,
+      "grad_norm": 1.9614331502072087,
+      "learning_rate": 7.454016145474918e-08,
+      "loss": 0.2651,
+      "step": 12577
+    },
+    {
+      "epoch": 1.9247130833970925,
+      "grad_norm": 2.1497780699564566,
+      "learning_rate": 7.423844831079941e-08,
+      "loss": 0.347,
+      "step": 12578
+    },
+    {
+      "epoch": 1.92486610558531,
+      "grad_norm": 2.240514112168765,
+      "learning_rate": 7.393734474669623e-08,
+      "loss": 0.2748,
+      "step": 12579
+    },
+    {
+      "epoch": 1.9250191277735271,
+      "grad_norm": 1.8984880370754438,
+      "learning_rate": 7.363685078093264e-08,
+      "loss": 0.2359,
+      "step": 12580
+    },
+    {
+      "epoch": 1.9251721499617445,
+      "grad_norm": 1.9864328339745811,
+      "learning_rate": 7.333696643196164e-08,
+      "loss": 0.236,
+      "step": 12581
+    },
+    {
+      "epoch": 1.9253251721499618,
+      "grad_norm": 2.234747741359613,
+      "learning_rate": 7.303769171820074e-08,
+      "loss": 0.252,
+      "step": 12582
+    },
+    {
+      "epoch": 1.925478194338179,
+      "grad_norm": 1.9754636671622134,
+      "learning_rate": 7.273902665802967e-08,
+      "loss": 0.2453,
+      "step": 12583
+    },
+    {
+      "epoch": 1.9256312165263965,
+      "grad_norm": 2.146107414003439,
+      "learning_rate": 7.244097126979044e-08,
+      "loss": 0.2792,
+      "step": 12584
+    },
+    {
+      "epoch": 1.9257842387146136,
+      "grad_norm": 1.9724035356572298,
+      "learning_rate": 7.214352557178838e-08,
+      "loss": 0.2872,
+      "step": 12585
+    },
+    {
+      "epoch": 1.925937260902831,
+      "grad_norm": 1.757835085743604,
+      "learning_rate": 7.184668958229003e-08,
+      "loss": 0.2815,
+      "step": 12586
+    },
+    {
+      "epoch": 1.9260902830910482,
+      "grad_norm": 2.1502075931636013,
+      "learning_rate": 7.15504633195263e-08,
+      "loss": 0.2656,
+      "step": 12587
+    },
+    {
+      "epoch": 1.9262433052792653,
+      "grad_norm": 2.0155875426810343,
+      "learning_rate": 7.125484680168826e-08,
+      "loss": 0.27,
+      "step": 12588
+    },
+    {
+      "epoch": 1.926396327467483,
+      "grad_norm": 1.8428482707965728,
+      "learning_rate": 7.095984004693246e-08,
+      "loss": 0.2136,
+      "step": 12589
+    },
+    {
+      "epoch": 1.9265493496557,
+      "grad_norm": 2.2304850376404213,
+      "learning_rate": 7.066544307337442e-08,
+      "loss": 0.2676,
+      "step": 12590
+    },
+    {
+      "epoch": 1.9267023718439173,
+      "grad_norm": 2.5671786865674706,
+      "learning_rate": 7.037165589909523e-08,
+      "loss": 0.316,
+      "step": 12591
+    },
+    {
+      "epoch": 1.9268553940321347,
+      "grad_norm": 2.2924701688919007,
+      "learning_rate": 7.007847854213822e-08,
+      "loss": 0.2505,
+      "step": 12592
+    },
+    {
+      "epoch": 1.9270084162203518,
+      "grad_norm": 2.0974154895266355,
+      "learning_rate": 6.978591102050791e-08,
+      "loss": 0.309,
+      "step": 12593
+    },
+    {
+      "epoch": 1.9271614384085693,
+      "grad_norm": 2.1726512258841764,
+      "learning_rate": 6.949395335217102e-08,
+      "loss": 0.3529,
+      "step": 12594
+    },
+    {
+      "epoch": 1.9273144605967865,
+      "grad_norm": 2.4202430555156775,
+      "learning_rate": 6.920260555506098e-08,
+      "loss": 0.3122,
+      "step": 12595
+    },
+    {
+      "epoch": 1.9274674827850038,
+      "grad_norm": 2.542246552788122,
+      "learning_rate": 6.891186764706681e-08,
+      "loss": 0.2913,
+      "step": 12596
+    },
+    {
+      "epoch": 1.9276205049732211,
+      "grad_norm": 2.3066907503970624,
+      "learning_rate": 6.862173964604535e-08,
+      "loss": 0.277,
+      "step": 12597
+    },
+    {
+      "epoch": 1.9277735271614385,
+      "grad_norm": 2.4028158038821417,
+      "learning_rate": 6.833222156981567e-08,
+      "loss": 0.2989,
+      "step": 12598
+    },
+    {
+      "epoch": 1.9279265493496558,
+      "grad_norm": 2.4353789679467313,
+      "learning_rate": 6.804331343615577e-08,
+      "loss": 0.2893,
+      "step": 12599
+    },
+    {
+      "epoch": 1.928079571537873,
+      "grad_norm": 2.2241183604402632,
+      "learning_rate": 6.775501526281037e-08,
+      "loss": 0.272,
+      "step": 12600
+    },
+    {
+      "epoch": 1.9282325937260902,
+      "grad_norm": 2.430754290427599,
+      "learning_rate": 6.746732706748637e-08,
+      "loss": 0.2673,
+      "step": 12601
+    },
+    {
+      "epoch": 1.9283856159143076,
+      "grad_norm": 2.281050405523614,
+      "learning_rate": 6.718024886784858e-08,
+      "loss": 0.2809,
+      "step": 12602
+    },
+    {
+      "epoch": 1.928538638102525,
+      "grad_norm": 2.1745371935601066,
+      "learning_rate": 6.689378068152841e-08,
+      "loss": 0.2604,
+      "step": 12603
+    },
+    {
+      "epoch": 1.9286916602907422,
+      "grad_norm": 2.1098243004454025,
+      "learning_rate": 6.660792252612069e-08,
+      "loss": 0.2782,
+      "step": 12604
+    },
+    {
+      "epoch": 1.9288446824789593,
+      "grad_norm": 2.298117415399452,
+      "learning_rate": 6.632267441918027e-08,
+      "loss": 0.2785,
+      "step": 12605
+    },
+    {
+      "epoch": 1.928997704667177,
+      "grad_norm": 2.486226541978198,
+      "learning_rate": 6.603803637822537e-08,
+      "loss": 0.3168,
+      "step": 12606
+    },
+    {
+      "epoch": 1.929150726855394,
+      "grad_norm": 2.236276600745981,
+      "learning_rate": 6.575400842073643e-08,
+      "loss": 0.293,
+      "step": 12607
+    },
+    {
+      "epoch": 1.9293037490436113,
+      "grad_norm": 2.3638793794792217,
+      "learning_rate": 6.547059056415617e-08,
+      "loss": 0.2954,
+      "step": 12608
+    },
+    {
+      "epoch": 1.9294567712318287,
+      "grad_norm": 2.2428113757322494,
+      "learning_rate": 6.518778282589067e-08,
+      "loss": 0.2923,
+      "step": 12609
+    },
+    {
+      "epoch": 1.9296097934200458,
+      "grad_norm": 2.5389227577644284,
+      "learning_rate": 6.490558522330936e-08,
+      "loss": 0.3074,
+      "step": 12610
+    },
+    {
+      "epoch": 1.9297628156082633,
+      "grad_norm": 2.090509434099971,
+      "learning_rate": 6.462399777374284e-08,
+      "loss": 0.2688,
+      "step": 12611
+    },
+    {
+      "epoch": 1.9299158377964805,
+      "grad_norm": 2.1129234004043957,
+      "learning_rate": 6.434302049448393e-08,
+      "loss": 0.2829,
+      "step": 12612
+    },
+    {
+      "epoch": 1.9300688599846978,
+      "grad_norm": 1.8011178532916772,
+      "learning_rate": 6.406265340278772e-08,
+      "loss": 0.2099,
+      "step": 12613
+    },
+    {
+      "epoch": 1.9302218821729151,
+      "grad_norm": 2.264044979485212,
+      "learning_rate": 6.378289651587377e-08,
+      "loss": 0.2486,
+      "step": 12614
+    },
+    {
+      "epoch": 1.9303749043611322,
+      "grad_norm": 2.1444665625136534,
+      "learning_rate": 6.350374985092389e-08,
+      "loss": 0.2625,
+      "step": 12615
+    },
+    {
+      "epoch": 1.9305279265493498,
+      "grad_norm": 1.8452951812548426,
+      "learning_rate": 6.322521342507993e-08,
+      "loss": 0.2091,
+      "step": 12616
+    },
+    {
+      "epoch": 1.930680948737567,
+      "grad_norm": 2.2619235464051566,
+      "learning_rate": 6.294728725544708e-08,
+      "loss": 0.2916,
+      "step": 12617
+    },
+    {
+      "epoch": 1.9308339709257842,
+      "grad_norm": 1.8293153143516894,
+      "learning_rate": 6.266997135909725e-08,
+      "loss": 0.2575,
+      "step": 12618
+    },
+    {
+      "epoch": 1.9309869931140016,
+      "grad_norm": 1.9826796507956632,
+      "learning_rate": 6.239326575305904e-08,
+      "loss": 0.2715,
+      "step": 12619
+    },
+    {
+      "epoch": 1.9311400153022187,
+      "grad_norm": 2.3014999251705257,
+      "learning_rate": 6.211717045432553e-08,
+      "loss": 0.3369,
+      "step": 12620
+    },
+    {
+      "epoch": 1.9312930374904362,
+      "grad_norm": 2.2258466809607436,
+      "learning_rate": 6.184168547985536e-08,
+      "loss": 0.299,
+      "step": 12621
+    },
+    {
+      "epoch": 1.9314460596786533,
+      "grad_norm": 1.994257024190829,
+      "learning_rate": 6.156681084656501e-08,
+      "loss": 0.2588,
+      "step": 12622
+    },
+    {
+      "epoch": 1.9315990818668707,
+      "grad_norm": 2.1592430726452148,
+      "learning_rate": 6.129254657133544e-08,
+      "loss": 0.2516,
+      "step": 12623
+    },
+    {
+      "epoch": 1.931752104055088,
+      "grad_norm": 1.931950663977846,
+      "learning_rate": 6.101889267101201e-08,
+      "loss": 0.2307,
+      "step": 12624
+    },
+    {
+      "epoch": 1.9319051262433051,
+      "grad_norm": 2.3942192745722406,
+      "learning_rate": 6.074584916239911e-08,
+      "loss": 0.3507,
+      "step": 12625
+    },
+    {
+      "epoch": 1.9320581484315227,
+      "grad_norm": 2.098354148192397,
+      "learning_rate": 6.047341606226664e-08,
+      "loss": 0.3061,
+      "step": 12626
+    },
+    {
+      "epoch": 1.9322111706197398,
+      "grad_norm": 2.163742519632644,
+      "learning_rate": 6.020159338734566e-08,
+      "loss": 0.2322,
+      "step": 12627
+    },
+    {
+      "epoch": 1.9323641928079571,
+      "grad_norm": 2.3218334894180708,
+      "learning_rate": 5.993038115432837e-08,
+      "loss": 0.3174,
+      "step": 12628
+    },
+    {
+      "epoch": 1.9325172149961745,
+      "grad_norm": 1.7732066918572658,
+      "learning_rate": 5.965977937987366e-08,
+      "loss": 0.2302,
+      "step": 12629
+    },
+    {
+      "epoch": 1.9326702371843916,
+      "grad_norm": 2.000109135862685,
+      "learning_rate": 5.938978808059825e-08,
+      "loss": 0.2409,
+      "step": 12630
+    },
+    {
+      "epoch": 1.9328232593726091,
+      "grad_norm": 2.333789338782402,
+      "learning_rate": 5.9120407273082215e-08,
+      "loss": 0.345,
+      "step": 12631
+    },
+    {
+      "epoch": 1.9329762815608262,
+      "grad_norm": 1.90728146018191,
+      "learning_rate": 5.8851636973872306e-08,
+      "loss": 0.2432,
+      "step": 12632
+    },
+    {
+      "epoch": 1.9331293037490436,
+      "grad_norm": 2.208001178052822,
+      "learning_rate": 5.8583477199474213e-08,
+      "loss": 0.294,
+      "step": 12633
+    },
+    {
+      "epoch": 1.933282325937261,
+      "grad_norm": 2.354523824473684,
+      "learning_rate": 5.831592796635477e-08,
+      "loss": 0.3385,
+      "step": 12634
+    },
+    {
+      "epoch": 1.9334353481254782,
+      "grad_norm": 1.8312645826200542,
+      "learning_rate": 5.804898929094638e-08,
+      "loss": 0.2109,
+      "step": 12635
+    },
+    {
+      "epoch": 1.9335883703136956,
+      "grad_norm": 2.0830635861639366,
+      "learning_rate": 5.77826611896426e-08,
+      "loss": 0.2663,
+      "step": 12636
+    },
+    {
+      "epoch": 1.9337413925019127,
+      "grad_norm": 2.419482269865649,
+      "learning_rate": 5.751694367879923e-08,
+      "loss": 0.2717,
+      "step": 12637
+    },
+    {
+      "epoch": 1.93389441469013,
+      "grad_norm": 2.0657515526693464,
+      "learning_rate": 5.725183677473656e-08,
+      "loss": 0.2888,
+      "step": 12638
+    },
+    {
+      "epoch": 1.9340474368783473,
+      "grad_norm": 2.0596978612467565,
+      "learning_rate": 5.698734049373489e-08,
+      "loss": 0.2544,
+      "step": 12639
+    },
+    {
+      "epoch": 1.9342004590665647,
+      "grad_norm": 2.5250076265493955,
+      "learning_rate": 5.6723454852036785e-08,
+      "loss": 0.3538,
+      "step": 12640
+    },
+    {
+      "epoch": 1.934353481254782,
+      "grad_norm": 2.2943276103623536,
+      "learning_rate": 5.64601798658504e-08,
+      "loss": 0.2966,
+      "step": 12641
+    },
+    {
+      "epoch": 1.9345065034429991,
+      "grad_norm": 2.1423583068445087,
+      "learning_rate": 5.6197515551343895e-08,
+      "loss": 0.3187,
+      "step": 12642
+    },
+    {
+      "epoch": 1.9346595256312167,
+      "grad_norm": 2.123351982906415,
+      "learning_rate": 5.593546192464883e-08,
+      "loss": 0.2805,
+      "step": 12643
+    },
+    {
+      "epoch": 1.9348125478194338,
+      "grad_norm": 2.1385231757290257,
+      "learning_rate": 5.567401900185787e-08,
+      "loss": 0.2985,
+      "step": 12644
+    },
+    {
+      "epoch": 1.9349655700076511,
+      "grad_norm": 2.283449728780997,
+      "learning_rate": 5.541318679902707e-08,
+      "loss": 0.3114,
+      "step": 12645
+    },
+    {
+      "epoch": 1.9351185921958685,
+      "grad_norm": 1.9670750492646343,
+      "learning_rate": 5.515296533217696e-08,
+      "loss": 0.2487,
+      "step": 12646
+    },
+    {
+      "epoch": 1.9352716143840856,
+      "grad_norm": 2.095455321897977,
+      "learning_rate": 5.4893354617286954e-08,
+      "loss": 0.2793,
+      "step": 12647
+    },
+    {
+      "epoch": 1.9354246365723031,
+      "grad_norm": 1.9765462462707317,
+      "learning_rate": 5.4634354670302095e-08,
+      "loss": 0.3064,
+      "step": 12648
+    },
+    {
+      "epoch": 1.9355776587605202,
+      "grad_norm": 1.8820560851270636,
+      "learning_rate": 5.437596550712743e-08,
+      "loss": 0.2657,
+      "step": 12649
+    },
+    {
+      "epoch": 1.9357306809487376,
+      "grad_norm": 2.3788353271474634,
+      "learning_rate": 5.411818714363248e-08,
+      "loss": 0.3056,
+      "step": 12650
+    },
+    {
+      "epoch": 1.935883703136955,
+      "grad_norm": 2.2261586061883953,
+      "learning_rate": 5.3861019595649045e-08,
+      "loss": 0.3066,
+      "step": 12651
+    },
+    {
+      "epoch": 1.936036725325172,
+      "grad_norm": 1.785826281369035,
+      "learning_rate": 5.360446287896781e-08,
+      "loss": 0.2866,
+      "step": 12652
+    },
+    {
+      "epoch": 1.9361897475133896,
+      "grad_norm": 2.331191024955345,
+      "learning_rate": 5.334851700934951e-08,
+      "loss": 0.3063,
+      "step": 12653
+    },
+    {
+      "epoch": 1.9363427697016067,
+      "grad_norm": 1.8719935808142152,
+      "learning_rate": 5.309318200250824e-08,
+      "loss": 0.2523,
+      "step": 12654
+    },
+    {
+      "epoch": 1.936495791889824,
+      "grad_norm": 2.223253250317612,
+      "learning_rate": 5.283845787412811e-08,
+      "loss": 0.3068,
+      "step": 12655
+    },
+    {
+      "epoch": 1.9366488140780413,
+      "grad_norm": 2.1951752055550853,
+      "learning_rate": 5.258434463985218e-08,
+      "loss": 0.2619,
+      "step": 12656
+    },
+    {
+      "epoch": 1.9368018362662585,
+      "grad_norm": 2.059501912533121,
+      "learning_rate": 5.233084231528574e-08,
+      "loss": 0.2819,
+      "step": 12657
+    },
+    {
+      "epoch": 1.936954858454476,
+      "grad_norm": 2.2992385321273896,
+      "learning_rate": 5.207795091599743e-08,
+      "loss": 0.2655,
+      "step": 12658
+    },
+    {
+      "epoch": 1.9371078806426931,
+      "grad_norm": 2.034712783110768,
+      "learning_rate": 5.182567045751929e-08,
+      "loss": 0.2648,
+      "step": 12659
+    },
+    {
+      "epoch": 1.9372609028309105,
+      "grad_norm": 2.0115061960293747,
+      "learning_rate": 5.157400095534337e-08,
+      "loss": 0.2186,
+      "step": 12660
+    },
+    {
+      "epoch": 1.9374139250191278,
+      "grad_norm": 1.9064111987400696,
+      "learning_rate": 5.1322942424927303e-08,
+      "loss": 0.2377,
+      "step": 12661
+    },
+    {
+      "epoch": 1.937566947207345,
+      "grad_norm": 2.1764256108666236,
+      "learning_rate": 5.107249488168875e-08,
+      "loss": 0.247,
+      "step": 12662
+    },
+    {
+      "epoch": 1.9377199693955625,
+      "grad_norm": 1.7123673286799752,
+      "learning_rate": 5.082265834100875e-08,
+      "loss": 0.2426,
+      "step": 12663
+    },
+    {
+      "epoch": 1.9378729915837796,
+      "grad_norm": 2.3757105271033483,
+      "learning_rate": 5.0573432818229464e-08,
+      "loss": 0.2783,
+      "step": 12664
+    },
+    {
+      "epoch": 1.938026013771997,
+      "grad_norm": 2.0579111088781223,
+      "learning_rate": 5.0324818328659766e-08,
+      "loss": 0.2608,
+      "step": 12665
+    },
+    {
+      "epoch": 1.9381790359602142,
+      "grad_norm": 2.160905591204374,
+      "learning_rate": 5.007681488756522e-08,
+      "loss": 0.2706,
+      "step": 12666
+    },
+    {
+      "epoch": 1.9383320581484316,
+      "grad_norm": 2.2877271140280393,
+      "learning_rate": 4.982942251017808e-08,
+      "loss": 0.2357,
+      "step": 12667
+    },
+    {
+      "epoch": 1.938485080336649,
+      "grad_norm": 2.1542809755239216,
+      "learning_rate": 4.958264121169176e-08,
+      "loss": 0.2625,
+      "step": 12668
+    },
+    {
+      "epoch": 1.938638102524866,
+      "grad_norm": 2.059582513808786,
+      "learning_rate": 4.93364710072608e-08,
+      "loss": 0.2883,
+      "step": 12669
+    },
+    {
+      "epoch": 1.9387911247130833,
+      "grad_norm": 2.1724063331417636,
+      "learning_rate": 4.9090911912005325e-08,
+      "loss": 0.2804,
+      "step": 12670
+    },
+    {
+      "epoch": 1.9389441469013007,
+      "grad_norm": 2.076507363841404,
+      "learning_rate": 4.884596394100438e-08,
+      "loss": 0.2798,
+      "step": 12671
+    },
+    {
+      "epoch": 1.939097169089518,
+      "grad_norm": 2.1280474655352286,
+      "learning_rate": 4.8601627109302605e-08,
+      "loss": 0.3037,
+      "step": 12672
+    },
+    {
+      "epoch": 1.9392501912777353,
+      "grad_norm": 2.191630607477236,
+      "learning_rate": 4.835790143190466e-08,
+      "loss": 0.2652,
+      "step": 12673
+    },
+    {
+      "epoch": 1.9394032134659525,
+      "grad_norm": 2.017617923200027,
+      "learning_rate": 4.8114786923778575e-08,
+      "loss": 0.2702,
+      "step": 12674
+    },
+    {
+      "epoch": 1.93955623565417,
+      "grad_norm": 2.1656108843312754,
+      "learning_rate": 4.787228359985685e-08,
+      "loss": 0.2703,
+      "step": 12675
+    },
+    {
+      "epoch": 1.9397092578423871,
+      "grad_norm": 2.056305398421697,
+      "learning_rate": 4.7630391475029794e-08,
+      "loss": 0.2748,
+      "step": 12676
+    },
+    {
+      "epoch": 1.9398622800306045,
+      "grad_norm": 2.0212531782535934,
+      "learning_rate": 4.7389110564154407e-08,
+      "loss": 0.2648,
+      "step": 12677
+    },
+    {
+      "epoch": 1.9400153022188218,
+      "grad_norm": 2.183826196568856,
+      "learning_rate": 4.714844088204884e-08,
+      "loss": 0.2861,
+      "step": 12678
+    },
+    {
+      "epoch": 1.940168324407039,
+      "grad_norm": 1.9595429105828446,
+      "learning_rate": 4.6908382443493496e-08,
+      "loss": 0.2806,
+      "step": 12679
+    },
+    {
+      "epoch": 1.9403213465952565,
+      "grad_norm": 2.104117403711713,
+      "learning_rate": 4.666893526323102e-08,
+      "loss": 0.3612,
+      "step": 12680
+    },
+    {
+      "epoch": 1.9404743687834736,
+      "grad_norm": 2.3085008583768447,
+      "learning_rate": 4.643009935596632e-08,
+      "loss": 0.3073,
+      "step": 12681
+    },
+    {
+      "epoch": 1.940627390971691,
+      "grad_norm": 2.146451918561183,
+      "learning_rate": 4.6191874736367656e-08,
+      "loss": 0.2558,
+      "step": 12682
+    },
+    {
+      "epoch": 1.9407804131599082,
+      "grad_norm": 2.3192462362711734,
+      "learning_rate": 4.595426141906667e-08,
+      "loss": 0.3116,
+      "step": 12683
+    },
+    {
+      "epoch": 1.9409334353481253,
+      "grad_norm": 2.4574745152024793,
+      "learning_rate": 4.57172594186539e-08,
+      "loss": 0.2959,
+      "step": 12684
+    },
+    {
+      "epoch": 1.941086457536343,
+      "grad_norm": 1.9916261660219228,
+      "learning_rate": 4.54808687496866e-08,
+      "loss": 0.2328,
+      "step": 12685
+    },
+    {
+      "epoch": 1.94123947972456,
+      "grad_norm": 1.8992522773265608,
+      "learning_rate": 4.5245089426680934e-08,
+      "loss": 0.2541,
+      "step": 12686
+    },
+    {
+      "epoch": 1.9413925019127773,
+      "grad_norm": 1.8079036077768846,
+      "learning_rate": 4.500992146411753e-08,
+      "loss": 0.2135,
+      "step": 12687
+    },
+    {
+      "epoch": 1.9415455241009947,
+      "grad_norm": 1.9889511801750943,
+      "learning_rate": 4.4775364876438185e-08,
+      "loss": 0.2427,
+      "step": 12688
+    },
+    {
+      "epoch": 1.9416985462892118,
+      "grad_norm": 1.751479263129326,
+      "learning_rate": 4.454141967805026e-08,
+      "loss": 0.2189,
+      "step": 12689
+    },
+    {
+      "epoch": 1.9418515684774293,
+      "grad_norm": 2.0423292553879957,
+      "learning_rate": 4.430808588331892e-08,
+      "loss": 0.257,
+      "step": 12690
+    },
+    {
+      "epoch": 1.9420045906656465,
+      "grad_norm": 2.0041499310977504,
+      "learning_rate": 4.407536350657493e-08,
+      "loss": 0.2035,
+      "step": 12691
+    },
+    {
+      "epoch": 1.9421576128538638,
+      "grad_norm": 2.312755751955661,
+      "learning_rate": 4.3843252562110193e-08,
+      "loss": 0.2828,
+      "step": 12692
+    },
+    {
+      "epoch": 1.9423106350420811,
+      "grad_norm": 1.971954947422424,
+      "learning_rate": 4.361175306418109e-08,
+      "loss": 0.2417,
+      "step": 12693
+    },
+    {
+      "epoch": 1.9424636572302982,
+      "grad_norm": 1.831074912801939,
+      "learning_rate": 4.33808650270029e-08,
+      "loss": 0.2528,
+      "step": 12694
+    },
+    {
+      "epoch": 1.9426166794185158,
+      "grad_norm": 2.0578830783806525,
+      "learning_rate": 4.3150588464756504e-08,
+      "loss": 0.2771,
+      "step": 12695
+    },
+    {
+      "epoch": 1.942769701606733,
+      "grad_norm": 1.961835942990468,
+      "learning_rate": 4.2920923391583937e-08,
+      "loss": 0.221,
+      "step": 12696
+    },
+    {
+      "epoch": 1.9429227237949502,
+      "grad_norm": 1.9705490121987508,
+      "learning_rate": 4.269186982159057e-08,
+      "loss": 0.2571,
+      "step": 12697
+    },
+    {
+      "epoch": 1.9430757459831676,
+      "grad_norm": 2.0673965559846312,
+      "learning_rate": 4.246342776884182e-08,
+      "loss": 0.2342,
+      "step": 12698
+    },
+    {
+      "epoch": 1.943228768171385,
+      "grad_norm": 2.027833565441114,
+      "learning_rate": 4.223559724736759e-08,
+      "loss": 0.2193,
+      "step": 12699
+    },
+    {
+      "epoch": 1.9433817903596022,
+      "grad_norm": 1.928042773055958,
+      "learning_rate": 4.200837827116111e-08,
+      "loss": 0.2249,
+      "step": 12700
+    },
+    {
+      "epoch": 1.9435348125478193,
+      "grad_norm": 2.0145095411771186,
+      "learning_rate": 4.1781770854175674e-08,
+      "loss": 0.2194,
+      "step": 12701
+    },
+    {
+      "epoch": 1.9436878347360367,
+      "grad_norm": 2.0755925996216606,
+      "learning_rate": 4.1555775010329034e-08,
+      "loss": 0.2577,
+      "step": 12702
+    },
+    {
+      "epoch": 1.943840856924254,
+      "grad_norm": 2.1088247690311714,
+      "learning_rate": 4.1330390753500094e-08,
+      "loss": 0.28,
+      "step": 12703
+    },
+    {
+      "epoch": 1.9439938791124713,
+      "grad_norm": 2.070960100346576,
+      "learning_rate": 4.110561809752889e-08,
+      "loss": 0.2509,
+      "step": 12704
+    },
+    {
+      "epoch": 1.9441469013006887,
+      "grad_norm": 2.0891100512185603,
+      "learning_rate": 4.0881457056222153e-08,
+      "loss": 0.2513,
+      "step": 12705
+    },
+    {
+      "epoch": 1.9442999234889058,
+      "grad_norm": 2.5962353577847566,
+      "learning_rate": 4.065790764334554e-08,
+      "loss": 0.3419,
+      "step": 12706
+    },
+    {
+      "epoch": 1.9444529456771233,
+      "grad_norm": 1.9640071319967864,
+      "learning_rate": 4.043496987262696e-08,
+      "loss": 0.2708,
+      "step": 12707
+    },
+    {
+      "epoch": 1.9446059678653405,
+      "grad_norm": 2.179001565244998,
+      "learning_rate": 4.021264375775991e-08,
+      "loss": 0.3385,
+      "step": 12708
+    },
+    {
+      "epoch": 1.9447589900535578,
+      "grad_norm": 2.1426220088913954,
+      "learning_rate": 3.99909293123979e-08,
+      "loss": 0.2687,
+      "step": 12709
+    },
+    {
+      "epoch": 1.9449120122417751,
+      "grad_norm": 1.8754076871641312,
+      "learning_rate": 3.97698265501556e-08,
+      "loss": 0.2513,
+      "step": 12710
+    },
+    {
+      "epoch": 1.9450650344299922,
+      "grad_norm": 2.042815422922247,
+      "learning_rate": 3.954933548461326e-08,
+      "loss": 0.2162,
+      "step": 12711
+    },
+    {
+      "epoch": 1.9452180566182098,
+      "grad_norm": 1.8552967583474373,
+      "learning_rate": 3.932945612931227e-08,
+      "loss": 0.2369,
+      "step": 12712
+    },
+    {
+      "epoch": 1.945371078806427,
+      "grad_norm": 1.907696643956236,
+      "learning_rate": 3.911018849775405e-08,
+      "loss": 0.2624,
+      "step": 12713
+    },
+    {
+      "epoch": 1.9455241009946442,
+      "grad_norm": 2.0798771441829116,
+      "learning_rate": 3.8891532603407834e-08,
+      "loss": 0.2829,
+      "step": 12714
+    },
+    {
+      "epoch": 1.9456771231828616,
+      "grad_norm": 2.3790723531215257,
+      "learning_rate": 3.867348845970065e-08,
+      "loss": 0.3033,
+      "step": 12715
+    },
+    {
+      "epoch": 1.9458301453710787,
+      "grad_norm": 1.8852530500177367,
+      "learning_rate": 3.84560560800229e-08,
+      "loss": 0.228,
+      "step": 12716
+    },
+    {
+      "epoch": 1.9459831675592962,
+      "grad_norm": 2.1820151584706613,
+      "learning_rate": 3.8239235477728344e-08,
+      "loss": 0.3082,
+      "step": 12717
+    },
+    {
+      "epoch": 1.9461361897475133,
+      "grad_norm": 2.2140014660667933,
+      "learning_rate": 3.8023026666133e-08,
+      "loss": 0.2537,
+      "step": 12718
+    },
+    {
+      "epoch": 1.9462892119357307,
+      "grad_norm": 2.2659188032539195,
+      "learning_rate": 3.7807429658516246e-08,
+      "loss": 0.2632,
+      "step": 12719
+    },
+    {
+      "epoch": 1.946442234123948,
+      "grad_norm": 2.5293218796462464,
+      "learning_rate": 3.7592444468116384e-08,
+      "loss": 0.2743,
+      "step": 12720
+    },
+    {
+      "epoch": 1.9465952563121651,
+      "grad_norm": 1.8671996561905655,
+      "learning_rate": 3.7378071108138405e-08,
+      "loss": 0.2409,
+      "step": 12721
+    },
+    {
+      "epoch": 1.9467482785003827,
+      "grad_norm": 2.239141860501856,
+      "learning_rate": 3.716430959174622e-08,
+      "loss": 0.3049,
+      "step": 12722
+    },
+    {
+      "epoch": 1.9469013006885998,
+      "grad_norm": 1.909309810127325,
+      "learning_rate": 3.6951159932069323e-08,
+      "loss": 0.2702,
+      "step": 12723
+    },
+    {
+      "epoch": 1.9470543228768171,
+      "grad_norm": 2.1027549057800097,
+      "learning_rate": 3.673862214219726e-08,
+      "loss": 0.2468,
+      "step": 12724
+    },
+    {
+      "epoch": 1.9472073450650345,
+      "grad_norm": 1.864832404974843,
+      "learning_rate": 3.65266962351829e-08,
+      "loss": 0.223,
+      "step": 12725
+    },
+    {
+      "epoch": 1.9473603672532516,
+      "grad_norm": 1.7866644661828868,
+      "learning_rate": 3.6315382224041406e-08,
+      "loss": 0.2208,
+      "step": 12726
+    },
+    {
+      "epoch": 1.9475133894414691,
+      "grad_norm": 2.125342586374641,
+      "learning_rate": 3.610468012175017e-08,
+      "loss": 0.3016,
+      "step": 12727
+    },
+    {
+      "epoch": 1.9476664116296862,
+      "grad_norm": 2.676246515950065,
+      "learning_rate": 3.5894589941248835e-08,
+      "loss": 0.3001,
+      "step": 12728
+    },
+    {
+      "epoch": 1.9478194338179036,
+      "grad_norm": 2.1133835150212024,
+      "learning_rate": 3.568511169544153e-08,
+      "loss": 0.3017,
+      "step": 12729
+    },
+    {
+      "epoch": 1.947972456006121,
+      "grad_norm": 1.9531071847013737,
+      "learning_rate": 3.547624539719241e-08,
+      "loss": 0.2712,
+      "step": 12730
+    },
+    {
+      "epoch": 1.9481254781943382,
+      "grad_norm": 2.3816136624101403,
+      "learning_rate": 3.526799105932788e-08,
+      "loss": 0.3098,
+      "step": 12731
+    },
+    {
+      "epoch": 1.9482785003825556,
+      "grad_norm": 2.0472889114981436,
+      "learning_rate": 3.506034869463881e-08,
+      "loss": 0.2275,
+      "step": 12732
+    },
+    {
+      "epoch": 1.9484315225707727,
+      "grad_norm": 2.005149594082394,
+      "learning_rate": 3.4853318315876126e-08,
+      "loss": 0.28,
+      "step": 12733
+    },
+    {
+      "epoch": 1.94858454475899,
+      "grad_norm": 2.307715143977197,
+      "learning_rate": 3.4646899935755205e-08,
+      "loss": 0.3267,
+      "step": 12734
+    },
+    {
+      "epoch": 1.9487375669472073,
+      "grad_norm": 2.6018657290133205,
+      "learning_rate": 3.4441093566953685e-08,
+      "loss": 0.2484,
+      "step": 12735
+    },
+    {
+      "epoch": 1.9488905891354247,
+      "grad_norm": 2.2228026058532255,
+      "learning_rate": 3.4235899222108126e-08,
+      "loss": 0.2843,
+      "step": 12736
+    },
+    {
+      "epoch": 1.949043611323642,
+      "grad_norm": 2.278609420116441,
+      "learning_rate": 3.4031316913824e-08,
+      "loss": 0.2884,
+      "step": 12737
+    },
+    {
+      "epoch": 1.9491966335118591,
+      "grad_norm": 2.3154914775864954,
+      "learning_rate": 3.3827346654663473e-08,
+      "loss": 0.324,
+      "step": 12738
+    },
+    {
+      "epoch": 1.9493496557000765,
+      "grad_norm": 2.6375485637022438,
+      "learning_rate": 3.36239884571532e-08,
+      "loss": 0.2916,
+      "step": 12739
+    },
+    {
+      "epoch": 1.9495026778882938,
+      "grad_norm": 1.9118207499010522,
+      "learning_rate": 3.342124233378319e-08,
+      "loss": 0.2729,
+      "step": 12740
+    },
+    {
+      "epoch": 1.9496557000765111,
+      "grad_norm": 1.9701216325844941,
+      "learning_rate": 3.3219108297003476e-08,
+      "loss": 0.2535,
+      "step": 12741
+    },
+    {
+      "epoch": 1.9498087222647285,
+      "grad_norm": 2.2188220603731286,
+      "learning_rate": 3.3017586359228584e-08,
+      "loss": 0.2926,
+      "step": 12742
+    },
+    {
+      "epoch": 1.9499617444529456,
+      "grad_norm": 1.828588995912328,
+      "learning_rate": 3.281667653283416e-08,
+      "loss": 0.2301,
+      "step": 12743
+    },
+    {
+      "epoch": 1.9501147666411631,
+      "grad_norm": 2.072114708886476,
+      "learning_rate": 3.2616378830160336e-08,
+      "loss": 0.2664,
+      "step": 12744
+    },
+    {
+      "epoch": 1.9502677888293802,
+      "grad_norm": 2.09207507253338,
+      "learning_rate": 3.241669326350727e-08,
+      "loss": 0.2392,
+      "step": 12745
+    },
+    {
+      "epoch": 1.9504208110175976,
+      "grad_norm": 2.222014245372224,
+      "learning_rate": 3.221761984513849e-08,
+      "loss": 0.3022,
+      "step": 12746
+    },
+    {
+      "epoch": 1.950573833205815,
+      "grad_norm": 2.4473576328151823,
+      "learning_rate": 3.2019158587279773e-08,
+      "loss": 0.3092,
+      "step": 12747
+    },
+    {
+      "epoch": 1.950726855394032,
+      "grad_norm": 2.1372539044169256,
+      "learning_rate": 3.1821309502119146e-08,
+      "loss": 0.2586,
+      "step": 12748
+    },
+    {
+      "epoch": 1.9508798775822496,
+      "grad_norm": 2.2224469014528383,
+      "learning_rate": 3.1624072601808e-08,
+      "loss": 0.3213,
+      "step": 12749
+    },
+    {
+      "epoch": 1.9510328997704667,
+      "grad_norm": 2.268220714282305,
+      "learning_rate": 3.142744789845997e-08,
+      "loss": 0.2969,
+      "step": 12750
+    },
+    {
+      "epoch": 1.951185921958684,
+      "grad_norm": 2.2497531369255426,
+      "learning_rate": 3.123143540414875e-08,
+      "loss": 0.2996,
+      "step": 12751
+    },
+    {
+      "epoch": 1.9513389441469013,
+      "grad_norm": 2.0618539697503935,
+      "learning_rate": 3.1036035130912464e-08,
+      "loss": 0.3029,
+      "step": 12752
+    },
+    {
+      "epoch": 1.9514919663351185,
+      "grad_norm": 1.9194248768655344,
+      "learning_rate": 3.0841247090752646e-08,
+      "loss": 0.2567,
+      "step": 12753
+    },
+    {
+      "epoch": 1.951644988523336,
+      "grad_norm": 2.2237292380811384,
+      "learning_rate": 3.064707129563194e-08,
+      "loss": 0.2751,
+      "step": 12754
+    },
+    {
+      "epoch": 1.9517980107115531,
+      "grad_norm": 2.1877555347738062,
+      "learning_rate": 3.045350775747413e-08,
+      "loss": 0.2889,
+      "step": 12755
+    },
+    {
+      "epoch": 1.9519510328997705,
+      "grad_norm": 2.1896793457465074,
+      "learning_rate": 3.02605564881675e-08,
+      "loss": 0.2728,
+      "step": 12756
+    },
+    {
+      "epoch": 1.9521040550879878,
+      "grad_norm": 1.6954979190860753,
+      "learning_rate": 3.006821749956146e-08,
+      "loss": 0.2552,
+      "step": 12757
+    },
+    {
+      "epoch": 1.952257077276205,
+      "grad_norm": 2.1512376385993712,
+      "learning_rate": 2.9876490803469884e-08,
+      "loss": 0.2612,
+      "step": 12758
+    },
+    {
+      "epoch": 1.9524100994644225,
+      "grad_norm": 2.4529291510609856,
+      "learning_rate": 2.968537641166669e-08,
+      "loss": 0.3468,
+      "step": 12759
+    },
+    {
+      "epoch": 1.9525631216526396,
+      "grad_norm": 2.25080392769247,
+      "learning_rate": 2.9494874335886935e-08,
+      "loss": 0.2853,
+      "step": 12760
+    },
+    {
+      "epoch": 1.952716143840857,
+      "grad_norm": 2.520938479932384,
+      "learning_rate": 2.9304984587833485e-08,
+      "loss": 0.3053,
+      "step": 12761
+    },
+    {
+      "epoch": 1.9528691660290742,
+      "grad_norm": 2.014803782393137,
+      "learning_rate": 2.9115707179165898e-08,
+      "loss": 0.2684,
+      "step": 12762
+    },
+    {
+      "epoch": 1.9530221882172913,
+      "grad_norm": 2.2036033789059104,
+      "learning_rate": 2.8927042121508208e-08,
+      "loss": 0.2638,
+      "step": 12763
+    },
+    {
+      "epoch": 1.953175210405509,
+      "grad_norm": 2.2003394728145618,
+      "learning_rate": 2.8738989426448928e-08,
+      "loss": 0.3004,
+      "step": 12764
+    },
+    {
+      "epoch": 1.953328232593726,
+      "grad_norm": 1.9303823675832934,
+      "learning_rate": 2.855154910553548e-08,
+      "loss": 0.2712,
+      "step": 12765
+    },
+    {
+      "epoch": 1.9534812547819433,
+      "grad_norm": 2.1463555044250775,
+      "learning_rate": 2.8364721170280883e-08,
+      "loss": 0.2339,
+      "step": 12766
+    },
+    {
+      "epoch": 1.9536342769701607,
+      "grad_norm": 2.2150266187516934,
+      "learning_rate": 2.817850563215707e-08,
+      "loss": 0.3033,
+      "step": 12767
+    },
+    {
+      "epoch": 1.953787299158378,
+      "grad_norm": 2.258512569394117,
+      "learning_rate": 2.799290250260156e-08,
+      "loss": 0.2643,
+      "step": 12768
+    },
+    {
+      "epoch": 1.9539403213465953,
+      "grad_norm": 2.300631573365148,
+      "learning_rate": 2.780791179301301e-08,
+      "loss": 0.3314,
+      "step": 12769
+    },
+    {
+      "epoch": 1.9540933435348125,
+      "grad_norm": 2.148802915378045,
+      "learning_rate": 2.7623533514751223e-08,
+      "loss": 0.2894,
+      "step": 12770
+    },
+    {
+      "epoch": 1.9542463657230298,
+      "grad_norm": 2.5611381404597,
+      "learning_rate": 2.7439767679140474e-08,
+      "loss": 0.2613,
+      "step": 12771
+    },
+    {
+      "epoch": 1.9543993879112471,
+      "grad_norm": 1.7155427881312075,
+      "learning_rate": 2.7256614297467287e-08,
+      "loss": 0.2885,
+      "step": 12772
+    },
+    {
+      "epoch": 1.9545524100994645,
+      "grad_norm": 1.8547576958237793,
+      "learning_rate": 2.707407338097823e-08,
+      "loss": 0.253,
+      "step": 12773
+    },
+    {
+      "epoch": 1.9547054322876818,
+      "grad_norm": 2.401444734532838,
+      "learning_rate": 2.6892144940885435e-08,
+      "loss": 0.3416,
+      "step": 12774
+    },
+    {
+      "epoch": 1.954858454475899,
+      "grad_norm": 2.3311045666651102,
+      "learning_rate": 2.6710828988359973e-08,
+      "loss": 0.3475,
+      "step": 12775
+    },
+    {
+      "epoch": 1.9550114766641165,
+      "grad_norm": 2.233725116048879,
+      "learning_rate": 2.653012553453849e-08,
+      "loss": 0.2838,
+      "step": 12776
+    },
+    {
+      "epoch": 1.9551644988523336,
+      "grad_norm": 2.06475531213535,
+      "learning_rate": 2.635003459051877e-08,
+      "loss": 0.3182,
+      "step": 12777
+    },
+    {
+      "epoch": 1.955317521040551,
+      "grad_norm": 2.2259592026875135,
+      "learning_rate": 2.617055616736086e-08,
+      "loss": 0.2861,
+      "step": 12778
+    },
+    {
+      "epoch": 1.9554705432287682,
+      "grad_norm": 2.3905401922864216,
+      "learning_rate": 2.5991690276087056e-08,
+      "loss": 0.297,
+      "step": 12779
+    },
+    {
+      "epoch": 1.9556235654169853,
+      "grad_norm": 1.7548109646120047,
+      "learning_rate": 2.581343692768079e-08,
+      "loss": 0.1689,
+      "step": 12780
+    },
+    {
+      "epoch": 1.955776587605203,
+      "grad_norm": 2.3205852966806804,
+      "learning_rate": 2.5635796133091084e-08,
+      "loss": 0.2712,
+      "step": 12781
+    },
+    {
+      "epoch": 1.95592960979342,
+      "grad_norm": 2.076512690272171,
+      "learning_rate": 2.54587679032281e-08,
+      "loss": 0.2737,
+      "step": 12782
+    },
+    {
+      "epoch": 1.9560826319816373,
+      "grad_norm": 1.9197513713440644,
+      "learning_rate": 2.528235224896203e-08,
+      "loss": 0.2378,
+      "step": 12783
+    },
+    {
+      "epoch": 1.9562356541698547,
+      "grad_norm": 2.137377531971493,
+      "learning_rate": 2.5106549181127536e-08,
+      "loss": 0.2342,
+      "step": 12784
+    },
+    {
+      "epoch": 1.9563886763580718,
+      "grad_norm": 2.4352962597543413,
+      "learning_rate": 2.4931358710522657e-08,
+      "loss": 0.2878,
+      "step": 12785
+    },
+    {
+      "epoch": 1.9565416985462893,
+      "grad_norm": 2.244106347035568,
+      "learning_rate": 2.4756780847905448e-08,
+      "loss": 0.2981,
+      "step": 12786
+    },
+    {
+      "epoch": 1.9566947207345065,
+      "grad_norm": 2.212890440083552,
+      "learning_rate": 2.4582815603998445e-08,
+      "loss": 0.2592,
+      "step": 12787
+    },
+    {
+      "epoch": 1.9568477429227238,
+      "grad_norm": 2.0677355320744435,
+      "learning_rate": 2.440946298948421e-08,
+      "loss": 0.2614,
+      "step": 12788
+    },
+    {
+      "epoch": 1.9570007651109411,
+      "grad_norm": 2.185133311922832,
+      "learning_rate": 2.4236723015008678e-08,
+      "loss": 0.2937,
+      "step": 12789
+    },
+    {
+      "epoch": 1.9571537872991582,
+      "grad_norm": 2.16851766484783,
+      "learning_rate": 2.4064595691182247e-08,
+      "loss": 0.2739,
+      "step": 12790
+    },
+    {
+      "epoch": 1.9573068094873758,
+      "grad_norm": 2.0968905865267,
+      "learning_rate": 2.3893081028575348e-08,
+      "loss": 0.2871,
+      "step": 12791
+    },
+    {
+      "epoch": 1.957459831675593,
+      "grad_norm": 2.1826172797738552,
+      "learning_rate": 2.3722179037720673e-08,
+      "loss": 0.287,
+      "step": 12792
+    },
+    {
+      "epoch": 1.9576128538638102,
+      "grad_norm": 2.0634184984657478,
+      "learning_rate": 2.3551889729114264e-08,
+      "loss": 0.2339,
+      "step": 12793
+    },
+    {
+      "epoch": 1.9577658760520276,
+      "grad_norm": 2.3855698367985756,
+      "learning_rate": 2.3382213113214426e-08,
+      "loss": 0.2683,
+      "step": 12794
+    },
+    {
+      "epoch": 1.9579188982402447,
+      "grad_norm": 2.2611003903495996,
+      "learning_rate": 2.321314920044171e-08,
+      "loss": 0.2936,
+      "step": 12795
+    },
+    {
+      "epoch": 1.9580719204284622,
+      "grad_norm": 2.1631865336468046,
+      "learning_rate": 2.3044698001178923e-08,
+      "loss": 0.295,
+      "step": 12796
+    },
+    {
+      "epoch": 1.9582249426166793,
+      "grad_norm": 2.012992323941021,
+      "learning_rate": 2.287685952577223e-08,
+      "loss": 0.3247,
+      "step": 12797
+    },
+    {
+      "epoch": 1.9583779648048967,
+      "grad_norm": 2.2715497490794054,
+      "learning_rate": 2.270963378452673e-08,
+      "loss": 0.2826,
+      "step": 12798
+    },
+    {
+      "epoch": 1.958530986993114,
+      "grad_norm": 2.3168577682957023,
+      "learning_rate": 2.2543020787715308e-08,
+      "loss": 0.3328,
+      "step": 12799
+    },
+    {
+      "epoch": 1.9586840091813313,
+      "grad_norm": 2.0418637103355786,
+      "learning_rate": 2.2377020545568673e-08,
+      "loss": 0.2509,
+      "step": 12800
+    },
+    {
+      "epoch": 1.9588370313695487,
+      "grad_norm": 2.1613693720849323,
+      "learning_rate": 2.221163306828089e-08,
+      "loss": 0.2569,
+      "step": 12801
+    },
+    {
+      "epoch": 1.9589900535577658,
+      "grad_norm": 1.972935734881763,
+      "learning_rate": 2.2046858366010503e-08,
+      "loss": 0.2469,
+      "step": 12802
+    },
+    {
+      "epoch": 1.9591430757459831,
+      "grad_norm": 1.8252344132384086,
+      "learning_rate": 2.1882696448877193e-08,
+      "loss": 0.2564,
+      "step": 12803
+    },
+    {
+      "epoch": 1.9592960979342005,
+      "grad_norm": 1.8533980617714614,
+      "learning_rate": 2.1719147326961786e-08,
+      "loss": 0.2677,
+      "step": 12804
+    },
+    {
+      "epoch": 1.9594491201224178,
+      "grad_norm": 1.9106183949752555,
+      "learning_rate": 2.155621101030847e-08,
+      "loss": 0.3246,
+      "step": 12805
+    },
+    {
+      "epoch": 1.9596021423106351,
+      "grad_norm": 2.4116686758652293,
+      "learning_rate": 2.1393887508924794e-08,
+      "loss": 0.2986,
+      "step": 12806
+    },
+    {
+      "epoch": 1.9597551644988522,
+      "grad_norm": 2.0370580101783755,
+      "learning_rate": 2.123217683277834e-08,
+      "loss": 0.2389,
+      "step": 12807
+    },
+    {
+      "epoch": 1.9599081866870698,
+      "grad_norm": 2.3374821126533636,
+      "learning_rate": 2.1071078991801164e-08,
+      "loss": 0.2636,
+      "step": 12808
+    },
+    {
+      "epoch": 1.960061208875287,
+      "grad_norm": 1.7950419140596388,
+      "learning_rate": 2.0910593995887573e-08,
+      "loss": 0.2105,
+      "step": 12809
+    },
+    {
+      "epoch": 1.9602142310635042,
+      "grad_norm": 2.073353178793678,
+      "learning_rate": 2.075072185489191e-08,
+      "loss": 0.2332,
+      "step": 12810
+    },
+    {
+      "epoch": 1.9603672532517216,
+      "grad_norm": 2.4863571935424273,
+      "learning_rate": 2.059146257863409e-08,
+      "loss": 0.308,
+      "step": 12811
+    },
+    {
+      "epoch": 1.9605202754399387,
+      "grad_norm": 2.2423145973838268,
+      "learning_rate": 2.0432816176894075e-08,
+      "loss": 0.2668,
+      "step": 12812
+    },
+    {
+      "epoch": 1.9606732976281562,
+      "grad_norm": 2.2613131124583283,
+      "learning_rate": 2.0274782659414072e-08,
+      "loss": 0.2899,
+      "step": 12813
+    },
+    {
+      "epoch": 1.9608263198163733,
+      "grad_norm": 2.2129547501462046,
+      "learning_rate": 2.0117362035901866e-08,
+      "loss": 0.2971,
+      "step": 12814
+    },
+    {
+      "epoch": 1.9609793420045907,
+      "grad_norm": 2.096062489754385,
+      "learning_rate": 1.9960554316024174e-08,
+      "loss": 0.2741,
+      "step": 12815
+    },
+    {
+      "epoch": 1.961132364192808,
+      "grad_norm": 1.853363747759997,
+      "learning_rate": 1.9804359509409954e-08,
+      "loss": 0.2461,
+      "step": 12816
+    },
+    {
+      "epoch": 1.9612853863810251,
+      "grad_norm": 2.1127207539609905,
+      "learning_rate": 1.9648777625653757e-08,
+      "loss": 0.2984,
+      "step": 12817
+    },
+    {
+      "epoch": 1.9614384085692427,
+      "grad_norm": 1.8244916883822468,
+      "learning_rate": 1.9493808674307945e-08,
+      "loss": 0.2208,
+      "step": 12818
+    },
+    {
+      "epoch": 1.9615914307574598,
+      "grad_norm": 1.9379785767665714,
+      "learning_rate": 1.9339452664892677e-08,
+      "loss": 0.2565,
+      "step": 12819
+    },
+    {
+      "epoch": 1.9617444529456771,
+      "grad_norm": 1.971258507742363,
+      "learning_rate": 1.918570960688593e-08,
+      "loss": 0.2639,
+      "step": 12820
+    },
+    {
+      "epoch": 1.9618974751338945,
+      "grad_norm": 2.206834865901176,
+      "learning_rate": 1.9032579509729033e-08,
+      "loss": 0.2336,
+      "step": 12821
+    },
+    {
+      "epoch": 1.9620504973221116,
+      "grad_norm": 2.2371951125913694,
+      "learning_rate": 1.88800623828278e-08,
+      "loss": 0.2632,
+      "step": 12822
+    },
+    {
+      "epoch": 1.9622035195103291,
+      "grad_norm": 2.214750729742328,
+      "learning_rate": 1.8728158235549187e-08,
+      "loss": 0.3168,
+      "step": 12823
+    },
+    {
+      "epoch": 1.9623565416985462,
+      "grad_norm": 1.8243923581833847,
+      "learning_rate": 1.8576867077220174e-08,
+      "loss": 0.2238,
+      "step": 12824
+    },
+    {
+      "epoch": 1.9625095638867636,
+      "grad_norm": 2.2820739863423527,
+      "learning_rate": 1.8426188917133324e-08,
+      "loss": 0.2798,
+      "step": 12825
+    },
+    {
+      "epoch": 1.962662586074981,
+      "grad_norm": 2.3473441828274417,
+      "learning_rate": 1.8276123764543463e-08,
+      "loss": 0.3383,
+      "step": 12826
+    },
+    {
+      "epoch": 1.962815608263198,
+      "grad_norm": 1.810468033827546,
+      "learning_rate": 1.8126671628664328e-08,
+      "loss": 0.2646,
+      "step": 12827
+    },
+    {
+      "epoch": 1.9629686304514156,
+      "grad_norm": 2.0161550530630774,
+      "learning_rate": 1.7977832518676354e-08,
+      "loss": 0.2554,
+      "step": 12828
+    },
+    {
+      "epoch": 1.9631216526396327,
+      "grad_norm": 2.0057193127755526,
+      "learning_rate": 1.7829606443718893e-08,
+      "loss": 0.2195,
+      "step": 12829
+    },
+    {
+      "epoch": 1.96327467482785,
+      "grad_norm": 1.9895992190324854,
+      "learning_rate": 1.7681993412895782e-08,
+      "loss": 0.1972,
+      "step": 12830
+    },
+    {
+      "epoch": 1.9634276970160673,
+      "grad_norm": 2.424438500859323,
+      "learning_rate": 1.7534993435273095e-08,
+      "loss": 0.284,
+      "step": 12831
+    },
+    {
+      "epoch": 1.9635807192042847,
+      "grad_norm": 2.342021551362485,
+      "learning_rate": 1.738860651987806e-08,
+      "loss": 0.2584,
+      "step": 12832
+    },
+    {
+      "epoch": 1.963733741392502,
+      "grad_norm": 1.952229890512349,
+      "learning_rate": 1.724283267570126e-08,
+      "loss": 0.3085,
+      "step": 12833
+    },
+    {
+      "epoch": 1.9638867635807191,
+      "grad_norm": 2.042895088129427,
+      "learning_rate": 1.7097671911693315e-08,
+      "loss": 0.236,
+      "step": 12834
+    },
+    {
+      "epoch": 1.9640397857689365,
+      "grad_norm": 2.325198304496035,
+      "learning_rate": 1.6953124236772645e-08,
+      "loss": 0.2503,
+      "step": 12835
+    },
+    {
+      "epoch": 1.9641928079571538,
+      "grad_norm": 2.1434028335453803,
+      "learning_rate": 1.6809189659813264e-08,
+      "loss": 0.2563,
+      "step": 12836
+    },
+    {
+      "epoch": 1.9643458301453711,
+      "grad_norm": 2.7341650968690705,
+      "learning_rate": 1.6665868189655878e-08,
+      "loss": 0.2641,
+      "step": 12837
+    },
+    {
+      "epoch": 1.9644988523335885,
+      "grad_norm": 2.1110716721025806,
+      "learning_rate": 1.652315983510344e-08,
+      "loss": 0.2664,
+      "step": 12838
+    },
+    {
+      "epoch": 1.9646518745218056,
+      "grad_norm": 2.4763953852189524,
+      "learning_rate": 1.638106460491895e-08,
+      "loss": 0.3259,
+      "step": 12839
+    },
+    {
+      "epoch": 1.9648048967100231,
+      "grad_norm": 1.9515477357591082,
+      "learning_rate": 1.6239582507828754e-08,
+      "loss": 0.2274,
+      "step": 12840
+    },
+    {
+      "epoch": 1.9649579188982402,
+      "grad_norm": 2.1874600207771144,
+      "learning_rate": 1.6098713552523683e-08,
+      "loss": 0.3048,
+      "step": 12841
+    },
+    {
+      "epoch": 1.9651109410864576,
+      "grad_norm": 2.163087834030164,
+      "learning_rate": 1.5958457747652368e-08,
+      "loss": 0.2767,
+      "step": 12842
+    },
+    {
+      "epoch": 1.965263963274675,
+      "grad_norm": 2.405722917761059,
+      "learning_rate": 1.5818815101831252e-08,
+      "loss": 0.2629,
+      "step": 12843
+    },
+    {
+      "epoch": 1.965416985462892,
+      "grad_norm": 2.3389774387611086,
+      "learning_rate": 1.567978562363459e-08,
+      "loss": 0.2692,
+      "step": 12844
+    },
+    {
+      "epoch": 1.9655700076511096,
+      "grad_norm": 1.774637055198869,
+      "learning_rate": 1.5541369321601107e-08,
+      "loss": 0.2302,
+      "step": 12845
+    },
+    {
+      "epoch": 1.9657230298393267,
+      "grad_norm": 2.2203580711233815,
+      "learning_rate": 1.5403566204231777e-08,
+      "loss": 0.2742,
+      "step": 12846
+    },
+    {
+      "epoch": 1.965876052027544,
+      "grad_norm": 2.0022244855679183,
+      "learning_rate": 1.5266376279988726e-08,
+      "loss": 0.2747,
+      "step": 12847
+    },
+    {
+      "epoch": 1.9660290742157613,
+      "grad_norm": 2.086556860314529,
+      "learning_rate": 1.512979955729854e-08,
+      "loss": 0.2604,
+      "step": 12848
+    },
+    {
+      "epoch": 1.9661820964039785,
+      "grad_norm": 2.044756387848911,
+      "learning_rate": 1.499383604454896e-08,
+      "loss": 0.2365,
+      "step": 12849
+    },
+    {
+      "epoch": 1.966335118592196,
+      "grad_norm": 2.2031325687200654,
+      "learning_rate": 1.4858485750088857e-08,
+      "loss": 0.2612,
+      "step": 12850
+    },
+    {
+      "epoch": 1.9664881407804131,
+      "grad_norm": 2.0442598118479873,
+      "learning_rate": 1.4723748682231587e-08,
+      "loss": 0.2223,
+      "step": 12851
+    },
+    {
+      "epoch": 1.9666411629686305,
+      "grad_norm": 2.024031388837172,
+      "learning_rate": 1.4589624849250527e-08,
+      "loss": 0.2676,
+      "step": 12852
+    },
+    {
+      "epoch": 1.9667941851568478,
+      "grad_norm": 2.042948797618122,
+      "learning_rate": 1.4456114259384647e-08,
+      "loss": 0.2578,
+      "step": 12853
+    },
+    {
+      "epoch": 1.966947207345065,
+      "grad_norm": 1.9906287211778895,
+      "learning_rate": 1.4323216920831829e-08,
+      "loss": 0.2132,
+      "step": 12854
+    },
+    {
+      "epoch": 1.9671002295332825,
+      "grad_norm": 2.482263149556431,
+      "learning_rate": 1.4190932841755544e-08,
+      "loss": 0.333,
+      "step": 12855
+    },
+    {
+      "epoch": 1.9672532517214996,
+      "grad_norm": 1.7243344603142814,
+      "learning_rate": 1.4059262030278186e-08,
+      "loss": 0.2489,
+      "step": 12856
+    },
+    {
+      "epoch": 1.967406273909717,
+      "grad_norm": 2.451829415489958,
+      "learning_rate": 1.392820449448662e-08,
+      "loss": 0.2961,
+      "step": 12857
+    },
+    {
+      "epoch": 1.9675592960979342,
+      "grad_norm": 2.3167822602608132,
+      "learning_rate": 1.3797760242429958e-08,
+      "loss": 0.2894,
+      "step": 12858
+    },
+    {
+      "epoch": 1.9677123182861513,
+      "grad_norm": 2.163481464374622,
+      "learning_rate": 1.3667929282118464e-08,
+      "loss": 0.2569,
+      "step": 12859
+    },
+    {
+      "epoch": 1.967865340474369,
+      "grad_norm": 1.8496510780178912,
+      "learning_rate": 1.3538711621526868e-08,
+      "loss": 0.2437,
+      "step": 12860
+    },
+    {
+      "epoch": 1.968018362662586,
+      "grad_norm": 2.229870372063878,
+      "learning_rate": 1.3410107268589934e-08,
+      "loss": 0.3158,
+      "step": 12861
+    },
+    {
+      "epoch": 1.9681713848508033,
+      "grad_norm": 2.30265422085295,
+      "learning_rate": 1.3282116231205789e-08,
+      "loss": 0.2171,
+      "step": 12862
+    },
+    {
+      "epoch": 1.9683244070390207,
+      "grad_norm": 2.053435870710879,
+      "learning_rate": 1.3154738517235921e-08,
+      "loss": 0.2806,
+      "step": 12863
+    },
+    {
+      "epoch": 1.9684774292272378,
+      "grad_norm": 2.1300685488173254,
+      "learning_rate": 1.3027974134501853e-08,
+      "loss": 0.3023,
+      "step": 12864
+    },
+    {
+      "epoch": 1.9686304514154553,
+      "grad_norm": 2.094110877470264,
+      "learning_rate": 1.2901823090789577e-08,
+      "loss": 0.2508,
+      "step": 12865
+    },
+    {
+      "epoch": 1.9687834736036725,
+      "grad_norm": 2.003064762109647,
+      "learning_rate": 1.2776285393845122e-08,
+      "loss": 0.258,
+      "step": 12866
+    },
+    {
+      "epoch": 1.9689364957918898,
+      "grad_norm": 2.4322997253204144,
+      "learning_rate": 1.2651361051380095e-08,
+      "loss": 0.3001,
+      "step": 12867
+    },
+    {
+      "epoch": 1.9690895179801071,
+      "grad_norm": 2.0065478541875463,
+      "learning_rate": 1.2527050071065028e-08,
+      "loss": 0.3071,
+      "step": 12868
+    },
+    {
+      "epoch": 1.9692425401683245,
+      "grad_norm": 1.9145659667385493,
+      "learning_rate": 1.2403352460536034e-08,
+      "loss": 0.267,
+      "step": 12869
+    },
+    {
+      "epoch": 1.9693955623565418,
+      "grad_norm": 1.998726280086352,
+      "learning_rate": 1.2280268227388148e-08,
+      "loss": 0.2625,
+      "step": 12870
+    },
+    {
+      "epoch": 1.969548584544759,
+      "grad_norm": 1.7810985547272813,
+      "learning_rate": 1.215779737918088e-08,
+      "loss": 0.2553,
+      "step": 12871
+    },
+    {
+      "epoch": 1.9697016067329762,
+      "grad_norm": 2.0506737515397133,
+      "learning_rate": 1.20359399234371e-08,
+      "loss": 0.2376,
+      "step": 12872
+    },
+    {
+      "epoch": 1.9698546289211936,
+      "grad_norm": 2.4243875323116053,
+      "learning_rate": 1.1914695867638603e-08,
+      "loss": 0.3341,
+      "step": 12873
+    },
+    {
+      "epoch": 1.970007651109411,
+      "grad_norm": 1.930622018969895,
+      "learning_rate": 1.1794065219231654e-08,
+      "loss": 0.2257,
+      "step": 12874
+    },
+    {
+      "epoch": 1.9701606732976282,
+      "grad_norm": 1.864437419947269,
+      "learning_rate": 1.167404798562588e-08,
+      "loss": 0.2139,
+      "step": 12875
+    },
+    {
+      "epoch": 1.9703136954858453,
+      "grad_norm": 1.897571077812364,
+      "learning_rate": 1.1554644174192053e-08,
+      "loss": 0.2248,
+      "step": 12876
+    },
+    {
+      "epoch": 1.970466717674063,
+      "grad_norm": 1.9202972446689073,
+      "learning_rate": 1.1435853792260976e-08,
+      "loss": 0.2011,
+      "step": 12877
+    },
+    {
+      "epoch": 1.97061973986228,
+      "grad_norm": 2.110629293594774,
+      "learning_rate": 1.1317676847131254e-08,
+      "loss": 0.3001,
+      "step": 12878
+    },
+    {
+      "epoch": 1.9707727620504973,
+      "grad_norm": 2.392523432977483,
+      "learning_rate": 1.1200113346058195e-08,
+      "loss": 0.3097,
+      "step": 12879
+    },
+    {
+      "epoch": 1.9709257842387147,
+      "grad_norm": 2.2066538540809435,
+      "learning_rate": 1.1083163296262689e-08,
+      "loss": 0.2743,
+      "step": 12880
+    },
+    {
+      "epoch": 1.9710788064269318,
+      "grad_norm": 1.9974484118386584,
+      "learning_rate": 1.096682670492677e-08,
+      "loss": 0.2149,
+      "step": 12881
+    },
+    {
+      "epoch": 1.9712318286151493,
+      "grad_norm": 2.0342333461670425,
+      "learning_rate": 1.0851103579194722e-08,
+      "loss": 0.2746,
+      "step": 12882
+    },
+    {
+      "epoch": 1.9713848508033665,
+      "grad_norm": 2.3163031751434295,
+      "learning_rate": 1.0735993926175304e-08,
+      "loss": 0.2656,
+      "step": 12883
+    },
+    {
+      "epoch": 1.9715378729915838,
+      "grad_norm": 1.9829201798149283,
+      "learning_rate": 1.0621497752936194e-08,
+      "loss": 0.2601,
+      "step": 12884
+    },
+    {
+      "epoch": 1.9716908951798011,
+      "grad_norm": 2.1571377906044735,
+      "learning_rate": 1.0507615066509547e-08,
+      "loss": 0.2833,
+      "step": 12885
+    },
+    {
+      "epoch": 1.9718439173680182,
+      "grad_norm": 2.07040945084194,
+      "learning_rate": 1.0394345873889766e-08,
+      "loss": 0.2879,
+      "step": 12886
+    },
+    {
+      "epoch": 1.9719969395562358,
+      "grad_norm": 2.2778466632041887,
+      "learning_rate": 1.0281690182032399e-08,
+      "loss": 0.2694,
+      "step": 12887
+    },
+    {
+      "epoch": 1.972149961744453,
+      "grad_norm": 1.9727267744304013,
+      "learning_rate": 1.0169647997856358e-08,
+      "loss": 0.2375,
+      "step": 12888
+    },
+    {
+      "epoch": 1.9723029839326702,
+      "grad_norm": 2.0960160436743003,
+      "learning_rate": 1.0058219328242802e-08,
+      "loss": 0.3329,
+      "step": 12889
+    },
+    {
+      "epoch": 1.9724560061208876,
+      "grad_norm": 2.046906516879214,
+      "learning_rate": 9.947404180035147e-09,
+      "loss": 0.2837,
+      "step": 12890
+    },
+    {
+      "epoch": 1.9726090283091047,
+      "grad_norm": 2.3133831775980847,
+      "learning_rate": 9.83720256003795e-09,
+      "loss": 0.3173,
+      "step": 12891
+    },
+    {
+      "epoch": 1.9727620504973222,
+      "grad_norm": 2.025923377837509,
+      "learning_rate": 9.727614475020241e-09,
+      "loss": 0.2715,
+      "step": 12892
+    },
+    {
+      "epoch": 1.9729150726855393,
+      "grad_norm": 2.1217931982552334,
+      "learning_rate": 9.618639931712193e-09,
+      "loss": 0.2897,
+      "step": 12893
+    },
+    {
+      "epoch": 1.9730680948737567,
+      "grad_norm": 2.3348882198976333,
+      "learning_rate": 9.510278936806227e-09,
+      "loss": 0.3275,
+      "step": 12894
+    },
+    {
+      "epoch": 1.973221117061974,
+      "grad_norm": 2.5591447995277403,
+      "learning_rate": 9.402531496957024e-09,
+      "loss": 0.3593,
+      "step": 12895
+    },
+    {
+      "epoch": 1.9733741392501911,
+      "grad_norm": 2.0794411023520243,
+      "learning_rate": 9.29539761878151e-09,
+      "loss": 0.2514,
+      "step": 12896
+    },
+    {
+      "epoch": 1.9735271614384087,
+      "grad_norm": 2.097465611753309,
+      "learning_rate": 9.188877308858867e-09,
+      "loss": 0.2895,
+      "step": 12897
+    },
+    {
+      "epoch": 1.9736801836266258,
+      "grad_norm": 2.0471184159575007,
+      "learning_rate": 9.082970573732752e-09,
+      "loss": 0.2243,
+      "step": 12898
+    },
+    {
+      "epoch": 1.9738332058148431,
+      "grad_norm": 2.1026098669748086,
+      "learning_rate": 8.97767741990574e-09,
+      "loss": 0.2623,
+      "step": 12899
+    },
+    {
+      "epoch": 1.9739862280030605,
+      "grad_norm": 2.007922658885538,
+      "learning_rate": 8.872997853843767e-09,
+      "loss": 0.2379,
+      "step": 12900
+    },
+    {
+      "epoch": 1.9741392501912778,
+      "grad_norm": 2.1522852184706345,
+      "learning_rate": 8.76893188197725e-09,
+      "loss": 0.2646,
+      "step": 12901
+    },
+    {
+      "epoch": 1.9742922723794951,
+      "grad_norm": 1.7457113137200364,
+      "learning_rate": 8.665479510696629e-09,
+      "loss": 0.2275,
+      "step": 12902
+    },
+    {
+      "epoch": 1.9744452945677122,
+      "grad_norm": 2.3852903678590494,
+      "learning_rate": 8.562640746354601e-09,
+      "loss": 0.2778,
+      "step": 12903
+    },
+    {
+      "epoch": 1.9745983167559296,
+      "grad_norm": 2.0897669244676216,
+      "learning_rate": 8.460415595268334e-09,
+      "loss": 0.2541,
+      "step": 12904
+    },
+    {
+      "epoch": 1.974751338944147,
+      "grad_norm": 2.21857340447905,
+      "learning_rate": 8.358804063715032e-09,
+      "loss": 0.2525,
+      "step": 12905
+    },
+    {
+      "epoch": 1.9749043611323642,
+      "grad_norm": 1.9060225156510942,
+      "learning_rate": 8.257806157934145e-09,
+      "loss": 0.276,
+      "step": 12906
+    },
+    {
+      "epoch": 1.9750573833205816,
+      "grad_norm": 2.062959820646989,
+      "learning_rate": 8.157421884129602e-09,
+      "loss": 0.2851,
+      "step": 12907
+    },
+    {
+      "epoch": 1.9752104055087987,
+      "grad_norm": 1.7379375669518808,
+      "learning_rate": 8.057651248466469e-09,
+      "loss": 0.2096,
+      "step": 12908
+    },
+    {
+      "epoch": 1.9753634276970162,
+      "grad_norm": 1.8775977832890811,
+      "learning_rate": 7.958494257072069e-09,
+      "loss": 0.2281,
+      "step": 12909
+    },
+    {
+      "epoch": 1.9755164498852333,
+      "grad_norm": 1.8697635311100709,
+      "learning_rate": 7.859950916034865e-09,
+      "loss": 0.2855,
+      "step": 12910
+    },
+    {
+      "epoch": 1.9756694720734507,
+      "grad_norm": 2.3981445826317906,
+      "learning_rate": 7.76202123140779e-09,
+      "loss": 0.3044,
+      "step": 12911
+    },
+    {
+      "epoch": 1.975822494261668,
+      "grad_norm": 2.168622460349219,
+      "learning_rate": 7.664705209204925e-09,
+      "loss": 0.2939,
+      "step": 12912
+    },
+    {
+      "epoch": 1.9759755164498851,
+      "grad_norm": 2.210540309755076,
+      "learning_rate": 7.568002855402602e-09,
+      "loss": 0.3456,
+      "step": 12913
+    },
+    {
+      "epoch": 1.9761285386381027,
+      "grad_norm": 2.47623544445302,
+      "learning_rate": 7.471914175940509e-09,
+      "loss": 0.3503,
+      "step": 12914
+    },
+    {
+      "epoch": 1.9762815608263198,
+      "grad_norm": 2.0110291162646146,
+      "learning_rate": 7.376439176718375e-09,
+      "loss": 0.2296,
+      "step": 12915
+    },
+    {
+      "epoch": 1.9764345830145371,
+      "grad_norm": 2.059626415844223,
+      "learning_rate": 7.2815778636003975e-09,
+      "loss": 0.261,
+      "step": 12916
+    },
+    {
+      "epoch": 1.9765876052027544,
+      "grad_norm": 1.9309744737462933,
+      "learning_rate": 7.187330242413026e-09,
+      "loss": 0.2504,
+      "step": 12917
+    },
+    {
+      "epoch": 1.9767406273909716,
+      "grad_norm": 2.153764527156966,
+      "learning_rate": 7.093696318943854e-09,
+      "loss": 0.2581,
+      "step": 12918
+    },
+    {
+      "epoch": 1.9768936495791891,
+      "grad_norm": 2.184851336719344,
+      "learning_rate": 7.000676098942727e-09,
+      "loss": 0.3208,
+      "step": 12919
+    },
+    {
+      "epoch": 1.9770466717674062,
+      "grad_norm": 2.1162936254329954,
+      "learning_rate": 6.9082695881228515e-09,
+      "loss": 0.2874,
+      "step": 12920
+    },
+    {
+      "epoch": 1.9771996939556236,
+      "grad_norm": 1.8457304539215134,
+      "learning_rate": 6.816476792159687e-09,
+      "loss": 0.2402,
+      "step": 12921
+    },
+    {
+      "epoch": 1.977352716143841,
+      "grad_norm": 2.2107074116766294,
+      "learning_rate": 6.725297716689838e-09,
+      "loss": 0.3502,
+      "step": 12922
+    },
+    {
+      "epoch": 1.977505738332058,
+      "grad_norm": 2.312117402985523,
+      "learning_rate": 6.6347323673143786e-09,
+      "loss": 0.3004,
+      "step": 12923
+    },
+    {
+      "epoch": 1.9776587605202756,
+      "grad_norm": 1.916658673354669,
+      "learning_rate": 6.544780749593305e-09,
+      "loss": 0.2303,
+      "step": 12924
+    },
+    {
+      "epoch": 1.9778117827084927,
+      "grad_norm": 1.8504219486404065,
+      "learning_rate": 6.455442869052197e-09,
+      "loss": 0.2086,
+      "step": 12925
+    },
+    {
+      "epoch": 1.97796480489671,
+      "grad_norm": 2.252796300111889,
+      "learning_rate": 6.366718731177779e-09,
+      "loss": 0.3257,
+      "step": 12926
+    },
+    {
+      "epoch": 1.9781178270849273,
+      "grad_norm": 2.0448647401610414,
+      "learning_rate": 6.278608341416803e-09,
+      "loss": 0.2517,
+      "step": 12927
+    },
+    {
+      "epoch": 1.9782708492731444,
+      "grad_norm": 1.9816261463855467,
+      "learning_rate": 6.1911117051838274e-09,
+      "loss": 0.2709,
+      "step": 12928
+    },
+    {
+      "epoch": 1.978423871461362,
+      "grad_norm": 2.1873499122980555,
+      "learning_rate": 6.104228827850111e-09,
+      "loss": 0.287,
+      "step": 12929
+    },
+    {
+      "epoch": 1.9785768936495791,
+      "grad_norm": 2.1254405287529763,
+      "learning_rate": 6.0179597147524975e-09,
+      "loss": 0.2381,
+      "step": 12930
+    },
+    {
+      "epoch": 1.9787299158377964,
+      "grad_norm": 2.0765132192816216,
+      "learning_rate": 5.932304371187858e-09,
+      "loss": 0.2851,
+      "step": 12931
+    },
+    {
+      "epoch": 1.9788829380260138,
+      "grad_norm": 2.2452308925752194,
+      "learning_rate": 5.847262802417542e-09,
+      "loss": 0.2512,
+      "step": 12932
+    },
+    {
+      "epoch": 1.9790359602142311,
+      "grad_norm": 2.1688896379609885,
+      "learning_rate": 5.7628350136640365e-09,
+      "loss": 0.3165,
+      "step": 12933
+    },
+    {
+      "epoch": 1.9791889824024484,
+      "grad_norm": 2.1543318719909053,
+      "learning_rate": 5.6790210101131945e-09,
+      "loss": 0.2452,
+      "step": 12934
+    },
+    {
+      "epoch": 1.9793420045906656,
+      "grad_norm": 1.9687252404414846,
+      "learning_rate": 5.595820796912011e-09,
+      "loss": 0.2407,
+      "step": 12935
+    },
+    {
+      "epoch": 1.979495026778883,
+      "grad_norm": 2.2367361565308825,
+      "learning_rate": 5.513234379168619e-09,
+      "loss": 0.2738,
+      "step": 12936
+    },
+    {
+      "epoch": 1.9796480489671002,
+      "grad_norm": 2.0907816123855185,
+      "learning_rate": 5.431261761956741e-09,
+      "loss": 0.2742,
+      "step": 12937
+    },
+    {
+      "epoch": 1.9798010711553176,
+      "grad_norm": 1.9706136989908423,
+      "learning_rate": 5.349902950310126e-09,
+      "loss": 0.2708,
+      "step": 12938
+    },
+    {
+      "epoch": 1.979954093343535,
+      "grad_norm": 1.9392410273061564,
+      "learning_rate": 5.269157949224779e-09,
+      "loss": 0.2548,
+      "step": 12939
+    },
+    {
+      "epoch": 1.980107115531752,
+      "grad_norm": 2.3816947627575873,
+      "learning_rate": 5.189026763661176e-09,
+      "loss": 0.2669,
+      "step": 12940
+    },
+    {
+      "epoch": 1.9802601377199696,
+      "grad_norm": 2.3186158240362045,
+      "learning_rate": 5.109509398538714e-09,
+      "loss": 0.3214,
+      "step": 12941
+    },
+    {
+      "epoch": 1.9804131599081867,
+      "grad_norm": 1.9554878445595376,
+      "learning_rate": 5.030605858740156e-09,
+      "loss": 0.2202,
+      "step": 12942
+    },
+    {
+      "epoch": 1.980566182096404,
+      "grad_norm": 2.0322729150228134,
+      "learning_rate": 4.952316149114955e-09,
+      "loss": 0.2662,
+      "step": 12943
+    },
+    {
+      "epoch": 1.9807192042846213,
+      "grad_norm": 1.8278177150486519,
+      "learning_rate": 4.874640274467046e-09,
+      "loss": 0.2244,
+      "step": 12944
+    },
+    {
+      "epoch": 1.9808722264728384,
+      "grad_norm": 2.174806576540863,
+      "learning_rate": 4.797578239569278e-09,
+      "loss": 0.2741,
+      "step": 12945
+    },
+    {
+      "epoch": 1.981025248661056,
+      "grad_norm": 2.2415653534967217,
+      "learning_rate": 4.721130049154532e-09,
+      "loss": 0.2785,
+      "step": 12946
+    },
+    {
+      "epoch": 1.9811782708492731,
+      "grad_norm": 2.166734196094784,
+      "learning_rate": 4.64529570791572e-09,
+      "loss": 0.2677,
+      "step": 12947
+    },
+    {
+      "epoch": 1.9813312930374904,
+      "grad_norm": 2.0414327018784006,
+      "learning_rate": 4.5700752205113385e-09,
+      "loss": 0.2732,
+      "step": 12948
+    },
+    {
+      "epoch": 1.9814843152257078,
+      "grad_norm": 2.256044758470271,
+      "learning_rate": 4.495468591562135e-09,
+      "loss": 0.3323,
+      "step": 12949
+    },
+    {
+      "epoch": 1.981637337413925,
+      "grad_norm": 2.1976314322826114,
+      "learning_rate": 4.421475825647781e-09,
+      "loss": 0.2598,
+      "step": 12950
+    },
+    {
+      "epoch": 1.9817903596021424,
+      "grad_norm": 1.7322213738511125,
+      "learning_rate": 4.3480969273135276e-09,
+      "loss": 0.2153,
+      "step": 12951
+    },
+    {
+      "epoch": 1.9819433817903596,
+      "grad_norm": 1.9892742007837072,
+      "learning_rate": 4.275331901066881e-09,
+      "loss": 0.2685,
+      "step": 12952
+    },
+    {
+      "epoch": 1.982096403978577,
+      "grad_norm": 2.201534831571064,
+      "learning_rate": 4.20318075137427e-09,
+      "loss": 0.2703,
+      "step": 12953
+    },
+    {
+      "epoch": 1.9822494261667942,
+      "grad_norm": 2.3807635297799283,
+      "learning_rate": 4.1316434826688124e-09,
+      "loss": 0.243,
+      "step": 12954
+    },
+    {
+      "epoch": 1.9824024483550113,
+      "grad_norm": 2.2659288950701346,
+      "learning_rate": 4.060720099343662e-09,
+      "loss": 0.3149,
+      "step": 12955
+    },
+    {
+      "epoch": 1.982555470543229,
+      "grad_norm": 2.1254319890729083,
+      "learning_rate": 3.990410605753115e-09,
+      "loss": 0.2547,
+      "step": 12956
+    },
+    {
+      "epoch": 1.982708492731446,
+      "grad_norm": 1.7746557192110035,
+      "learning_rate": 3.920715006217047e-09,
+      "loss": 0.2284,
+      "step": 12957
+    },
+    {
+      "epoch": 1.9828615149196633,
+      "grad_norm": 2.070971432899368,
+      "learning_rate": 3.851633305014257e-09,
+      "loss": 0.2593,
+      "step": 12958
+    },
+    {
+      "epoch": 1.9830145371078807,
+      "grad_norm": 2.2118533165855534,
+      "learning_rate": 3.78316550638691e-09,
+      "loss": 0.3016,
+      "step": 12959
+    },
+    {
+      "epoch": 1.9831675592960978,
+      "grad_norm": 1.9279899584450073,
+      "learning_rate": 3.715311614541639e-09,
+      "loss": 0.2188,
+      "step": 12960
+    },
+    {
+      "epoch": 1.9833205814843153,
+      "grad_norm": 2.071591635449214,
+      "learning_rate": 3.648071633645112e-09,
+      "loss": 0.2602,
+      "step": 12961
+    },
+    {
+      "epoch": 1.9834736036725324,
+      "grad_norm": 2.147092374012748,
+      "learning_rate": 3.5814455678262473e-09,
+      "loss": 0.3022,
+      "step": 12962
+    },
+    {
+      "epoch": 1.9836266258607498,
+      "grad_norm": 2.022467934629971,
+      "learning_rate": 3.5154334211762174e-09,
+      "loss": 0.2448,
+      "step": 12963
+    },
+    {
+      "epoch": 1.9837796480489671,
+      "grad_norm": 2.130643029150222,
+      "learning_rate": 3.4500351977506675e-09,
+      "loss": 0.2165,
+      "step": 12964
+    },
+    {
+      "epoch": 1.9839326702371842,
+      "grad_norm": 2.0960930753843914,
+      "learning_rate": 3.3852509015652734e-09,
+      "loss": 0.2506,
+      "step": 12965
+    },
+    {
+      "epoch": 1.9840856924254018,
+      "grad_norm": 1.9855073543678754,
+      "learning_rate": 3.3210805365979648e-09,
+      "loss": 0.2233,
+      "step": 12966
+    },
+    {
+      "epoch": 1.984238714613619,
+      "grad_norm": 2.051055830940307,
+      "learning_rate": 3.2575241067911435e-09,
+      "loss": 0.2452,
+      "step": 12967
+    },
+    {
+      "epoch": 1.9843917368018362,
+      "grad_norm": 2.0931128703850708,
+      "learning_rate": 3.194581616046133e-09,
+      "loss": 0.2692,
+      "step": 12968
+    },
+    {
+      "epoch": 1.9845447589900536,
+      "grad_norm": 1.736009426122614,
+      "learning_rate": 3.1322530682309506e-09,
+      "loss": 0.1924,
+      "step": 12969
+    },
+    {
+      "epoch": 1.984697781178271,
+      "grad_norm": 2.124895379549314,
+      "learning_rate": 3.0705384671714245e-09,
+      "loss": 0.2864,
+      "step": 12970
+    },
+    {
+      "epoch": 1.9848508033664882,
+      "grad_norm": 2.0882242357227283,
+      "learning_rate": 3.0094378166578563e-09,
+      "loss": 0.2996,
+      "step": 12971
+    },
+    {
+      "epoch": 1.9850038255547053,
+      "grad_norm": 1.9487330157379934,
+      "learning_rate": 2.9489511204439105e-09,
+      "loss": 0.2617,
+      "step": 12972
+    },
+    {
+      "epoch": 1.9851568477429227,
+      "grad_norm": 2.2667445577756444,
+      "learning_rate": 2.8890783822432823e-09,
+      "loss": 0.2701,
+      "step": 12973
+    },
+    {
+      "epoch": 1.98530986993114,
+      "grad_norm": 2.3116383396322986,
+      "learning_rate": 2.8298196057330308e-09,
+      "loss": 0.2981,
+      "step": 12974
+    },
+    {
+      "epoch": 1.9854628921193573,
+      "grad_norm": 1.7453561995500706,
+      "learning_rate": 2.7711747945524668e-09,
+      "loss": 0.1991,
+      "step": 12975
+    },
+    {
+      "epoch": 1.9856159143075747,
+      "grad_norm": 1.9390207186144572,
+      "learning_rate": 2.7131439523042646e-09,
+      "loss": 0.2393,
+      "step": 12976
+    },
+    {
+      "epoch": 1.9857689364957918,
+      "grad_norm": 2.3644792875455125,
+      "learning_rate": 2.6557270825511293e-09,
+      "loss": 0.2784,
+      "step": 12977
+    },
+    {
+      "epoch": 1.9859219586840093,
+      "grad_norm": 2.178932135966078,
+      "learning_rate": 2.5989241888191296e-09,
+      "loss": 0.343,
+      "step": 12978
+    },
+    {
+      "epoch": 1.9860749808722264,
+      "grad_norm": 2.1307936669410132,
+      "learning_rate": 2.542735274597696e-09,
+      "loss": 0.2718,
+      "step": 12979
+    },
+    {
+      "epoch": 1.9862280030604438,
+      "grad_norm": 1.9874821763775943,
+      "learning_rate": 2.4871603433374026e-09,
+      "loss": 0.2545,
+      "step": 12980
+    },
+    {
+      "epoch": 1.9863810252486611,
+      "grad_norm": 1.790519718630377,
+      "learning_rate": 2.432199398451074e-09,
+      "loss": 0.188,
+      "step": 12981
+    },
+    {
+      "epoch": 1.9865340474368782,
+      "grad_norm": 2.3174362383429723,
+      "learning_rate": 2.3778524433137882e-09,
+      "loss": 0.2746,
+      "step": 12982
+    },
+    {
+      "epoch": 1.9866870696250958,
+      "grad_norm": 2.346253570180665,
+      "learning_rate": 2.3241194812639868e-09,
+      "loss": 0.3198,
+      "step": 12983
+    },
+    {
+      "epoch": 1.986840091813313,
+      "grad_norm": 2.1172501130386716,
+      "learning_rate": 2.2710005156001415e-09,
+      "loss": 0.2755,
+      "step": 12984
+    },
+    {
+      "epoch": 1.9869931140015302,
+      "grad_norm": 2.2663121503499677,
+      "learning_rate": 2.218495549586308e-09,
+      "loss": 0.2343,
+      "step": 12985
+    },
+    {
+      "epoch": 1.9871461361897476,
+      "grad_norm": 1.9793839185682316,
+      "learning_rate": 2.166604586445464e-09,
+      "loss": 0.2497,
+      "step": 12986
+    },
+    {
+      "epoch": 1.9872991583779647,
+      "grad_norm": 2.2530991162584217,
+      "learning_rate": 2.1153276293661708e-09,
+      "loss": 0.2837,
+      "step": 12987
+    },
+    {
+      "epoch": 1.9874521805661822,
+      "grad_norm": 2.3052278983511703,
+      "learning_rate": 2.0646646814959093e-09,
+      "loss": 0.3059,
+      "step": 12988
+    },
+    {
+      "epoch": 1.9876052027543993,
+      "grad_norm": 2.0974839786486728,
+      "learning_rate": 2.014615745946635e-09,
+      "loss": 0.3217,
+      "step": 12989
+    },
+    {
+      "epoch": 1.9877582249426167,
+      "grad_norm": 2.233579314139941,
+      "learning_rate": 1.9651808257925564e-09,
+      "loss": 0.2597,
+      "step": 12990
+    },
+    {
+      "epoch": 1.987911247130834,
+      "grad_norm": 1.969583542552495,
+      "learning_rate": 1.9163599240690225e-09,
+      "loss": 0.2457,
+      "step": 12991
+    },
+    {
+      "epoch": 1.9880642693190511,
+      "grad_norm": 2.2150331863004236,
+      "learning_rate": 1.868153043774745e-09,
+      "loss": 0.2915,
+      "step": 12992
+    },
+    {
+      "epoch": 1.9882172915072687,
+      "grad_norm": 2.0676187007120923,
+      "learning_rate": 1.82056018786958e-09,
+      "loss": 0.2538,
+      "step": 12993
+    },
+    {
+      "epoch": 1.9883703136954858,
+      "grad_norm": 1.9421580439547903,
+      "learning_rate": 1.773581359277854e-09,
+      "loss": 0.2766,
+      "step": 12994
+    },
+    {
+      "epoch": 1.9885233358837031,
+      "grad_norm": 1.913701865466778,
+      "learning_rate": 1.727216560882816e-09,
+      "loss": 0.2257,
+      "step": 12995
+    },
+    {
+      "epoch": 1.9886763580719204,
+      "grad_norm": 2.220818373435268,
+      "learning_rate": 1.6814657955332992e-09,
+      "loss": 0.3052,
+      "step": 12996
+    },
+    {
+      "epoch": 1.9888293802601376,
+      "grad_norm": 2.179646529566567,
+      "learning_rate": 1.6363290660392773e-09,
+      "loss": 0.3342,
+      "step": 12997
+    },
+    {
+      "epoch": 1.9889824024483551,
+      "grad_norm": 2.435209578355301,
+      "learning_rate": 1.591806375170757e-09,
+      "loss": 0.3043,
+      "step": 12998
+    },
+    {
+      "epoch": 1.9891354246365722,
+      "grad_norm": 2.353359008708445,
+      "learning_rate": 1.5478977256644379e-09,
+      "loss": 0.2977,
+      "step": 12999
+    },
+    {
+      "epoch": 1.9892884468247896,
+      "grad_norm": 2.2126170930853903,
+      "learning_rate": 1.5046031202159417e-09,
+      "loss": 0.2601,
+      "step": 13000
+    },
+    {
+      "epoch": 1.989441469013007,
+      "grad_norm": 2.1203776595014348,
+      "learning_rate": 1.4619225614831424e-09,
+      "loss": 0.2572,
+      "step": 13001
+    },
+    {
+      "epoch": 1.9895944912012242,
+      "grad_norm": 2.253613767274853,
+      "learning_rate": 1.4198560520883865e-09,
+      "loss": 0.2622,
+      "step": 13002
+    },
+    {
+      "epoch": 1.9897475133894416,
+      "grad_norm": 2.11290047824741,
+      "learning_rate": 1.378403594615163e-09,
+      "loss": 0.3005,
+      "step": 13003
+    },
+    {
+      "epoch": 1.9899005355776587,
+      "grad_norm": 2.1312196286465244,
+      "learning_rate": 1.3375651916092135e-09,
+      "loss": 0.2949,
+      "step": 13004
+    },
+    {
+      "epoch": 1.990053557765876,
+      "grad_norm": 2.3139439421090287,
+      "learning_rate": 1.297340845578532e-09,
+      "loss": 0.2974,
+      "step": 13005
+    },
+    {
+      "epoch": 1.9902065799540933,
+      "grad_norm": 1.9727633886230713,
+      "learning_rate": 1.2577305589933642e-09,
+      "loss": 0.2459,
+      "step": 13006
+    },
+    {
+      "epoch": 1.9903596021423107,
+      "grad_norm": 2.0438809630472794,
+      "learning_rate": 1.218734334286209e-09,
+      "loss": 0.2784,
+      "step": 13007
+    },
+    {
+      "epoch": 1.990512624330528,
+      "grad_norm": 1.9332274761646762,
+      "learning_rate": 1.1803521738507072e-09,
+      "loss": 0.3422,
+      "step": 13008
+    },
+    {
+      "epoch": 1.9906656465187451,
+      "grad_norm": 1.9007967207990857,
+      "learning_rate": 1.1425840800471933e-09,
+      "loss": 0.254,
+      "step": 13009
+    },
+    {
+      "epoch": 1.9908186687069627,
+      "grad_norm": 2.1458707099063057,
+      "learning_rate": 1.1054300551927022e-09,
+      "loss": 0.2593,
+      "step": 13010
+    },
+    {
+      "epoch": 1.9909716908951798,
+      "grad_norm": 1.9056536998552065,
+      "learning_rate": 1.0688901015687425e-09,
+      "loss": 0.233,
+      "step": 13011
+    },
+    {
+      "epoch": 1.9911247130833971,
+      "grad_norm": 2.3772144145607546,
+      "learning_rate": 1.0329642214212953e-09,
+      "loss": 0.313,
+      "step": 13012
+    },
+    {
+      "epoch": 1.9912777352716144,
+      "grad_norm": 2.0814580331247927,
+      "learning_rate": 9.976524169552638e-10,
+      "loss": 0.2907,
+      "step": 13013
+    },
+    {
+      "epoch": 1.9914307574598316,
+      "grad_norm": 2.0239695611574064,
+      "learning_rate": 9.629546903400232e-10,
+      "loss": 0.2643,
+      "step": 13014
+    },
+    {
+      "epoch": 1.9915837796480491,
+      "grad_norm": 2.5593845320658444,
+      "learning_rate": 9.288710437060922e-10,
+      "loss": 0.3326,
+      "step": 13015
+    },
+    {
+      "epoch": 1.9917368018362662,
+      "grad_norm": 2.48462633721508,
+      "learning_rate": 8.954014791473509e-10,
+      "loss": 0.2874,
+      "step": 13016
+    },
+    {
+      "epoch": 1.9918898240244836,
+      "grad_norm": 1.883559842168018,
+      "learning_rate": 8.62545998717712e-10,
+      "loss": 0.1955,
+      "step": 13017
+    },
+    {
+      "epoch": 1.992042846212701,
+      "grad_norm": 1.7267492856950604,
+      "learning_rate": 8.303046044366713e-10,
+      "loss": 0.2448,
+      "step": 13018
+    },
+    {
+      "epoch": 1.992195868400918,
+      "grad_norm": 2.033471506716017,
+      "learning_rate": 7.986772982826463e-10,
+      "loss": 0.2754,
+      "step": 13019
+    },
+    {
+      "epoch": 1.9923488905891356,
+      "grad_norm": 1.8562161515444404,
+      "learning_rate": 7.676640821996373e-10,
+      "loss": 0.2387,
+      "step": 13020
+    },
+    {
+      "epoch": 1.9925019127773527,
+      "grad_norm": 2.150187407301561,
+      "learning_rate": 7.372649580916768e-10,
+      "loss": 0.2375,
+      "step": 13021
+    },
+    {
+      "epoch": 1.99265493496557,
+      "grad_norm": 1.9946054431849656,
+      "learning_rate": 7.074799278261602e-10,
+      "loss": 0.2818,
+      "step": 13022
+    },
+    {
+      "epoch": 1.9928079571537873,
+      "grad_norm": 1.7498242022545936,
+      "learning_rate": 6.783089932305142e-10,
+      "loss": 0.207,
+      "step": 13023
+    },
+    {
+      "epoch": 1.9929609793420044,
+      "grad_norm": 2.1900542288629032,
+      "learning_rate": 6.497521560977494e-10,
+      "loss": 0.278,
+      "step": 13024
+    },
+    {
+      "epoch": 1.993114001530222,
+      "grad_norm": 2.1975617817398447,
+      "learning_rate": 6.218094181820177e-10,
+      "loss": 0.237,
+      "step": 13025
+    },
+    {
+      "epoch": 1.9932670237184391,
+      "grad_norm": 1.8776794834042783,
+      "learning_rate": 5.944807811986142e-10,
+      "loss": 0.2505,
+      "step": 13026
+    },
+    {
+      "epoch": 1.9934200459066564,
+      "grad_norm": 1.9385495670411852,
+      "learning_rate": 5.677662468250855e-10,
+      "loss": 0.2213,
+      "step": 13027
+    },
+    {
+      "epoch": 1.9935730680948738,
+      "grad_norm": 2.0533741177883713,
+      "learning_rate": 5.416658167045619e-10,
+      "loss": 0.285,
+      "step": 13028
+    },
+    {
+      "epoch": 1.993726090283091,
+      "grad_norm": 2.1770054768883202,
+      "learning_rate": 5.161794924368746e-10,
+      "loss": 0.2512,
+      "step": 13029
+    },
+    {
+      "epoch": 1.9938791124713084,
+      "grad_norm": 2.305230017581347,
+      "learning_rate": 4.913072755896586e-10,
+      "loss": 0.3359,
+      "step": 13030
+    },
+    {
+      "epoch": 1.9940321346595256,
+      "grad_norm": 2.002526031764342,
+      "learning_rate": 4.670491676894706e-10,
+      "loss": 0.2198,
+      "step": 13031
+    },
+    {
+      "epoch": 1.994185156847743,
+      "grad_norm": 2.0269799838385905,
+      "learning_rate": 4.434051702262299e-10,
+      "loss": 0.2523,
+      "step": 13032
+    },
+    {
+      "epoch": 1.9943381790359602,
+      "grad_norm": 1.9888908831604968,
+      "learning_rate": 4.203752846521081e-10,
+      "loss": 0.2865,
+      "step": 13033
+    },
+    {
+      "epoch": 1.9944912012241776,
+      "grad_norm": 1.9144641539035228,
+      "learning_rate": 3.979595123815294e-10,
+      "loss": 0.2994,
+      "step": 13034
+    },
+    {
+      "epoch": 1.994644223412395,
+      "grad_norm": 1.776279721281176,
+      "learning_rate": 3.761578547900602e-10,
+      "loss": 0.1904,
+      "step": 13035
+    },
+    {
+      "epoch": 1.994797245600612,
+      "grad_norm": 2.298937100217351,
+      "learning_rate": 3.549703132188498e-10,
+      "loss": 0.2974,
+      "step": 13036
+    },
+    {
+      "epoch": 1.9949502677888293,
+      "grad_norm": 1.919240065983227,
+      "learning_rate": 3.343968889668592e-10,
+      "loss": 0.2379,
+      "step": 13037
+    },
+    {
+      "epoch": 1.9951032899770467,
+      "grad_norm": 2.0107139718990346,
+      "learning_rate": 3.1443758329752214e-10,
+      "loss": 0.2555,
+      "step": 13038
+    },
+    {
+      "epoch": 1.995256312165264,
+      "grad_norm": 2.1252763337912985,
+      "learning_rate": 2.950923974387454e-10,
+      "loss": 0.2751,
+      "step": 13039
+    },
+    {
+      "epoch": 1.9954093343534813,
+      "grad_norm": 2.411812539166475,
+      "learning_rate": 2.763613325773573e-10,
+      "loss": 0.3436,
+      "step": 13040
+    },
+    {
+      "epoch": 1.9955623565416984,
+      "grad_norm": 1.9107329649484788,
+      "learning_rate": 2.5824438986354895e-10,
+      "loss": 0.2166,
+      "step": 13041
+    },
+    {
+      "epoch": 1.995715378729916,
+      "grad_norm": 2.358934465004862,
+      "learning_rate": 2.4074157040976377e-10,
+      "loss": 0.2485,
+      "step": 13042
+    },
+    {
+      "epoch": 1.9958684009181331,
+      "grad_norm": 2.215031190581882,
+      "learning_rate": 2.2385287529180788e-10,
+      "loss": 0.2784,
+      "step": 13043
+    },
+    {
+      "epoch": 1.9960214231063504,
+      "grad_norm": 2.236157444646783,
+      "learning_rate": 2.075783055466296e-10,
+      "loss": 0.4084,
+      "step": 13044
+    },
+    {
+      "epoch": 1.9961744452945678,
+      "grad_norm": 2.0773031268943996,
+      "learning_rate": 1.9191786217342967e-10,
+      "loss": 0.2735,
+      "step": 13045
+    },
+    {
+      "epoch": 1.996327467482785,
+      "grad_norm": 1.783273240684617,
+      "learning_rate": 1.768715461336612e-10,
+      "loss": 0.2927,
+      "step": 13046
+    },
+    {
+      "epoch": 1.9964804896710024,
+      "grad_norm": 1.9397137676314884,
+      "learning_rate": 1.6243935835213998e-10,
+      "loss": 0.2428,
+      "step": 13047
+    },
+    {
+      "epoch": 1.9966335118592196,
+      "grad_norm": 1.8943175353851476,
+      "learning_rate": 1.4862129971371375e-10,
+      "loss": 0.2924,
+      "step": 13048
+    },
+    {
+      "epoch": 1.996786534047437,
+      "grad_norm": 2.1136284004737766,
+      "learning_rate": 1.3541737106881336e-10,
+      "loss": 0.2756,
+      "step": 13049
+    },
+    {
+      "epoch": 1.9969395562356542,
+      "grad_norm": 1.8973543054141213,
+      "learning_rate": 1.2282757322790162e-10,
+      "loss": 0.2196,
+      "step": 13050
+    },
+    {
+      "epoch": 1.9970925784238713,
+      "grad_norm": 2.0657394153839554,
+      "learning_rate": 1.1085190696369374e-10,
+      "loss": 0.2822,
+      "step": 13051
+    },
+    {
+      "epoch": 1.997245600612089,
+      "grad_norm": 1.8308499398834328,
+      "learning_rate": 9.94903730122676e-11,
+      "loss": 0.2253,
+      "step": 13052
+    },
+    {
+      "epoch": 1.997398622800306,
+      "grad_norm": 1.8283706378478568,
+      "learning_rate": 8.874297207084326e-11,
+      "loss": 0.2385,
+      "step": 13053
+    },
+    {
+      "epoch": 1.9975516449885233,
+      "grad_norm": 1.9678986030662173,
+      "learning_rate": 7.860970479889318e-11,
+      "loss": 0.3014,
+      "step": 13054
+    },
+    {
+      "epoch": 1.9977046671767407,
+      "grad_norm": 1.8653265671577366,
+      "learning_rate": 6.909057181925249e-11,
+      "loss": 0.2535,
+      "step": 13055
+    },
+    {
+      "epoch": 1.9978576893649578,
+      "grad_norm": 2.581058041189355,
+      "learning_rate": 6.018557371811895e-11,
+      "loss": 0.3192,
+      "step": 13056
+    },
+    {
+      "epoch": 1.9980107115531753,
+      "grad_norm": 2.005465358101594,
+      "learning_rate": 5.1894711039501836e-11,
+      "loss": 0.219,
+      "step": 13057
+    },
+    {
+      "epoch": 1.9981637337413924,
+      "grad_norm": 2.0811354665325625,
+      "learning_rate": 4.421798429521396e-11,
+      "loss": 0.2814,
+      "step": 13058
+    },
+    {
+      "epoch": 1.9983167559296098,
+      "grad_norm": 2.1460837784604236,
+      "learning_rate": 3.715539395487966e-11,
+      "loss": 0.2241,
+      "step": 13059
+    },
+    {
+      "epoch": 1.9984697781178271,
+      "grad_norm": 2.081856996888235,
+      "learning_rate": 3.0706940452596145e-11,
+      "loss": 0.242,
+      "step": 13060
+    },
+    {
+      "epoch": 1.9986228003060442,
+      "grad_norm": 2.227922939830152,
+      "learning_rate": 2.4872624185823258e-11,
+      "loss": 0.2824,
+      "step": 13061
+    },
+    {
+      "epoch": 1.9987758224942618,
+      "grad_norm": 2.3071203921932977,
+      "learning_rate": 1.9652445510942586e-11,
+      "loss": 0.2726,
+      "step": 13062
+    },
+    {
+      "epoch": 1.998928844682479,
+      "grad_norm": 2.4540772967670152,
+      "learning_rate": 1.5046404748808585e-11,
+      "loss": 0.2448,
+      "step": 13063
+    },
+    {
+      "epoch": 1.9990818668706962,
+      "grad_norm": 1.9558988952277159,
+      "learning_rate": 1.1054502183638349e-11,
+      "loss": 0.2816,
+      "step": 13064
+    },
+    {
+      "epoch": 1.9992348890589136,
+      "grad_norm": 2.437269378289497,
+      "learning_rate": 7.676738059680944e-12,
+      "loss": 0.366,
+      "step": 13065
+    },
+    {
+      "epoch": 1.999387911247131,
+      "grad_norm": 2.309948760100383,
+      "learning_rate": 4.913112584548074e-12,
+      "loss": 0.2803,
+      "step": 13066
+    },
+    {
+      "epoch": 1.9995409334353482,
+      "grad_norm": 2.4610462328848994,
+      "learning_rate": 2.7636259281038637e-12,
+      "loss": 0.3342,
+      "step": 13067
+    },
+    {
+      "epoch": 1.9996939556235653,
+      "grad_norm": 2.4319473298190433,
+      "learning_rate": 1.2282782213546284e-12,
+      "loss": 0.2665,
+      "step": 13068
+    },
+    {
+      "epoch": 1.9998469778117827,
+      "grad_norm": 1.9865301756122895,
+      "learning_rate": 3.070695597795492e-13,
+      "loss": 0.2713,
+      "step": 13069
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.437039631382003,
+      "learning_rate": 0.0,
+      "loss": 0.3371,
+      "step": 13070
+    },
+    {
+      "epoch": 2.0,
+      "step": 13070,
+      "total_flos": 2.021711277939753e+19,
+      "train_loss": 0.5611064869115256,
+      "train_runtime": 74118.7755,
+      "train_samples_per_second": 39.019,
+      "train_steps_per_second": 0.176
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 13070,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.021711277939753e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}