{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5871,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005109862033725089,
      "grad_norm": 57.87140382113745,
      "learning_rate": 5e-06,
      "loss": 2.3817,
      "step": 10
    },
    {
      "epoch": 0.010219724067450179,
      "grad_norm": 22.753974610258698,
      "learning_rate": 5e-06,
      "loss": 1.7816,
      "step": 20
    },
    {
      "epoch": 0.015329586101175269,
      "grad_norm": 10.200028852824934,
      "learning_rate": 5e-06,
      "loss": 1.4328,
      "step": 30
    },
    {
      "epoch": 0.020439448134900357,
      "grad_norm": 10.263199576103078,
      "learning_rate": 5e-06,
      "loss": 1.2383,
      "step": 40
    },
    {
      "epoch": 0.025549310168625446,
      "grad_norm": 4.8615267890156915,
      "learning_rate": 5e-06,
      "loss": 1.131,
      "step": 50
    },
    {
      "epoch": 0.030659172202350538,
      "grad_norm": 4.118120498213385,
      "learning_rate": 5e-06,
      "loss": 1.0356,
      "step": 60
    },
    {
      "epoch": 0.03576903423607562,
      "grad_norm": 2.8275806453598338,
      "learning_rate": 5e-06,
      "loss": 1.0006,
      "step": 70
    },
    {
      "epoch": 0.040878896269800714,
      "grad_norm": 2.315792934153761,
      "learning_rate": 5e-06,
      "loss": 0.9655,
      "step": 80
    },
    {
      "epoch": 0.045988758303525806,
      "grad_norm": 2.165672978783245,
      "learning_rate": 5e-06,
      "loss": 0.9467,
      "step": 90
    },
    {
      "epoch": 0.05109862033725089,
      "grad_norm": 1.5488035180810096,
      "learning_rate": 5e-06,
      "loss": 0.9281,
      "step": 100
    },
    {
      "epoch": 0.05620848237097598,
      "grad_norm": 1.7243341087325224,
      "learning_rate": 5e-06,
      "loss": 0.9128,
      "step": 110
    },
    {
      "epoch": 0.061318344404701075,
      "grad_norm": 1.5359895854005425,
      "learning_rate": 5e-06,
      "loss": 0.9031,
      "step": 120
    },
    {
      "epoch": 0.06642820643842616,
      "grad_norm": 1.435850065371256,
      "learning_rate": 5e-06,
      "loss": 0.8922,
      "step": 130
    },
    {
      "epoch": 0.07153806847215124,
      "grad_norm": 1.2790521202391505,
      "learning_rate": 5e-06,
      "loss": 0.8765,
      "step": 140
    },
    {
      "epoch": 0.07664793050587634,
      "grad_norm": 1.5917071745489466,
      "learning_rate": 5e-06,
      "loss": 0.883,
      "step": 150
    },
    {
      "epoch": 0.08175779253960143,
      "grad_norm": 1.4686990635112733,
      "learning_rate": 5e-06,
      "loss": 0.8874,
      "step": 160
    },
    {
      "epoch": 0.08686765457332651,
      "grad_norm": 1.2956583323307207,
      "learning_rate": 5e-06,
      "loss": 0.8411,
      "step": 170
    },
    {
      "epoch": 0.09197751660705161,
      "grad_norm": 1.2357587318309058,
      "learning_rate": 5e-06,
      "loss": 0.8726,
      "step": 180
    },
    {
      "epoch": 0.0970873786407767,
      "grad_norm": 1.1075519305139339,
      "learning_rate": 5e-06,
      "loss": 0.8548,
      "step": 190
    },
    {
      "epoch": 0.10219724067450178,
      "grad_norm": 1.086727474202395,
      "learning_rate": 5e-06,
      "loss": 0.8493,
      "step": 200
    },
    {
      "epoch": 0.10730710270822688,
      "grad_norm": 1.0270014162441539,
      "learning_rate": 5e-06,
      "loss": 0.8545,
      "step": 210
    },
    {
      "epoch": 0.11241696474195197,
      "grad_norm": 0.8505362329385927,
      "learning_rate": 5e-06,
      "loss": 0.8388,
      "step": 220
    },
    {
      "epoch": 0.11752682677567705,
      "grad_norm": 1.109098072362057,
      "learning_rate": 5e-06,
      "loss": 0.8384,
      "step": 230
    },
    {
      "epoch": 0.12263668880940215,
      "grad_norm": 1.2809391093206264,
      "learning_rate": 5e-06,
      "loss": 0.8471,
      "step": 240
    },
    {
      "epoch": 0.12774655084312725,
      "grad_norm": 0.9245822751345097,
      "learning_rate": 5e-06,
      "loss": 0.836,
      "step": 250
    },
    {
      "epoch": 0.13285641287685232,
      "grad_norm": 0.8873874884959316,
      "learning_rate": 5e-06,
      "loss": 0.8283,
      "step": 260
    },
    {
      "epoch": 0.13796627491057742,
      "grad_norm": 0.6839627248028421,
      "learning_rate": 5e-06,
      "loss": 0.8191,
      "step": 270
    },
    {
      "epoch": 0.1430761369443025,
      "grad_norm": 0.8006134542844254,
      "learning_rate": 5e-06,
      "loss": 0.82,
      "step": 280
    },
    {
      "epoch": 0.1481859989780276,
      "grad_norm": 0.7249813175028869,
      "learning_rate": 5e-06,
      "loss": 0.8254,
      "step": 290
    },
    {
      "epoch": 0.1532958610117527,
      "grad_norm": 0.611471064933341,
      "learning_rate": 5e-06,
      "loss": 0.8148,
      "step": 300
    },
    {
      "epoch": 0.15840572304547776,
      "grad_norm": 0.8403753613185879,
      "learning_rate": 5e-06,
      "loss": 0.8091,
      "step": 310
    },
    {
      "epoch": 0.16351558507920286,
      "grad_norm": 0.5250234838001376,
      "learning_rate": 5e-06,
      "loss": 0.8328,
      "step": 320
    },
    {
      "epoch": 0.16862544711292796,
      "grad_norm": 0.5411140081911021,
      "learning_rate": 5e-06,
      "loss": 0.8171,
      "step": 330
    },
    {
      "epoch": 0.17373530914665303,
      "grad_norm": 0.7002721154510084,
      "learning_rate": 5e-06,
      "loss": 0.8073,
      "step": 340
    },
    {
      "epoch": 0.17884517118037813,
      "grad_norm": 0.7366019714677299,
      "learning_rate": 5e-06,
      "loss": 0.8139,
      "step": 350
    },
    {
      "epoch": 0.18395503321410323,
      "grad_norm": 0.4960107622038265,
      "learning_rate": 5e-06,
      "loss": 0.8195,
      "step": 360
    },
    {
      "epoch": 0.1890648952478283,
      "grad_norm": 0.590921742782821,
      "learning_rate": 5e-06,
      "loss": 0.8131,
      "step": 370
    },
    {
      "epoch": 0.1941747572815534,
      "grad_norm": 0.43046345355299165,
      "learning_rate": 5e-06,
      "loss": 0.8166,
      "step": 380
    },
    {
      "epoch": 0.1992846193152785,
      "grad_norm": 0.37156340060616266,
      "learning_rate": 5e-06,
      "loss": 0.8082,
      "step": 390
    },
    {
      "epoch": 0.20439448134900357,
      "grad_norm": 0.5029827349762587,
      "learning_rate": 5e-06,
      "loss": 0.8187,
      "step": 400
    },
    {
      "epoch": 0.20950434338272866,
      "grad_norm": 0.39268177137347837,
      "learning_rate": 5e-06,
      "loss": 0.8132,
      "step": 410
    },
    {
      "epoch": 0.21461420541645376,
      "grad_norm": 0.38787210509179787,
      "learning_rate": 5e-06,
      "loss": 0.8153,
      "step": 420
    },
    {
      "epoch": 0.21972406745017883,
      "grad_norm": 0.5073537687615494,
      "learning_rate": 5e-06,
      "loss": 0.8032,
      "step": 430
    },
    {
      "epoch": 0.22483392948390393,
      "grad_norm": 0.36232141966438186,
      "learning_rate": 5e-06,
      "loss": 0.786,
      "step": 440
    },
    {
      "epoch": 0.22994379151762903,
      "grad_norm": 0.3362374048828632,
      "learning_rate": 5e-06,
      "loss": 0.8019,
      "step": 450
    },
    {
      "epoch": 0.2350536535513541,
      "grad_norm": 0.3661481336901712,
      "learning_rate": 5e-06,
      "loss": 0.8012,
      "step": 460
    },
    {
      "epoch": 0.2401635155850792,
      "grad_norm": 0.3186478374299407,
      "learning_rate": 5e-06,
      "loss": 0.8088,
      "step": 470
    },
    {
      "epoch": 0.2452733776188043,
      "grad_norm": 0.3742412176358808,
      "learning_rate": 5e-06,
      "loss": 0.795,
      "step": 480
    },
    {
      "epoch": 0.2503832396525294,
      "grad_norm": 0.3687134051282934,
      "learning_rate": 5e-06,
      "loss": 0.7899,
      "step": 490
    },
    {
      "epoch": 0.2554931016862545,
      "grad_norm": 0.2844684124400157,
      "learning_rate": 5e-06,
      "loss": 0.7992,
      "step": 500
    },
    {
      "epoch": 0.26060296371997954,
      "grad_norm": 0.3445705328013273,
      "learning_rate": 5e-06,
      "loss": 0.7908,
      "step": 510
    },
    {
      "epoch": 0.26571282575370464,
      "grad_norm": 0.3100104564359116,
      "learning_rate": 5e-06,
      "loss": 0.7874,
      "step": 520
    },
    {
      "epoch": 0.27082268778742974,
      "grad_norm": 0.32495110543783357,
      "learning_rate": 5e-06,
      "loss": 0.7982,
      "step": 530
    },
    {
      "epoch": 0.27593254982115484,
      "grad_norm": 0.27997656185948167,
      "learning_rate": 5e-06,
      "loss": 0.7954,
      "step": 540
    },
    {
      "epoch": 0.28104241185487994,
      "grad_norm": 0.4055280072220847,
      "learning_rate": 5e-06,
      "loss": 0.7895,
      "step": 550
    },
    {
      "epoch": 0.286152273888605,
      "grad_norm": 0.2751499734062912,
      "learning_rate": 5e-06,
      "loss": 0.7931,
      "step": 560
    },
    {
      "epoch": 0.2912621359223301,
      "grad_norm": 0.3232277546201772,
      "learning_rate": 5e-06,
      "loss": 0.7972,
      "step": 570
    },
    {
      "epoch": 0.2963719979560552,
      "grad_norm": 0.2915040535303487,
      "learning_rate": 5e-06,
      "loss": 0.7912,
      "step": 580
    },
    {
      "epoch": 0.3014818599897803,
      "grad_norm": 0.2637267252402153,
      "learning_rate": 5e-06,
      "loss": 0.7903,
      "step": 590
    },
    {
      "epoch": 0.3065917220235054,
      "grad_norm": 0.329139944193737,
      "learning_rate": 5e-06,
      "loss": 0.7906,
      "step": 600
    },
    {
      "epoch": 0.3117015840572305,
      "grad_norm": 0.24558162809212852,
      "learning_rate": 5e-06,
      "loss": 0.7947,
      "step": 610
    },
    {
      "epoch": 0.3168114460909555,
      "grad_norm": 0.27224578177967845,
      "learning_rate": 5e-06,
      "loss": 0.7998,
      "step": 620
    },
    {
      "epoch": 0.3219213081246806,
      "grad_norm": 0.2379116871443806,
      "learning_rate": 5e-06,
      "loss": 0.7702,
      "step": 630
    },
    {
      "epoch": 0.3270311701584057,
      "grad_norm": 0.2609800304013704,
      "learning_rate": 5e-06,
      "loss": 0.7799,
      "step": 640
    },
    {
      "epoch": 0.3321410321921308,
      "grad_norm": 0.23810454322283608,
      "learning_rate": 5e-06,
      "loss": 0.7897,
      "step": 650
    },
    {
      "epoch": 0.3372508942258559,
      "grad_norm": 0.252487995051141,
      "learning_rate": 5e-06,
      "loss": 0.7726,
      "step": 660
    },
    {
      "epoch": 0.342360756259581,
      "grad_norm": 0.24925719360242624,
      "learning_rate": 5e-06,
      "loss": 0.7978,
      "step": 670
    },
    {
      "epoch": 0.34747061829330605,
      "grad_norm": 0.2656149332274455,
      "learning_rate": 5e-06,
      "loss": 0.7915,
      "step": 680
    },
    {
      "epoch": 0.35258048032703115,
      "grad_norm": 0.2661464494899381,
      "learning_rate": 5e-06,
      "loss": 0.78,
      "step": 690
    },
    {
      "epoch": 0.35769034236075625,
      "grad_norm": 0.20145837945729328,
      "learning_rate": 5e-06,
      "loss": 0.7734,
      "step": 700
    },
    {
      "epoch": 0.36280020439448135,
      "grad_norm": 0.26148666494272177,
      "learning_rate": 5e-06,
      "loss": 0.7792,
      "step": 710
    },
    {
      "epoch": 0.36791006642820645,
      "grad_norm": 0.24346683962909085,
      "learning_rate": 5e-06,
      "loss": 0.7816,
      "step": 720
    },
    {
      "epoch": 0.37301992846193155,
      "grad_norm": 0.22278218589963927,
      "learning_rate": 5e-06,
      "loss": 0.7827,
      "step": 730
    },
    {
      "epoch": 0.3781297904956566,
      "grad_norm": 0.2303051334399858,
      "learning_rate": 5e-06,
      "loss": 0.7707,
      "step": 740
    },
    {
      "epoch": 0.3832396525293817,
      "grad_norm": 0.2543406962739334,
      "learning_rate": 5e-06,
      "loss": 0.7721,
      "step": 750
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 0.22568609222745584,
      "learning_rate": 5e-06,
      "loss": 0.7836,
      "step": 760
    },
    {
      "epoch": 0.3934593765968319,
      "grad_norm": 0.2294845661519681,
      "learning_rate": 5e-06,
      "loss": 0.7872,
      "step": 770
    },
    {
      "epoch": 0.398569238630557,
      "grad_norm": 0.2197759747155408,
      "learning_rate": 5e-06,
      "loss": 0.7681,
      "step": 780
    },
    {
      "epoch": 0.4036791006642821,
      "grad_norm": 0.22484907906751414,
      "learning_rate": 5e-06,
      "loss": 0.7797,
      "step": 790
    },
    {
      "epoch": 0.40878896269800713,
      "grad_norm": 0.21381439198625019,
      "learning_rate": 5e-06,
      "loss": 0.7891,
      "step": 800
    },
    {
      "epoch": 0.41389882473173223,
      "grad_norm": 0.21726381466920616,
      "learning_rate": 5e-06,
      "loss": 0.7779,
      "step": 810
    },
    {
      "epoch": 0.4190086867654573,
      "grad_norm": 0.21218649602843365,
      "learning_rate": 5e-06,
      "loss": 0.7756,
      "step": 820
    },
    {
      "epoch": 0.4241185487991824,
      "grad_norm": 0.2321703903701098,
      "learning_rate": 5e-06,
      "loss": 0.778,
      "step": 830
    },
    {
      "epoch": 0.4292284108329075,
      "grad_norm": 0.21376686462742833,
      "learning_rate": 5e-06,
      "loss": 0.7754,
      "step": 840
    },
    {
      "epoch": 0.4343382728666326,
      "grad_norm": 0.20773121342846562,
      "learning_rate": 5e-06,
      "loss": 0.7729,
      "step": 850
    },
    {
      "epoch": 0.43944813490035767,
      "grad_norm": 0.23075209429146118,
      "learning_rate": 5e-06,
      "loss": 0.7743,
      "step": 860
    },
    {
      "epoch": 0.44455799693408277,
      "grad_norm": 0.21889813942522324,
      "learning_rate": 5e-06,
      "loss": 0.7872,
      "step": 870
    },
    {
      "epoch": 0.44966785896780787,
      "grad_norm": 0.23158111213452945,
      "learning_rate": 5e-06,
      "loss": 0.7729,
      "step": 880
    },
    {
      "epoch": 0.45477772100153296,
      "grad_norm": 0.23759359366214816,
      "learning_rate": 5e-06,
      "loss": 0.7781,
      "step": 890
    },
    {
      "epoch": 0.45988758303525806,
      "grad_norm": 0.20737512188117463,
      "learning_rate": 5e-06,
      "loss": 0.7733,
      "step": 900
    },
    {
      "epoch": 0.46499744506898316,
      "grad_norm": 0.21978896076440746,
      "learning_rate": 5e-06,
      "loss": 0.7724,
      "step": 910
    },
    {
      "epoch": 0.4701073071027082,
      "grad_norm": 0.21326048741734183,
      "learning_rate": 5e-06,
      "loss": 0.7779,
      "step": 920
    },
    {
      "epoch": 0.4752171691364333,
      "grad_norm": 0.2127733995734116,
      "learning_rate": 5e-06,
      "loss": 0.7862,
      "step": 930
    },
    {
      "epoch": 0.4803270311701584,
      "grad_norm": 0.20545254783011138,
      "learning_rate": 5e-06,
      "loss": 0.7741,
      "step": 940
    },
    {
      "epoch": 0.4854368932038835,
      "grad_norm": 0.22401938020775877,
      "learning_rate": 5e-06,
      "loss": 0.7879,
      "step": 950
    },
    {
      "epoch": 0.4905467552376086,
      "grad_norm": 0.20313108726519555,
      "learning_rate": 5e-06,
      "loss": 0.7795,
      "step": 960
    },
    {
      "epoch": 0.4956566172713337,
      "grad_norm": 0.20676236489485517,
      "learning_rate": 5e-06,
      "loss": 0.7658,
      "step": 970
    },
    {
      "epoch": 0.5007664793050588,
      "grad_norm": 0.1997349100496237,
      "learning_rate": 5e-06,
      "loss": 0.7665,
      "step": 980
    },
    {
      "epoch": 0.5058763413387839,
      "grad_norm": 0.18582330695996696,
      "learning_rate": 5e-06,
      "loss": 0.7678,
      "step": 990
    },
    {
      "epoch": 0.510986203372509,
      "grad_norm": 0.20480578385489007,
      "learning_rate": 5e-06,
      "loss": 0.7916,
      "step": 1000
    },
    {
      "epoch": 0.516096065406234,
      "grad_norm": 0.2068595244192038,
      "learning_rate": 5e-06,
      "loss": 0.7728,
      "step": 1010
    },
    {
      "epoch": 0.5212059274399591,
      "grad_norm": 0.19202510728167327,
      "learning_rate": 5e-06,
      "loss": 0.7875,
      "step": 1020
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.20142193391262955,
      "learning_rate": 5e-06,
      "loss": 0.7691,
      "step": 1030
    },
    {
      "epoch": 0.5314256515074093,
      "grad_norm": 0.19153357968487428,
      "learning_rate": 5e-06,
      "loss": 0.7717,
      "step": 1040
    },
    {
      "epoch": 0.5365355135411344,
      "grad_norm": 0.21070221575444148,
      "learning_rate": 5e-06,
      "loss": 0.771,
      "step": 1050
    },
    {
      "epoch": 0.5416453755748595,
      "grad_norm": 0.19388651883167224,
      "learning_rate": 5e-06,
      "loss": 0.7674,
      "step": 1060
    },
    {
      "epoch": 0.5467552376085846,
      "grad_norm": 0.19235966239305358,
      "learning_rate": 5e-06,
      "loss": 0.7639,
      "step": 1070
    },
    {
      "epoch": 0.5518650996423097,
      "grad_norm": 0.2020128053771939,
      "learning_rate": 5e-06,
      "loss": 0.7838,
      "step": 1080
    },
    {
      "epoch": 0.5569749616760348,
      "grad_norm": 0.19898211373916952,
      "learning_rate": 5e-06,
      "loss": 0.7734,
      "step": 1090
    },
    {
      "epoch": 0.5620848237097599,
      "grad_norm": 0.19556108367687597,
      "learning_rate": 5e-06,
      "loss": 0.7668,
      "step": 1100
    },
    {
      "epoch": 0.567194685743485,
      "grad_norm": 0.19409586285487196,
      "learning_rate": 5e-06,
      "loss": 0.7778,
      "step": 1110
    },
    {
      "epoch": 0.57230454777721,
      "grad_norm": 0.20681598650606026,
      "learning_rate": 5e-06,
      "loss": 0.779,
      "step": 1120
    },
    {
      "epoch": 0.5774144098109351,
      "grad_norm": 0.2017546177375358,
      "learning_rate": 5e-06,
      "loss": 0.7573,
      "step": 1130
    },
    {
      "epoch": 0.5825242718446602,
      "grad_norm": 0.1913755514728684,
      "learning_rate": 5e-06,
      "loss": 0.7625,
      "step": 1140
    },
    {
      "epoch": 0.5876341338783853,
      "grad_norm": 0.18303523769215124,
      "learning_rate": 5e-06,
      "loss": 0.7752,
      "step": 1150
    },
    {
      "epoch": 0.5927439959121104,
      "grad_norm": 0.19646699079126273,
      "learning_rate": 5e-06,
      "loss": 0.7734,
      "step": 1160
    },
    {
      "epoch": 0.5978538579458355,
      "grad_norm": 0.18589450588787854,
      "learning_rate": 5e-06,
      "loss": 0.7631,
      "step": 1170
    },
    {
      "epoch": 0.6029637199795606,
      "grad_norm": 0.19141416094028266,
      "learning_rate": 5e-06,
      "loss": 0.7721,
      "step": 1180
    },
    {
      "epoch": 0.6080735820132857,
      "grad_norm": 0.2010700371236207,
      "learning_rate": 5e-06,
      "loss": 0.7723,
      "step": 1190
    },
    {
      "epoch": 0.6131834440470108,
      "grad_norm": 0.209693144160535,
      "learning_rate": 5e-06,
      "loss": 0.7709,
      "step": 1200
    },
    {
      "epoch": 0.6182933060807358,
      "grad_norm": 0.19006264249604685,
      "learning_rate": 5e-06,
      "loss": 0.7731,
      "step": 1210
    },
    {
      "epoch": 0.623403168114461,
      "grad_norm": 0.19625031587081004,
      "learning_rate": 5e-06,
      "loss": 0.7683,
      "step": 1220
    },
    {
      "epoch": 0.628513030148186,
      "grad_norm": 0.19078272960681916,
      "learning_rate": 5e-06,
      "loss": 0.773,
      "step": 1230
    },
    {
      "epoch": 0.633622892181911,
      "grad_norm": 0.2007688157803181,
      "learning_rate": 5e-06,
      "loss": 0.7533,
      "step": 1240
    },
    {
      "epoch": 0.6387327542156361,
      "grad_norm": 0.20004504142951482,
      "learning_rate": 5e-06,
      "loss": 0.7721,
      "step": 1250
    },
    {
      "epoch": 0.6438426162493612,
      "grad_norm": 0.18940428161938747,
      "learning_rate": 5e-06,
      "loss": 0.7525,
      "step": 1260
    },
    {
      "epoch": 0.6489524782830863,
      "grad_norm": 0.19192245014604287,
      "learning_rate": 5e-06,
      "loss": 0.7743,
      "step": 1270
    },
    {
      "epoch": 0.6540623403168114,
      "grad_norm": 0.18803291404020508,
      "learning_rate": 5e-06,
      "loss": 0.7781,
      "step": 1280
    },
    {
      "epoch": 0.6591722023505365,
      "grad_norm": 0.19034763073417169,
      "learning_rate": 5e-06,
      "loss": 0.7609,
      "step": 1290
    },
    {
      "epoch": 0.6642820643842616,
      "grad_norm": 0.2000260745898994,
      "learning_rate": 5e-06,
      "loss": 0.7493,
      "step": 1300
    },
    {
      "epoch": 0.6693919264179867,
      "grad_norm": 0.19721212735320226,
      "learning_rate": 5e-06,
      "loss": 0.759,
      "step": 1310
    },
    {
      "epoch": 0.6745017884517118,
      "grad_norm": 0.1840594216152482,
      "learning_rate": 5e-06,
      "loss": 0.7554,
      "step": 1320
    },
    {
      "epoch": 0.6796116504854369,
      "grad_norm": 0.19294533218500587,
      "learning_rate": 5e-06,
      "loss": 0.7601,
      "step": 1330
    },
    {
      "epoch": 0.684721512519162,
      "grad_norm": 0.2003437533474394,
      "learning_rate": 5e-06,
      "loss": 0.7665,
      "step": 1340
    },
    {
      "epoch": 0.6898313745528871,
      "grad_norm": 0.19522682692250634,
      "learning_rate": 5e-06,
      "loss": 0.7682,
      "step": 1350
    },
    {
      "epoch": 0.6949412365866121,
      "grad_norm": 0.19114649796975278,
      "learning_rate": 5e-06,
      "loss": 0.7675,
      "step": 1360
    },
    {
      "epoch": 0.7000510986203372,
      "grad_norm": 0.18795490185304542,
      "learning_rate": 5e-06,
      "loss": 0.7514,
      "step": 1370
    },
    {
      "epoch": 0.7051609606540623,
      "grad_norm": 0.19982770311498202,
      "learning_rate": 5e-06,
      "loss": 0.7587,
      "step": 1380
    },
    {
      "epoch": 0.7102708226877874,
      "grad_norm": 0.2094868748722276,
      "learning_rate": 5e-06,
      "loss": 0.7424,
      "step": 1390
    },
    {
      "epoch": 0.7153806847215125,
      "grad_norm": 0.18141939898186682,
      "learning_rate": 5e-06,
      "loss": 0.7668,
      "step": 1400
    },
    {
      "epoch": 0.7204905467552376,
      "grad_norm": 0.184317543793234,
      "learning_rate": 5e-06,
      "loss": 0.7659,
      "step": 1410
    },
    {
      "epoch": 0.7256004087889627,
      "grad_norm": 0.19482359157799217,
      "learning_rate": 5e-06,
      "loss": 0.766,
      "step": 1420
    },
    {
      "epoch": 0.7307102708226878,
      "grad_norm": 0.20609416559465576,
      "learning_rate": 5e-06,
      "loss": 0.7591,
      "step": 1430
    },
    {
      "epoch": 0.7358201328564129,
      "grad_norm": 0.1733817918744796,
      "learning_rate": 5e-06,
      "loss": 0.7657,
      "step": 1440
    },
    {
      "epoch": 0.740929994890138,
      "grad_norm": 0.20231819059208814,
      "learning_rate": 5e-06,
      "loss": 0.7693,
      "step": 1450
    },
    {
      "epoch": 0.7460398569238631,
      "grad_norm": 0.19384075901742115,
      "learning_rate": 5e-06,
      "loss": 0.7677,
      "step": 1460
    },
    {
      "epoch": 0.7511497189575882,
      "grad_norm": 0.20242534213073207,
      "learning_rate": 5e-06,
      "loss": 0.7658,
      "step": 1470
    },
    {
      "epoch": 0.7562595809913132,
      "grad_norm": 0.18992152096280124,
      "learning_rate": 5e-06,
      "loss": 0.7604,
      "step": 1480
    },
    {
      "epoch": 0.7613694430250383,
      "grad_norm": 0.20300312286644698,
      "learning_rate": 5e-06,
      "loss": 0.7654,
      "step": 1490
    },
    {
      "epoch": 0.7664793050587634,
      "grad_norm": 0.2110214114358105,
      "learning_rate": 5e-06,
      "loss": 0.7579,
      "step": 1500
    },
    {
      "epoch": 0.7715891670924885,
      "grad_norm": 0.18507470375993035,
      "learning_rate": 5e-06,
      "loss": 0.7535,
      "step": 1510
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 0.20773697020515,
      "learning_rate": 5e-06,
      "loss": 0.7681,
      "step": 1520
    },
    {
      "epoch": 0.7818088911599387,
      "grad_norm": 0.1846942481775099,
      "learning_rate": 5e-06,
      "loss": 0.7589,
      "step": 1530
    },
    {
      "epoch": 0.7869187531936638,
      "grad_norm": 0.19873266556747132,
      "learning_rate": 5e-06,
      "loss": 0.7701,
      "step": 1540
    },
    {
      "epoch": 0.7920286152273889,
      "grad_norm": 0.1889722343751879,
      "learning_rate": 5e-06,
      "loss": 0.7637,
      "step": 1550
    },
    {
      "epoch": 0.797138477261114,
      "grad_norm": 0.2160555515101488,
      "learning_rate": 5e-06,
      "loss": 0.7459,
      "step": 1560
    },
    {
      "epoch": 0.8022483392948391,
      "grad_norm": 0.18958425843094595,
      "learning_rate": 5e-06,
      "loss": 0.7499,
      "step": 1570
    },
    {
      "epoch": 0.8073582013285642,
      "grad_norm": 0.18917128647246217,
      "learning_rate": 5e-06,
      "loss": 0.7572,
      "step": 1580
    },
    {
      "epoch": 0.8124680633622893,
      "grad_norm": 0.17914225162735836,
      "learning_rate": 5e-06,
      "loss": 0.7506,
      "step": 1590
    },
    {
      "epoch": 0.8175779253960143,
      "grad_norm": 0.19189511614416635,
      "learning_rate": 5e-06,
      "loss": 0.7613,
      "step": 1600
    },
    {
      "epoch": 0.8226877874297394,
      "grad_norm": 0.19059344363394998,
      "learning_rate": 5e-06,
      "loss": 0.7604,
      "step": 1610
    },
    {
      "epoch": 0.8277976494634645,
      "grad_norm": 0.20852250489781288,
      "learning_rate": 5e-06,
      "loss": 0.7737,
      "step": 1620
    },
    {
      "epoch": 0.8329075114971896,
      "grad_norm": 0.1877964215413997,
      "learning_rate": 5e-06,
      "loss": 0.7725,
      "step": 1630
    },
    {
      "epoch": 0.8380173735309147,
      "grad_norm": 0.18419774730049385,
      "learning_rate": 5e-06,
      "loss": 0.757,
      "step": 1640
    },
    {
      "epoch": 0.8431272355646398,
      "grad_norm": 0.1926784804428757,
      "learning_rate": 5e-06,
      "loss": 0.7515,
      "step": 1650
    },
    {
      "epoch": 0.8482370975983649,
      "grad_norm": 0.19627214391258455,
      "learning_rate": 5e-06,
      "loss": 0.7718,
      "step": 1660
    },
    {
      "epoch": 0.85334695963209,
      "grad_norm": 0.1944045450874202,
      "learning_rate": 5e-06,
      "loss": 0.7449,
      "step": 1670
    },
    {
      "epoch": 0.858456821665815,
      "grad_norm": 0.18125259142482736,
      "learning_rate": 5e-06,
      "loss": 0.7604,
      "step": 1680
    },
    {
      "epoch": 0.8635666836995401,
      "grad_norm": 0.17817637475202877,
      "learning_rate": 5e-06,
      "loss": 0.7641,
      "step": 1690
    },
    {
      "epoch": 0.8686765457332652,
      "grad_norm": 0.20471658407314103,
      "learning_rate": 5e-06,
      "loss": 0.7734,
      "step": 1700
    },
    {
      "epoch": 0.8737864077669902,
      "grad_norm": 0.1878161208418682,
      "learning_rate": 5e-06,
      "loss": 0.7696,
      "step": 1710
    },
    {
      "epoch": 0.8788962698007153,
      "grad_norm": 0.18334210713527221,
      "learning_rate": 5e-06,
      "loss": 0.7536,
      "step": 1720
    },
    {
      "epoch": 0.8840061318344404,
      "grad_norm": 0.18076699199021762,
      "learning_rate": 5e-06,
      "loss": 0.7522,
      "step": 1730
    },
    {
      "epoch": 0.8891159938681655,
      "grad_norm": 0.19181861440876702,
      "learning_rate": 5e-06,
      "loss": 0.7625,
      "step": 1740
    },
    {
      "epoch": 0.8942258559018906,
      "grad_norm": 0.17965104525055625,
      "learning_rate": 5e-06,
      "loss": 0.754,
      "step": 1750
    },
    {
      "epoch": 0.8993357179356157,
      "grad_norm": 0.19942739585455946,
      "learning_rate": 5e-06,
      "loss": 0.7663,
      "step": 1760
    },
    {
      "epoch": 0.9044455799693408,
      "grad_norm": 0.19441177912329213,
      "learning_rate": 5e-06,
      "loss": 0.7643,
      "step": 1770
    },
    {
      "epoch": 0.9095554420030659,
      "grad_norm": 0.1989153914067934,
      "learning_rate": 5e-06,
      "loss": 0.7538,
      "step": 1780
    },
    {
      "epoch": 0.914665304036791,
      "grad_norm": 0.18779250602110079,
      "learning_rate": 5e-06,
      "loss": 0.7582,
      "step": 1790
    },
    {
      "epoch": 0.9197751660705161,
      "grad_norm": 0.18880847142963628,
      "learning_rate": 5e-06,
      "loss": 0.7834,
      "step": 1800
    },
    {
      "epoch": 0.9248850281042412,
      "grad_norm": 0.19100671149679851,
      "learning_rate": 5e-06,
      "loss": 0.7633,
      "step": 1810
    },
    {
      "epoch": 0.9299948901379663,
      "grad_norm": 0.19156019315031683,
      "learning_rate": 5e-06,
      "loss": 0.7644,
      "step": 1820
    },
    {
      "epoch": 0.9351047521716913,
      "grad_norm": 0.18435442552610007,
      "learning_rate": 5e-06,
      "loss": 0.7597,
      "step": 1830
    },
    {
      "epoch": 0.9402146142054164,
      "grad_norm": 0.2051724242213117,
      "learning_rate": 5e-06,
      "loss": 0.7504,
      "step": 1840
    },
    {
      "epoch": 0.9453244762391415,
      "grad_norm": 0.18536587248191086,
      "learning_rate": 5e-06,
      "loss": 0.7396,
      "step": 1850
    },
    {
      "epoch": 0.9504343382728666,
      "grad_norm": 0.17780024213447235,
      "learning_rate": 5e-06,
      "loss": 0.7506,
      "step": 1860
    },
    {
      "epoch": 0.9555442003065917,
      "grad_norm": 0.18708967218779626,
      "learning_rate": 5e-06,
      "loss": 0.7455,
      "step": 1870
    },
    {
      "epoch": 0.9606540623403168,
      "grad_norm": 0.17747103600840475,
      "learning_rate": 5e-06,
      "loss": 0.7444,
      "step": 1880
    },
    {
      "epoch": 0.9657639243740419,
      "grad_norm": 0.19063629309146018,
      "learning_rate": 5e-06,
      "loss": 0.7533,
      "step": 1890
    },
    {
      "epoch": 0.970873786407767,
      "grad_norm": 0.19471871824403422,
      "learning_rate": 5e-06,
      "loss": 0.7539,
      "step": 1900
    },
    {
      "epoch": 0.9759836484414921,
      "grad_norm": 0.18403984668742995,
      "learning_rate": 5e-06,
      "loss": 0.766,
      "step": 1910
    },
    {
      "epoch": 0.9810935104752172,
      "grad_norm": 0.19270062603489418,
      "learning_rate": 5e-06,
      "loss": 0.7661,
      "step": 1920
    },
    {
      "epoch": 0.9862033725089423,
      "grad_norm": 0.19463685028697894,
      "learning_rate": 5e-06,
      "loss": 0.7591,
      "step": 1930
    },
    {
      "epoch": 0.9913132345426674,
      "grad_norm": 0.18870267371498323,
      "learning_rate": 5e-06,
      "loss": 0.7619,
      "step": 1940
    },
    {
      "epoch": 0.9964230965763924,
      "grad_norm": 0.19561509144751293,
      "learning_rate": 5e-06,
      "loss": 0.7727,
      "step": 1950
    },
    {
      "epoch": 1.0015329586101176,
      "grad_norm": 0.1865614045173204,
      "learning_rate": 5e-06,
      "loss": 0.753,
      "step": 1960
    },
    {
      "epoch": 1.0066428206438427,
      "grad_norm": 0.17671337095527262,
      "learning_rate": 5e-06,
      "loss": 0.7557,
      "step": 1970
    },
    {
      "epoch": 1.0117526826775678,
      "grad_norm": 0.19010483409505236,
      "learning_rate": 5e-06,
      "loss": 0.7446,
      "step": 1980
    },
    {
      "epoch": 1.016862544711293,
      "grad_norm": 0.17413564167981435,
      "learning_rate": 5e-06,
      "loss": 0.7389,
      "step": 1990
    },
    {
      "epoch": 1.021972406745018,
      "grad_norm": 0.18070567046481728,
      "learning_rate": 5e-06,
      "loss": 0.7451,
      "step": 2000
    },
    {
      "epoch": 1.0270822687787429,
      "grad_norm": 0.2047148011452083,
      "learning_rate": 5e-06,
      "loss": 0.7404,
      "step": 2010
    },
    {
      "epoch": 1.032192130812468,
      "grad_norm": 0.19369702939207809,
      "learning_rate": 5e-06,
      "loss": 0.7432,
      "step": 2020
    },
    {
      "epoch": 1.037301992846193,
      "grad_norm": 0.19067195646001925,
      "learning_rate": 5e-06,
      "loss": 0.7396,
      "step": 2030
    },
    {
      "epoch": 1.0424118548799182,
      "grad_norm": 0.20517054385871566,
      "learning_rate": 5e-06,
      "loss": 0.7312,
      "step": 2040
    },
    {
      "epoch": 1.0475217169136433,
      "grad_norm": 0.19412259781936536,
      "learning_rate": 5e-06,
      "loss": 0.7434,
      "step": 2050
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.1873188104877111,
      "learning_rate": 5e-06,
      "loss": 0.7411,
      "step": 2060
    },
    {
      "epoch": 1.0577414409810935,
      "grad_norm": 0.1897369700854769,
      "learning_rate": 5e-06,
      "loss": 0.7345,
      "step": 2070
    },
    {
      "epoch": 1.0628513030148186,
      "grad_norm": 0.1853976118509793,
      "learning_rate": 5e-06,
      "loss": 0.7549,
      "step": 2080
    },
    {
      "epoch": 1.0679611650485437,
      "grad_norm": 0.17933568978007214,
      "learning_rate": 5e-06,
      "loss": 0.7531,
      "step": 2090
    },
    {
      "epoch": 1.0730710270822688,
      "grad_norm": 0.20312467005419368,
      "learning_rate": 5e-06,
      "loss": 0.7498,
      "step": 2100
    },
    {
      "epoch": 1.0781808891159939,
      "grad_norm": 0.18428135458855252,
      "learning_rate": 5e-06,
      "loss": 0.7404,
      "step": 2110
    },
    {
      "epoch": 1.083290751149719,
      "grad_norm": 0.1973755972738785,
      "learning_rate": 5e-06,
      "loss": 0.7329,
      "step": 2120
    },
    {
      "epoch": 1.088400613183444,
      "grad_norm": 0.1917623224859124,
      "learning_rate": 5e-06,
      "loss": 0.7555,
      "step": 2130
    },
    {
      "epoch": 1.0935104752171692,
      "grad_norm": 0.18406588696688597,
      "learning_rate": 5e-06,
      "loss": 0.7433,
      "step": 2140
    },
    {
      "epoch": 1.0986203372508943,
      "grad_norm": 0.17921040061727433,
      "learning_rate": 5e-06,
      "loss": 0.7305,
      "step": 2150
    },
    {
      "epoch": 1.1037301992846194,
      "grad_norm": 0.18963146030246644,
      "learning_rate": 5e-06,
      "loss": 0.7397,
      "step": 2160
    },
    {
      "epoch": 1.1088400613183444,
      "grad_norm": 0.18712686418913257,
      "learning_rate": 5e-06,
      "loss": 0.7409,
      "step": 2170
    },
    {
      "epoch": 1.1139499233520695,
      "grad_norm": 0.19771359798461643,
      "learning_rate": 5e-06,
      "loss": 0.7427,
      "step": 2180
    },
    {
      "epoch": 1.1190597853857946,
      "grad_norm": 0.1840406013875161,
      "learning_rate": 5e-06,
      "loss": 0.7611,
      "step": 2190
    },
    {
      "epoch": 1.1241696474195197,
      "grad_norm": 0.18827676786628125,
      "learning_rate": 5e-06,
      "loss": 0.7364,
      "step": 2200
    },
    {
      "epoch": 1.1292795094532448,
      "grad_norm": 0.191153055094231,
      "learning_rate": 5e-06,
      "loss": 0.7539,
      "step": 2210
    },
    {
      "epoch": 1.13438937148697,
      "grad_norm": 0.1822666967889064,
      "learning_rate": 5e-06,
      "loss": 0.7414,
      "step": 2220
    },
    {
      "epoch": 1.139499233520695,
      "grad_norm": 0.19699245088881503,
      "learning_rate": 5e-06,
      "loss": 0.738,
      "step": 2230
    },
    {
      "epoch": 1.14460909555442,
      "grad_norm": 0.1898722587971926,
      "learning_rate": 5e-06,
      "loss": 0.7461,
      "step": 2240
    },
    {
      "epoch": 1.1497189575881452,
      "grad_norm": 0.18927249674651098,
      "learning_rate": 5e-06,
      "loss": 0.7429,
      "step": 2250
    },
    {
      "epoch": 1.1548288196218701,
      "grad_norm": 0.18896514284627008,
      "learning_rate": 5e-06,
      "loss": 0.7549,
      "step": 2260
    },
    {
      "epoch": 1.1599386816555952,
      "grad_norm": 0.18739970646008372,
      "learning_rate": 5e-06,
      "loss": 0.7499,
      "step": 2270
    },
    {
      "epoch": 1.1650485436893203,
      "grad_norm": 0.1910100009129843,
      "learning_rate": 5e-06,
      "loss": 0.747,
      "step": 2280
    },
    {
      "epoch": 1.1701584057230454,
      "grad_norm": 0.20198153170551428,
      "learning_rate": 5e-06,
      "loss": 0.7386,
      "step": 2290
    },
    {
      "epoch": 1.1752682677567705,
      "grad_norm": 0.18720641288978465,
      "learning_rate": 5e-06,
      "loss": 0.7578,
      "step": 2300
    },
    {
      "epoch": 1.1803781297904956,
      "grad_norm": 0.18961987449195758,
      "learning_rate": 5e-06,
      "loss": 0.7529,
      "step": 2310
    },
    {
      "epoch": 1.1854879918242207,
      "grad_norm": 0.17712198248177036,
      "learning_rate": 5e-06,
      "loss": 0.7476,
      "step": 2320
    },
    {
      "epoch": 1.1905978538579458,
      "grad_norm": 0.18490722732969878,
      "learning_rate": 5e-06,
      "loss": 0.7437,
      "step": 2330
    },
    {
      "epoch": 1.195707715891671,
      "grad_norm": 0.18822524406653396,
      "learning_rate": 5e-06,
      "loss": 0.7469,
      "step": 2340
    },
    {
      "epoch": 1.200817577925396,
      "grad_norm": 0.17979619340691932,
      "learning_rate": 5e-06,
      "loss": 0.7476,
      "step": 2350
    },
    {
      "epoch": 1.205927439959121,
      "grad_norm": 0.19302201862857232,
      "learning_rate": 5e-06,
      "loss": 0.7519,
      "step": 2360
    },
    {
      "epoch": 1.2110373019928462,
      "grad_norm": 0.17331596795239365,
      "learning_rate": 5e-06,
      "loss": 0.7421,
      "step": 2370
    },
    {
      "epoch": 1.2161471640265713,
      "grad_norm": 0.1900974772777454,
      "learning_rate": 5e-06,
      "loss": 0.7491,
      "step": 2380
    },
    {
      "epoch": 1.2212570260602964,
      "grad_norm": 0.18235079887869074,
      "learning_rate": 5e-06,
      "loss": 0.7398,
      "step": 2390
    },
    {
      "epoch": 1.2263668880940215,
      "grad_norm": 0.1990024061618005,
      "learning_rate": 5e-06,
      "loss": 0.7367,
      "step": 2400
    },
    {
      "epoch": 1.2314767501277466,
      "grad_norm": 0.19774880108315787,
      "learning_rate": 5e-06,
      "loss": 0.7509,
      "step": 2410
    },
    {
      "epoch": 1.2365866121614717,
      "grad_norm": 0.18038613594979708,
      "learning_rate": 5e-06,
      "loss": 0.7397,
      "step": 2420
    },
    {
      "epoch": 1.2416964741951968,
      "grad_norm": 0.19148490320343095,
      "learning_rate": 5e-06,
      "loss": 0.738,
      "step": 2430
    },
    {
      "epoch": 1.246806336228922,
      "grad_norm": 0.1764579818726389,
      "learning_rate": 5e-06,
      "loss": 0.7344,
      "step": 2440
    },
    {
      "epoch": 1.2519161982626468,
      "grad_norm": 0.19292171667184566,
      "learning_rate": 5e-06,
      "loss": 0.7386,
      "step": 2450
    },
    {
      "epoch": 1.257026060296372,
      "grad_norm": 0.1830054013064937,
      "learning_rate": 5e-06,
      "loss": 0.7428,
      "step": 2460
    },
    {
      "epoch": 1.262135922330097,
      "grad_norm": 0.1771188524320852,
      "learning_rate": 5e-06,
      "loss": 0.7349,
      "step": 2470
    },
    {
      "epoch": 1.2672457843638223,
      "grad_norm": 0.19308200340380383,
      "learning_rate": 5e-06,
      "loss": 0.7428,
      "step": 2480
    },
    {
      "epoch": 1.2723556463975472,
      "grad_norm": 0.17723395601019812,
      "learning_rate": 5e-06,
      "loss": 0.7271,
      "step": 2490
    },
    {
      "epoch": 1.2774655084312725,
      "grad_norm": 0.21808963432632347,
      "learning_rate": 5e-06,
      "loss": 0.7551,
      "step": 2500
    },
    {
      "epoch": 1.2825753704649974,
      "grad_norm": 0.1944324454402299,
      "learning_rate": 5e-06,
      "loss": 0.7388,
      "step": 2510
    },
    {
      "epoch": 1.2876852324987225,
      "grad_norm": 0.1737167892046253,
      "learning_rate": 5e-06,
      "loss": 0.7383,
      "step": 2520
    },
    {
      "epoch": 1.2927950945324476,
      "grad_norm": 0.19316243219818216,
      "learning_rate": 5e-06,
      "loss": 0.7447,
      "step": 2530
    },
    {
      "epoch": 1.2979049565661727,
      "grad_norm": 0.18271045253382115,
      "learning_rate": 5e-06,
      "loss": 0.7439,
      "step": 2540
    },
    {
      "epoch": 1.3030148185998978,
      "grad_norm": 0.19235060247622612,
      "learning_rate": 5e-06,
      "loss": 0.7362,
      "step": 2550
    },
    {
      "epoch": 1.3081246806336229,
      "grad_norm": 0.17975422534798727,
      "learning_rate": 5e-06,
      "loss": 0.7355,
      "step": 2560
    },
    {
      "epoch": 1.313234542667348,
      "grad_norm": 0.19133431284185412,
      "learning_rate": 5e-06,
      "loss": 0.7474,
      "step": 2570
    },
    {
      "epoch": 1.318344404701073,
      "grad_norm": 0.18525107476229646,
      "learning_rate": 5e-06,
      "loss": 0.7362,
      "step": 2580
    },
    {
      "epoch": 1.3234542667347982,
      "grad_norm": 0.17940598778920888,
      "learning_rate": 5e-06,
      "loss": 0.7456,
      "step": 2590
    },
    {
      "epoch": 1.3285641287685233,
      "grad_norm": 0.18377019808836909,
      "learning_rate": 5e-06,
      "loss": 0.7371,
      "step": 2600
    },
    {
      "epoch": 1.3336739908022484,
      "grad_norm": 0.19380006248186266,
      "learning_rate": 5e-06,
      "loss": 0.7382,
      "step": 2610
    },
    {
      "epoch": 1.3387838528359735,
      "grad_norm": 0.19130341477919996,
      "learning_rate": 5e-06,
      "loss": 0.7249,
      "step": 2620
    },
    {
      "epoch": 1.3438937148696986,
      "grad_norm": 0.19399937386692442,
      "learning_rate": 5e-06,
      "loss": 0.7461,
      "step": 2630
    },
    {
      "epoch": 1.3490035769034237,
      "grad_norm": 0.19026606995853784,
      "learning_rate": 5e-06,
      "loss": 0.744,
      "step": 2640
    },
    {
      "epoch": 1.3541134389371488,
      "grad_norm": 0.17865393066414276,
      "learning_rate": 5e-06,
      "loss": 0.7562,
      "step": 2650
    },
    {
      "epoch": 1.3592233009708738,
      "grad_norm": 0.18268806699965215,
      "learning_rate": 5e-06,
      "loss": 0.7369,
      "step": 2660
    },
    {
      "epoch": 1.364333163004599,
      "grad_norm": 0.1859894512893362,
      "learning_rate": 5e-06,
      "loss": 0.7303,
      "step": 2670
    },
    {
      "epoch": 1.369443025038324,
      "grad_norm": 0.1858579729895718,
      "learning_rate": 5e-06,
      "loss": 0.7338,
      "step": 2680
    },
    {
      "epoch": 1.3745528870720491,
      "grad_norm": 0.18844334675064925,
      "learning_rate": 5e-06,
      "loss": 0.7399,
      "step": 2690
    },
    {
      "epoch": 1.379662749105774,
      "grad_norm": 0.19488780104365555,
      "learning_rate": 5e-06,
      "loss": 0.7297,
      "step": 2700
    },
    {
      "epoch": 1.3847726111394993,
      "grad_norm": 0.2012609184785339,
      "learning_rate": 5e-06,
      "loss": 0.7438,
      "step": 2710
    },
    {
      "epoch": 1.3898824731732242,
      "grad_norm": 0.19240664121181153,
      "learning_rate": 5e-06,
      "loss": 0.724,
      "step": 2720
    },
    {
      "epoch": 1.3949923352069495,
      "grad_norm": 0.1989866337354731,
      "learning_rate": 5e-06,
      "loss": 0.7387,
      "step": 2730
    },
    {
      "epoch": 1.4001021972406744,
      "grad_norm": 0.19173593318229185,
      "learning_rate": 5e-06,
      "loss": 0.7389,
      "step": 2740
    },
    {
      "epoch": 1.4052120592743995,
      "grad_norm": 0.18490218706957628,
      "learning_rate": 5e-06,
      "loss": 0.7354,
      "step": 2750
    },
    {
      "epoch": 1.4103219213081246,
      "grad_norm": 0.19596927441115194,
      "learning_rate": 5e-06,
      "loss": 0.7294,
      "step": 2760
    },
    {
      "epoch": 1.4154317833418497,
      "grad_norm": 0.1884149575008095,
      "learning_rate": 5e-06,
      "loss": 0.7535,
      "step": 2770
    },
    {
      "epoch": 1.4205416453755748,
      "grad_norm": 0.19011351454021505,
      "learning_rate": 5e-06,
      "loss": 0.7311,
      "step": 2780
    },
    {
      "epoch": 1.4256515074093,
      "grad_norm": 0.18330086878582655,
      "learning_rate": 5e-06,
      "loss": 0.7238,
      "step": 2790
    },
    {
      "epoch": 1.430761369443025,
      "grad_norm": 0.18245127655957494,
      "learning_rate": 5e-06,
      "loss": 0.7297,
      "step": 2800
    },
    {
      "epoch": 1.43587123147675,
      "grad_norm": 0.1899683440493854,
      "learning_rate": 5e-06,
      "loss": 0.7388,
      "step": 2810
    },
    {
      "epoch": 1.4409810935104752,
      "grad_norm": 0.17955923753560576,
      "learning_rate": 5e-06,
      "loss": 0.7382,
      "step": 2820
    },
    {
      "epoch": 1.4460909555442003,
      "grad_norm": 0.19252118964036657,
      "learning_rate": 5e-06,
      "loss": 0.7428,
      "step": 2830
    },
    {
      "epoch": 1.4512008175779254,
      "grad_norm": 0.1993095814580447,
      "learning_rate": 5e-06,
      "loss": 0.7453,
      "step": 2840
    },
    {
      "epoch": 1.4563106796116505,
      "grad_norm": 0.19453340379561043,
      "learning_rate": 5e-06,
      "loss": 0.7316,
      "step": 2850
    },
    {
      "epoch": 1.4614205416453756,
      "grad_norm": 0.18991487065634022,
      "learning_rate": 5e-06,
      "loss": 0.7386,
      "step": 2860
    },
    {
      "epoch": 1.4665304036791007,
      "grad_norm": 0.18973483396294438,
      "learning_rate": 5e-06,
      "loss": 0.7466,
      "step": 2870
    },
    {
      "epoch": 1.4716402657128258,
      "grad_norm": 0.1919078737205217,
      "learning_rate": 5e-06,
      "loss": 0.7495,
      "step": 2880
    },
    {
      "epoch": 1.476750127746551,
      "grad_norm": 0.1839372662925669,
      "learning_rate": 5e-06,
      "loss": 0.7389,
      "step": 2890
    },
    {
      "epoch": 1.481859989780276,
      "grad_norm": 0.19249510537236922,
      "learning_rate": 5e-06,
      "loss": 0.749,
      "step": 2900
    },
    {
      "epoch": 1.486969851814001,
      "grad_norm": 0.179246890906481,
      "learning_rate": 5e-06,
      "loss": 0.7282,
      "step": 2910
    },
    {
      "epoch": 1.4920797138477262,
      "grad_norm": 0.19514232006253573,
      "learning_rate": 5e-06,
      "loss": 0.7446,
      "step": 2920
    },
    {
      "epoch": 1.497189575881451,
      "grad_norm": 0.18596165643891152,
      "learning_rate": 5e-06,
      "loss": 0.7334,
      "step": 2930
    },
    {
      "epoch": 1.5022994379151764,
      "grad_norm": 0.18786595484474303,
      "learning_rate": 5e-06,
      "loss": 0.7279,
      "step": 2940
    },
    {
      "epoch": 1.5074092999489013,
      "grad_norm": 0.1971738538736254,
      "learning_rate": 5e-06,
      "loss": 0.7461,
      "step": 2950
    },
    {
      "epoch": 1.5125191619826266,
      "grad_norm": 0.18866040282412957,
      "learning_rate": 5e-06,
      "loss": 0.7474,
      "step": 2960
    },
    {
      "epoch": 1.5176290240163515,
      "grad_norm": 0.19040908132898832,
      "learning_rate": 5e-06,
      "loss": 0.7445,
      "step": 2970
    },
    {
      "epoch": 1.5227388860500768,
      "grad_norm": 0.1929810400978154,
      "learning_rate": 5e-06,
      "loss": 0.7362,
      "step": 2980
    },
    {
      "epoch": 1.5278487480838017,
      "grad_norm": 0.19442375737357984,
      "learning_rate": 5e-06,
      "loss": 0.7305,
      "step": 2990
    },
    {
      "epoch": 1.532958610117527,
      "grad_norm": 0.18546826858387847,
      "learning_rate": 5e-06,
      "loss": 0.7311,
      "step": 3000
    },
    {
      "epoch": 1.5380684721512519,
      "grad_norm": 0.18542784404927515,
      "learning_rate": 5e-06,
      "loss": 0.7447,
      "step": 3010
    },
    {
      "epoch": 1.543178334184977,
      "grad_norm": 0.2020846723209545,
      "learning_rate": 5e-06,
      "loss": 0.7113,
      "step": 3020
    },
    {
      "epoch": 1.548288196218702,
      "grad_norm": 0.19026884893782828,
      "learning_rate": 5e-06,
      "loss": 0.7157,
      "step": 3030
    },
    {
      "epoch": 1.5533980582524272,
      "grad_norm": 0.18111421662938304,
      "learning_rate": 5e-06,
      "loss": 0.7323,
      "step": 3040
    },
    {
      "epoch": 1.5585079202861523,
      "grad_norm": 0.19367385202342016,
      "learning_rate": 5e-06,
      "loss": 0.7305,
      "step": 3050
    },
    {
      "epoch": 1.5636177823198774,
      "grad_norm": 0.18590394121821466,
      "learning_rate": 5e-06,
      "loss": 0.7341,
      "step": 3060
    },
    {
      "epoch": 1.5687276443536025,
      "grad_norm": 0.18488441186992707,
      "learning_rate": 5e-06,
      "loss": 0.7482,
      "step": 3070
    },
    {
      "epoch": 1.5738375063873276,
      "grad_norm": 0.18226306867076514,
      "learning_rate": 5e-06,
      "loss": 0.7334,
      "step": 3080
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.20053856155426641,
      "learning_rate": 5e-06,
      "loss": 0.7414,
      "step": 3090
    },
    {
      "epoch": 1.5840572304547778,
      "grad_norm": 0.19672564131420983,
      "learning_rate": 5e-06,
      "loss": 0.7508,
      "step": 3100
    },
    {
      "epoch": 1.5891670924885029,
      "grad_norm": 0.1790851772225089,
      "learning_rate": 5e-06,
      "loss": 0.7471,
      "step": 3110
    },
    {
      "epoch": 1.594276954522228,
      "grad_norm": 0.1900047612676954,
      "learning_rate": 5e-06,
      "loss": 0.7421,
      "step": 3120
    },
    {
      "epoch": 1.599386816555953,
      "grad_norm": 0.19746465955340986,
      "learning_rate": 5e-06,
      "loss": 0.7471,
      "step": 3130
    },
    {
      "epoch": 1.604496678589678,
      "grad_norm": 0.186549540683221,
      "learning_rate": 5e-06,
      "loss": 0.7275,
      "step": 3140
    },
    {
      "epoch": 1.6096065406234032,
      "grad_norm": 0.1876261054598287,
      "learning_rate": 5e-06,
      "loss": 0.7359,
      "step": 3150
    },
    {
      "epoch": 1.6147164026571281,
      "grad_norm": 0.19082325492370317,
      "learning_rate": 5e-06,
      "loss": 0.7268,
      "step": 3160
    },
    {
      "epoch": 1.6198262646908534,
      "grad_norm": 0.2016402119888201,
      "learning_rate": 5e-06,
      "loss": 0.7377,
      "step": 3170
    },
    {
      "epoch": 1.6249361267245783,
      "grad_norm": 0.1888126317070555,
      "learning_rate": 5e-06,
      "loss": 0.7265,
      "step": 3180
    },
    {
      "epoch": 1.6300459887583036,
      "grad_norm": 0.17743730583327474,
      "learning_rate": 5e-06,
      "loss": 0.7203,
      "step": 3190
    },
    {
      "epoch": 1.6351558507920285,
      "grad_norm": 0.1826162903853255,
      "learning_rate": 5e-06,
      "loss": 0.7215,
      "step": 3200
    },
    {
      "epoch": 1.6402657128257538,
      "grad_norm": 0.19419266754404552,
      "learning_rate": 5e-06,
      "loss": 0.7376,
      "step": 3210
    },
    {
      "epoch": 1.6453755748594787,
      "grad_norm": 0.1956453565355767,
      "learning_rate": 5e-06,
      "loss": 0.7316,
      "step": 3220
    },
    {
      "epoch": 1.650485436893204,
      "grad_norm": 0.19765143129125318,
      "learning_rate": 5e-06,
      "loss": 0.7374,
      "step": 3230
    },
    {
      "epoch": 1.655595298926929,
      "grad_norm": 0.19127982430051427,
      "learning_rate": 5e-06,
      "loss": 0.7405,
      "step": 3240
    },
    {
      "epoch": 1.660705160960654,
      "grad_norm": 0.1847472801458583,
      "learning_rate": 5e-06,
      "loss": 0.724,
      "step": 3250
    },
    {
      "epoch": 1.6658150229943791,
      "grad_norm": 0.18698307703261788,
      "learning_rate": 5e-06,
      "loss": 0.7374,
      "step": 3260
    },
    {
      "epoch": 1.6709248850281042,
      "grad_norm": 0.17533798523255767,
      "learning_rate": 5e-06,
      "loss": 0.7202,
      "step": 3270
    },
    {
      "epoch": 1.6760347470618293,
      "grad_norm": 0.1806351825859557,
      "learning_rate": 5e-06,
      "loss": 0.7452,
      "step": 3280
    },
    {
      "epoch": 1.6811446090955544,
      "grad_norm": 0.1767976805961292,
      "learning_rate": 5e-06,
      "loss": 0.7338,
      "step": 3290
    },
    {
      "epoch": 1.6862544711292795,
      "grad_norm": 0.19498984484111873,
      "learning_rate": 5e-06,
      "loss": 0.7403,
      "step": 3300
    },
    {
      "epoch": 1.6913643331630046,
      "grad_norm": 0.17701307669892918,
      "learning_rate": 5e-06,
      "loss": 0.7299,
      "step": 3310
    },
    {
      "epoch": 1.6964741951967297,
      "grad_norm": 0.19220216566407472,
      "learning_rate": 5e-06,
      "loss": 0.7314,
      "step": 3320
    },
    {
      "epoch": 1.7015840572304548,
      "grad_norm": 0.1829279730231264,
      "learning_rate": 5e-06,
      "loss": 0.743,
      "step": 3330
    },
    {
      "epoch": 1.70669391926418,
      "grad_norm": 0.19526766653061225,
      "learning_rate": 5e-06,
      "loss": 0.7222,
      "step": 3340
    },
    {
      "epoch": 1.711803781297905,
      "grad_norm": 0.19455609962672274,
      "learning_rate": 5e-06,
      "loss": 0.7253,
      "step": 3350
    },
    {
      "epoch": 1.71691364333163,
      "grad_norm": 0.21002379536162816,
      "learning_rate": 5e-06,
      "loss": 0.7429,
      "step": 3360
    },
    {
      "epoch": 1.722023505365355,
      "grad_norm": 0.1990882316461353,
      "learning_rate": 5e-06,
      "loss": 0.7443,
      "step": 3370
    },
    {
      "epoch": 1.7271333673990803,
      "grad_norm": 0.17934672167038826,
      "learning_rate": 5e-06,
      "loss": 0.7497,
      "step": 3380
    },
    {
      "epoch": 1.7322432294328052,
      "grad_norm": 0.19501165745940965,
      "learning_rate": 5e-06,
      "loss": 0.7425,
      "step": 3390
    },
    {
      "epoch": 1.7373530914665305,
      "grad_norm": 0.19248650606756543,
      "learning_rate": 5e-06,
      "loss": 0.7297,
      "step": 3400
    },
    {
      "epoch": 1.7424629535002554,
      "grad_norm": 0.17721599710417338,
      "learning_rate": 5e-06,
      "loss": 0.7251,
      "step": 3410
    },
    {
      "epoch": 1.7475728155339807,
      "grad_norm": 0.18509365156353424,
      "learning_rate": 5e-06,
      "loss": 0.7221,
      "step": 3420
    },
    {
      "epoch": 1.7526826775677056,
      "grad_norm": 0.18289284691122754,
      "learning_rate": 5e-06,
      "loss": 0.7327,
      "step": 3430
    },
    {
      "epoch": 1.757792539601431,
      "grad_norm": 0.18756279151165123,
      "learning_rate": 5e-06,
      "loss": 0.732,
      "step": 3440
    },
    {
      "epoch": 1.7629024016351558,
      "grad_norm": 0.17439303769229625,
      "learning_rate": 5e-06,
      "loss": 0.7344,
      "step": 3450
    },
    {
      "epoch": 1.768012263668881,
      "grad_norm": 0.17783376482824478,
      "learning_rate": 5e-06,
      "loss": 0.7323,
      "step": 3460
    },
    {
      "epoch": 1.773122125702606,
      "grad_norm": 0.19448194078586717,
      "learning_rate": 5e-06,
      "loss": 0.7292,
      "step": 3470
    },
    {
      "epoch": 1.778231987736331,
      "grad_norm": 0.18000237860427712,
      "learning_rate": 5e-06,
      "loss": 0.7219,
      "step": 3480
    },
    {
      "epoch": 1.7833418497700562,
      "grad_norm": 0.18519882940772772,
      "learning_rate": 5e-06,
      "loss": 0.7354,
      "step": 3490
    },
    {
      "epoch": 1.7884517118037813,
      "grad_norm": 0.19301292549147336,
      "learning_rate": 5e-06,
      "loss": 0.7487,
      "step": 3500
    },
    {
      "epoch": 1.7935615738375064,
      "grad_norm": 0.17758380897102066,
      "learning_rate": 5e-06,
      "loss": 0.7373,
      "step": 3510
    },
    {
      "epoch": 1.7986714358712315,
      "grad_norm": 0.1794720757802905,
      "learning_rate": 5e-06,
      "loss": 0.7226,
      "step": 3520
    },
    {
      "epoch": 1.8037812979049566,
      "grad_norm": 0.18100374933694008,
      "learning_rate": 5e-06,
      "loss": 0.7256,
      "step": 3530
    },
    {
      "epoch": 1.8088911599386817,
      "grad_norm": 0.1954603145633284,
      "learning_rate": 5e-06,
      "loss": 0.7283,
      "step": 3540
    },
    {
      "epoch": 1.8140010219724068,
      "grad_norm": 0.19558607958635285,
      "learning_rate": 5e-06,
      "loss": 0.7384,
      "step": 3550
    },
    {
      "epoch": 1.8191108840061319,
      "grad_norm": 0.1772107537935853,
      "learning_rate": 5e-06,
      "loss": 0.7382,
      "step": 3560
    },
    {
      "epoch": 1.824220746039857,
      "grad_norm": 0.17916901000763397,
      "learning_rate": 5e-06,
      "loss": 0.7367,
      "step": 3570
    },
    {
      "epoch": 1.829330608073582,
      "grad_norm": 0.19083482072843658,
      "learning_rate": 5e-06,
      "loss": 0.7247,
      "step": 3580
    },
    {
      "epoch": 1.8344404701073072,
      "grad_norm": 0.1770449813805881,
      "learning_rate": 5e-06,
      "loss": 0.7187,
      "step": 3590
    },
    {
      "epoch": 1.839550332141032,
      "grad_norm": 0.18790158384523442,
      "learning_rate": 5e-06,
      "loss": 0.7403,
      "step": 3600
    },
    {
      "epoch": 1.8446601941747574,
      "grad_norm": 0.17892362311216914,
      "learning_rate": 5e-06,
      "loss": 0.7368,
      "step": 3610
    },
    {
      "epoch": 1.8497700562084822,
      "grad_norm": 0.1839659251785667,
      "learning_rate": 5e-06,
      "loss": 0.7255,
      "step": 3620
    },
    {
      "epoch": 1.8548799182422075,
      "grad_norm": 0.19138496555502849,
      "learning_rate": 5e-06,
      "loss": 0.7453,
      "step": 3630
    },
    {
      "epoch": 1.8599897802759324,
      "grad_norm": 0.18135734354491537,
      "learning_rate": 5e-06,
      "loss": 0.7166,
      "step": 3640
    },
    {
      "epoch": 1.8650996423096577,
      "grad_norm": 0.1859082044449026,
      "learning_rate": 5e-06,
      "loss": 0.7285,
      "step": 3650
    },
    {
      "epoch": 1.8702095043433826,
      "grad_norm": 0.1913280855307758,
      "learning_rate": 5e-06,
      "loss": 0.7279,
      "step": 3660
    },
    {
      "epoch": 1.875319366377108,
      "grad_norm": 0.19148047998384163,
      "learning_rate": 5e-06,
      "loss": 0.7381,
      "step": 3670
    },
    {
      "epoch": 1.8804292284108328,
      "grad_norm": 0.190776629149848,
      "learning_rate": 5e-06,
      "loss": 0.7347,
      "step": 3680
    },
    {
      "epoch": 1.8855390904445581,
      "grad_norm": 0.18748890464701637,
      "learning_rate": 5e-06,
      "loss": 0.7214,
      "step": 3690
    },
    {
      "epoch": 1.890648952478283,
      "grad_norm": 0.19691029617370956,
      "learning_rate": 5e-06,
      "loss": 0.7396,
      "step": 3700
    },
    {
      "epoch": 1.8957588145120083,
      "grad_norm": 0.17143385370457725,
      "learning_rate": 5e-06,
      "loss": 0.7305,
      "step": 3710
    },
    {
      "epoch": 1.9008686765457332,
      "grad_norm": 0.19115494111352774,
      "learning_rate": 5e-06,
      "loss": 0.7252,
      "step": 3720
    },
    {
      "epoch": 1.9059785385794583,
      "grad_norm": 0.18589155104150923,
      "learning_rate": 5e-06,
      "loss": 0.728,
      "step": 3730
    },
    {
      "epoch": 1.9110884006131834,
      "grad_norm": 0.1870078279938856,
      "learning_rate": 5e-06,
      "loss": 0.7259,
      "step": 3740
    },
    {
      "epoch": 1.9161982626469085,
      "grad_norm": 0.17909224396912188,
      "learning_rate": 5e-06,
      "loss": 0.7345,
      "step": 3750
    },
    {
      "epoch": 1.9213081246806336,
      "grad_norm": 0.19885632401705697,
      "learning_rate": 5e-06,
      "loss": 0.7289,
      "step": 3760
    },
    {
      "epoch": 1.9264179867143587,
      "grad_norm": 0.18778552722329356,
      "learning_rate": 5e-06,
      "loss": 0.725,
      "step": 3770
    },
    {
      "epoch": 1.9315278487480838,
      "grad_norm": 0.2019790249495369,
      "learning_rate": 5e-06,
      "loss": 0.7398,
      "step": 3780
    },
    {
      "epoch": 1.936637710781809,
      "grad_norm": 0.19338701904495897,
      "learning_rate": 5e-06,
      "loss": 0.7265,
      "step": 3790
    },
    {
      "epoch": 1.941747572815534,
      "grad_norm": 0.18703296264974872,
      "learning_rate": 5e-06,
      "loss": 0.7332,
      "step": 3800
    },
    {
      "epoch": 1.946857434849259,
      "grad_norm": 0.1700175440342506,
      "learning_rate": 5e-06,
      "loss": 0.7205,
      "step": 3810
    },
    {
      "epoch": 1.9519672968829842,
      "grad_norm": 0.18636496202992153,
      "learning_rate": 5e-06,
      "loss": 0.7154,
      "step": 3820
    },
    {
      "epoch": 1.9570771589167093,
      "grad_norm": 0.1826391337083993,
      "learning_rate": 5e-06,
      "loss": 0.7356,
      "step": 3830
    },
    {
      "epoch": 1.9621870209504344,
      "grad_norm": 0.17766191115765154,
      "learning_rate": 5e-06,
      "loss": 0.7369,
      "step": 3840
    },
    {
      "epoch": 1.9672968829841593,
      "grad_norm": 0.18034528150782342,
      "learning_rate": 5e-06,
      "loss": 0.7288,
      "step": 3850
    },
    {
      "epoch": 1.9724067450178846,
      "grad_norm": 0.18541894497456152,
      "learning_rate": 5e-06,
      "loss": 0.7296,
      "step": 3860
    },
    {
      "epoch": 1.9775166070516095,
      "grad_norm": 0.19539634425789987,
      "learning_rate": 5e-06,
      "loss": 0.7374,
      "step": 3870
    },
    {
      "epoch": 1.9826264690853348,
      "grad_norm": 0.1928837424204438,
      "learning_rate": 5e-06,
      "loss": 0.732,
      "step": 3880
    },
    {
      "epoch": 1.9877363311190597,
      "grad_norm": 0.18813671735265705,
      "learning_rate": 5e-06,
      "loss": 0.7285,
      "step": 3890
    },
    {
      "epoch": 1.992846193152785,
      "grad_norm": 0.19024591983517306,
      "learning_rate": 5e-06,
      "loss": 0.7386,
      "step": 3900
    },
    {
      "epoch": 1.9979560551865099,
      "grad_norm": 0.1791330764130833,
      "learning_rate": 5e-06,
      "loss": 0.7324,
      "step": 3910
    },
    {
      "epoch": 2.003065917220235,
      "grad_norm": 0.18457472513280188,
      "learning_rate": 5e-06,
      "loss": 0.7303,
      "step": 3920
    },
    {
      "epoch": 2.00817577925396,
      "grad_norm": 0.18684788166920566,
      "learning_rate": 5e-06,
      "loss": 0.735,
      "step": 3930
    },
    {
      "epoch": 2.0132856412876854,
      "grad_norm": 0.18385152341485855,
      "learning_rate": 5e-06,
      "loss": 0.7204,
      "step": 3940
    },
    {
      "epoch": 2.0183955033214103,
      "grad_norm": 0.1893087134675762,
      "learning_rate": 5e-06,
      "loss": 0.7271,
      "step": 3950
    },
    {
      "epoch": 2.0235053653551356,
      "grad_norm": 0.19001500071497598,
      "learning_rate": 5e-06,
      "loss": 0.7286,
      "step": 3960
    },
    {
      "epoch": 2.0286152273888605,
      "grad_norm": 0.18883227887144083,
      "learning_rate": 5e-06,
      "loss": 0.7157,
      "step": 3970
    },
    {
      "epoch": 2.033725089422586,
      "grad_norm": 0.17689498199700174,
      "learning_rate": 5e-06,
      "loss": 0.7063,
      "step": 3980
    },
    {
      "epoch": 2.0388349514563107,
      "grad_norm": 0.19270059169949594,
      "learning_rate": 5e-06,
      "loss": 0.7295,
      "step": 3990
    },
    {
      "epoch": 2.043944813490036,
      "grad_norm": 0.19597358123850073,
      "learning_rate": 5e-06,
      "loss": 0.7268,
      "step": 4000
    },
    {
      "epoch": 2.049054675523761,
      "grad_norm": 0.19711968415004932,
      "learning_rate": 5e-06,
      "loss": 0.7229,
      "step": 4010
    },
    {
      "epoch": 2.0541645375574857,
      "grad_norm": 0.18776496261358783,
      "learning_rate": 5e-06,
      "loss": 0.7158,
      "step": 4020
    },
    {
      "epoch": 2.059274399591211,
      "grad_norm": 0.18818737294591004,
      "learning_rate": 5e-06,
      "loss": 0.7194,
      "step": 4030
    },
    {
      "epoch": 2.064384261624936,
      "grad_norm": 0.19845798280803176,
      "learning_rate": 5e-06,
      "loss": 0.7265,
      "step": 4040
    },
    {
      "epoch": 2.0694941236586613,
      "grad_norm": 0.18713228958457867,
      "learning_rate": 5e-06,
      "loss": 0.7095,
      "step": 4050
    },
    {
      "epoch": 2.074603985692386,
      "grad_norm": 0.17787870833728897,
      "learning_rate": 5e-06,
      "loss": 0.7149,
      "step": 4060
    },
    {
      "epoch": 2.0797138477261115,
      "grad_norm": 0.19472810880013228,
      "learning_rate": 5e-06,
      "loss": 0.7178,
      "step": 4070
    },
    {
      "epoch": 2.0848237097598363,
      "grad_norm": 0.19429450371850024,
      "learning_rate": 5e-06,
      "loss": 0.7118,
      "step": 4080
    },
    {
      "epoch": 2.0899335717935617,
      "grad_norm": 0.1941609760118733,
      "learning_rate": 5e-06,
      "loss": 0.7265,
      "step": 4090
    },
    {
      "epoch": 2.0950434338272865,
      "grad_norm": 0.19290976310635458,
      "learning_rate": 5e-06,
      "loss": 0.7085,
      "step": 4100
    },
    {
      "epoch": 2.100153295861012,
      "grad_norm": 0.1765241017205207,
      "learning_rate": 5e-06,
      "loss": 0.7301,
      "step": 4110
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.17846133756954982,
      "learning_rate": 5e-06,
      "loss": 0.7105,
      "step": 4120
    },
    {
      "epoch": 2.110373019928462,
      "grad_norm": 0.1990655813201847,
      "learning_rate": 5e-06,
      "loss": 0.7197,
      "step": 4130
    },
    {
      "epoch": 2.115482881962187,
      "grad_norm": 0.18227727070573727,
      "learning_rate": 5e-06,
      "loss": 0.7225,
      "step": 4140
    },
    {
      "epoch": 2.1205927439959122,
      "grad_norm": 0.1950861579503913,
      "learning_rate": 5e-06,
      "loss": 0.7151,
      "step": 4150
    },
    {
      "epoch": 2.125702606029637,
      "grad_norm": 0.1852273630173904,
      "learning_rate": 5e-06,
      "loss": 0.7349,
      "step": 4160
    },
    {
      "epoch": 2.1308124680633624,
      "grad_norm": 0.19463032921601597,
      "learning_rate": 5e-06,
      "loss": 0.7209,
      "step": 4170
    },
    {
      "epoch": 2.1359223300970873,
      "grad_norm": 0.20049767813158514,
      "learning_rate": 5e-06,
      "loss": 0.7272,
      "step": 4180
    },
    {
      "epoch": 2.1410321921308126,
      "grad_norm": 0.1964989410548727,
      "learning_rate": 5e-06,
      "loss": 0.7205,
      "step": 4190
    },
    {
      "epoch": 2.1461420541645375,
      "grad_norm": 0.18656787454813695,
      "learning_rate": 5e-06,
      "loss": 0.73,
      "step": 4200
    },
    {
      "epoch": 2.151251916198263,
      "grad_norm": 0.1869811709706087,
      "learning_rate": 5e-06,
      "loss": 0.7211,
      "step": 4210
    },
    {
      "epoch": 2.1563617782319877,
      "grad_norm": 0.1822452717215839,
      "learning_rate": 5e-06,
      "loss": 0.7149,
      "step": 4220
    },
    {
      "epoch": 2.161471640265713,
      "grad_norm": 0.18992052776084958,
      "learning_rate": 5e-06,
      "loss": 0.7343,
      "step": 4230
    },
    {
      "epoch": 2.166581502299438,
      "grad_norm": 0.18438525645647527,
      "learning_rate": 5e-06,
      "loss": 0.7094,
      "step": 4240
    },
    {
      "epoch": 2.171691364333163,
      "grad_norm": 0.19529306952253045,
      "learning_rate": 5e-06,
      "loss": 0.7309,
      "step": 4250
    },
    {
      "epoch": 2.176801226366888,
      "grad_norm": 0.19540047885297857,
      "learning_rate": 5e-06,
      "loss": 0.7128,
      "step": 4260
    },
    {
      "epoch": 2.181911088400613,
      "grad_norm": 0.18411641667621176,
      "learning_rate": 5e-06,
      "loss": 0.7225,
      "step": 4270
    },
    {
      "epoch": 2.1870209504343383,
      "grad_norm": 0.19493918166759708,
      "learning_rate": 5e-06,
      "loss": 0.723,
      "step": 4280
    },
    {
      "epoch": 2.192130812468063,
      "grad_norm": 0.1843201705851165,
      "learning_rate": 5e-06,
      "loss": 0.7133,
      "step": 4290
    },
    {
      "epoch": 2.1972406745017885,
      "grad_norm": 0.1945434906423164,
      "learning_rate": 5e-06,
      "loss": 0.7196,
      "step": 4300
    },
    {
      "epoch": 2.2023505365355134,
      "grad_norm": 0.19008929161664118,
      "learning_rate": 5e-06,
      "loss": 0.7224,
      "step": 4310
    },
    {
      "epoch": 2.2074603985692387,
      "grad_norm": 0.18391558057092405,
      "learning_rate": 5e-06,
      "loss": 0.723,
      "step": 4320
    },
    {
      "epoch": 2.2125702606029636,
      "grad_norm": 0.19348061164149957,
      "learning_rate": 5e-06,
      "loss": 0.7331,
      "step": 4330
    },
    {
      "epoch": 2.217680122636689,
      "grad_norm": 0.18607054173383442,
      "learning_rate": 5e-06,
      "loss": 0.7244,
      "step": 4340
    },
    {
      "epoch": 2.2227899846704138,
      "grad_norm": 0.18618780601208676,
      "learning_rate": 5e-06,
      "loss": 0.7272,
      "step": 4350
    },
    {
      "epoch": 2.227899846704139,
      "grad_norm": 0.17823884880532717,
      "learning_rate": 5e-06,
      "loss": 0.7091,
      "step": 4360
    },
    {
      "epoch": 2.233009708737864,
      "grad_norm": 0.18561422862650115,
      "learning_rate": 5e-06,
      "loss": 0.7205,
      "step": 4370
    },
    {
      "epoch": 2.2381195707715893,
      "grad_norm": 0.17953482905796772,
      "learning_rate": 5e-06,
      "loss": 0.729,
      "step": 4380
    },
    {
      "epoch": 2.243229432805314,
      "grad_norm": 0.18191363924178688,
      "learning_rate": 5e-06,
      "loss": 0.7111,
      "step": 4390
    },
    {
      "epoch": 2.2483392948390395,
      "grad_norm": 0.180435799688224,
      "learning_rate": 5e-06,
      "loss": 0.7107,
      "step": 4400
    },
    {
      "epoch": 2.2534491568727644,
      "grad_norm": 0.18960270587770217,
      "learning_rate": 5e-06,
      "loss": 0.725,
      "step": 4410
    },
    {
      "epoch": 2.2585590189064897,
      "grad_norm": 0.17769968679729267,
      "learning_rate": 5e-06,
      "loss": 0.7143,
      "step": 4420
    },
    {
      "epoch": 2.2636688809402146,
      "grad_norm": 0.19163278997385685,
      "learning_rate": 5e-06,
      "loss": 0.7182,
      "step": 4430
    },
    {
      "epoch": 2.26877874297394,
      "grad_norm": 0.1940788858796441,
      "learning_rate": 5e-06,
      "loss": 0.7366,
      "step": 4440
    },
    {
      "epoch": 2.2738886050076648,
      "grad_norm": 0.1776800674678045,
      "learning_rate": 5e-06,
      "loss": 0.7069,
      "step": 4450
    },
    {
      "epoch": 2.27899846704139,
      "grad_norm": 0.17843289642160187,
      "learning_rate": 5e-06,
      "loss": 0.724,
      "step": 4460
    },
    {
      "epoch": 2.284108329075115,
      "grad_norm": 0.18057741833938729,
      "learning_rate": 5e-06,
      "loss": 0.7137,
      "step": 4470
    },
    {
      "epoch": 2.28921819110884,
      "grad_norm": 0.1866133357180047,
      "learning_rate": 5e-06,
      "loss": 0.7166,
      "step": 4480
    },
    {
      "epoch": 2.294328053142565,
      "grad_norm": 0.18193348825294622,
      "learning_rate": 5e-06,
      "loss": 0.7244,
      "step": 4490
    },
    {
      "epoch": 2.2994379151762905,
      "grad_norm": 0.1929190761683958,
      "learning_rate": 5e-06,
      "loss": 0.7277,
      "step": 4500
    },
    {
      "epoch": 2.3045477772100154,
      "grad_norm": 0.19392177452359835,
      "learning_rate": 5e-06,
      "loss": 0.7198,
      "step": 4510
    },
    {
      "epoch": 2.3096576392437402,
      "grad_norm": 0.18143577291270357,
      "learning_rate": 5e-06,
      "loss": 0.728,
      "step": 4520
    },
    {
      "epoch": 2.3147675012774656,
      "grad_norm": 0.19443872804506757,
      "learning_rate": 5e-06,
      "loss": 0.7272,
      "step": 4530
    },
    {
      "epoch": 2.3198773633111904,
      "grad_norm": 0.1851328945489432,
      "learning_rate": 5e-06,
      "loss": 0.7234,
      "step": 4540
    },
    {
      "epoch": 2.3249872253449158,
      "grad_norm": 0.19038110214278162,
      "learning_rate": 5e-06,
      "loss": 0.713,
      "step": 4550
    },
    {
      "epoch": 2.3300970873786406,
      "grad_norm": 0.1874741106018047,
      "learning_rate": 5e-06,
      "loss": 0.7161,
      "step": 4560
    },
    {
      "epoch": 2.335206949412366,
      "grad_norm": 0.19581977792981697,
      "learning_rate": 5e-06,
      "loss": 0.7162,
      "step": 4570
    },
    {
      "epoch": 2.340316811446091,
      "grad_norm": 0.17541497625191085,
      "learning_rate": 5e-06,
      "loss": 0.723,
      "step": 4580
    },
    {
      "epoch": 2.345426673479816,
      "grad_norm": 0.20023862697490177,
      "learning_rate": 5e-06,
      "loss": 0.7185,
      "step": 4590
    },
    {
      "epoch": 2.350536535513541,
      "grad_norm": 0.1959974413991938,
      "learning_rate": 5e-06,
      "loss": 0.7088,
      "step": 4600
    },
    {
      "epoch": 2.3556463975472663,
      "grad_norm": 0.18473631827351808,
      "learning_rate": 5e-06,
      "loss": 0.719,
      "step": 4610
    },
    {
      "epoch": 2.3607562595809912,
      "grad_norm": 0.19338189496040809,
      "learning_rate": 5e-06,
      "loss": 0.7239,
      "step": 4620
    },
    {
      "epoch": 2.3658661216147165,
      "grad_norm": 0.19066355345819264,
      "learning_rate": 5e-06,
      "loss": 0.7266,
      "step": 4630
    },
    {
      "epoch": 2.3709759836484414,
      "grad_norm": 0.1883524242314222,
      "learning_rate": 5e-06,
      "loss": 0.7186,
      "step": 4640
    },
    {
      "epoch": 2.3760858456821667,
      "grad_norm": 0.18502939250064523,
      "learning_rate": 5e-06,
      "loss": 0.7199,
      "step": 4650
    },
    {
      "epoch": 2.3811957077158916,
      "grad_norm": 0.1792178984338457,
      "learning_rate": 5e-06,
      "loss": 0.714,
      "step": 4660
    },
    {
      "epoch": 2.386305569749617,
      "grad_norm": 0.18226944439325007,
      "learning_rate": 5e-06,
      "loss": 0.7041,
      "step": 4670
    },
    {
      "epoch": 2.391415431783342,
      "grad_norm": 0.18015473190902764,
      "learning_rate": 5e-06,
      "loss": 0.716,
      "step": 4680
    },
    {
      "epoch": 2.396525293817067,
      "grad_norm": 0.18115429600775665,
      "learning_rate": 5e-06,
      "loss": 0.7086,
      "step": 4690
    },
    {
      "epoch": 2.401635155850792,
      "grad_norm": 0.2043473041616948,
      "learning_rate": 5e-06,
      "loss": 0.7165,
      "step": 4700
    },
    {
      "epoch": 2.406745017884517,
      "grad_norm": 0.20268566073488395,
      "learning_rate": 5e-06,
      "loss": 0.7153,
      "step": 4710
    },
    {
      "epoch": 2.411854879918242,
      "grad_norm": 0.19488489934465453,
      "learning_rate": 5e-06,
      "loss": 0.7261,
      "step": 4720
    },
    {
      "epoch": 2.4169647419519675,
      "grad_norm": 0.1869426176451986,
      "learning_rate": 5e-06,
      "loss": 0.6977,
      "step": 4730
    },
    {
      "epoch": 2.4220746039856924,
      "grad_norm": 0.19817580694708986,
      "learning_rate": 5e-06,
      "loss": 0.71,
      "step": 4740
    },
    {
      "epoch": 2.4271844660194173,
      "grad_norm": 0.2099570721696579,
      "learning_rate": 5e-06,
      "loss": 0.7204,
      "step": 4750
    },
    {
      "epoch": 2.4322943280531426,
      "grad_norm": 0.1964657272176851,
      "learning_rate": 5e-06,
      "loss": 0.7104,
      "step": 4760
    },
    {
      "epoch": 2.4374041900868675,
      "grad_norm": 0.1886905609835038,
      "learning_rate": 5e-06,
      "loss": 0.7085,
      "step": 4770
    },
    {
      "epoch": 2.442514052120593,
      "grad_norm": 0.18220683381465547,
      "learning_rate": 5e-06,
      "loss": 0.722,
      "step": 4780
    },
    {
      "epoch": 2.4476239141543177,
      "grad_norm": 0.18487998134947345,
      "learning_rate": 5e-06,
      "loss": 0.7112,
      "step": 4790
    },
    {
      "epoch": 2.452733776188043,
      "grad_norm": 0.16830917755794478,
      "learning_rate": 5e-06,
      "loss": 0.7157,
      "step": 4800
    },
    {
      "epoch": 2.457843638221768,
      "grad_norm": 0.18354244943334422,
      "learning_rate": 5e-06,
      "loss": 0.7152,
      "step": 4810
    },
    {
      "epoch": 2.462953500255493,
      "grad_norm": 0.18986840496033355,
      "learning_rate": 5e-06,
      "loss": 0.7161,
      "step": 4820
    },
    {
      "epoch": 2.468063362289218,
      "grad_norm": 0.1826318712937385,
      "learning_rate": 5e-06,
      "loss": 0.7295,
      "step": 4830
    },
    {
      "epoch": 2.4731732243229434,
      "grad_norm": 0.18527008923688568,
      "learning_rate": 5e-06,
      "loss": 0.7115,
      "step": 4840
    },
    {
      "epoch": 2.4782830863566683,
      "grad_norm": 0.18224411387144135,
      "learning_rate": 5e-06,
      "loss": 0.7173,
      "step": 4850
    },
    {
      "epoch": 2.4833929483903936,
      "grad_norm": 0.17811035407666295,
      "learning_rate": 5e-06,
      "loss": 0.7095,
      "step": 4860
    },
    {
      "epoch": 2.4885028104241185,
      "grad_norm": 0.18261169308001168,
      "learning_rate": 5e-06,
      "loss": 0.7162,
      "step": 4870
    },
    {
      "epoch": 2.493612672457844,
      "grad_norm": 0.2002887514635378,
      "learning_rate": 5e-06,
      "loss": 0.7208,
      "step": 4880
    },
    {
      "epoch": 2.4987225344915687,
      "grad_norm": 0.1869680657646229,
      "learning_rate": 5e-06,
      "loss": 0.7081,
      "step": 4890
    },
    {
      "epoch": 2.5038323965252935,
      "grad_norm": 0.18675570886820952,
      "learning_rate": 5e-06,
      "loss": 0.7117,
      "step": 4900
    },
    {
      "epoch": 2.508942258559019,
      "grad_norm": 0.19944201285579877,
      "learning_rate": 5e-06,
      "loss": 0.7198,
      "step": 4910
    },
    {
      "epoch": 2.514052120592744,
      "grad_norm": 0.18956975903685294,
      "learning_rate": 5e-06,
      "loss": 0.7179,
      "step": 4920
    },
    {
      "epoch": 2.519161982626469,
      "grad_norm": 0.1942120087148544,
      "learning_rate": 5e-06,
      "loss": 0.7216,
      "step": 4930
    },
    {
      "epoch": 2.524271844660194,
      "grad_norm": 0.18623164151592378,
      "learning_rate": 5e-06,
      "loss": 0.6997,
      "step": 4940
    },
    {
      "epoch": 2.5293817066939193,
      "grad_norm": 0.19081196748322382,
      "learning_rate": 5e-06,
      "loss": 0.7091,
      "step": 4950
    },
    {
      "epoch": 2.5344915687276446,
      "grad_norm": 0.191889152108264,
      "learning_rate": 5e-06,
      "loss": 0.7122,
      "step": 4960
    },
    {
      "epoch": 2.5396014307613695,
      "grad_norm": 0.19018109192168367,
      "learning_rate": 5e-06,
      "loss": 0.7195,
      "step": 4970
    },
    {
      "epoch": 2.5447112927950943,
      "grad_norm": 0.18029975730571046,
      "learning_rate": 5e-06,
      "loss": 0.7067,
      "step": 4980
    },
    {
      "epoch": 2.5498211548288197,
      "grad_norm": 0.19081905131548996,
      "learning_rate": 5e-06,
      "loss": 0.7184,
      "step": 4990
    },
    {
      "epoch": 2.554931016862545,
      "grad_norm": 0.18184068561404432,
      "learning_rate": 5e-06,
      "loss": 0.711,
      "step": 5000
    },
    {
      "epoch": 2.56004087889627,
      "grad_norm": 0.18141265501857734,
      "learning_rate": 5e-06,
      "loss": 0.7298,
      "step": 5010
    },
    {
      "epoch": 2.5651507409299947,
      "grad_norm": 0.19059042586600472,
      "learning_rate": 5e-06,
      "loss": 0.7114,
      "step": 5020
    },
    {
      "epoch": 2.57026060296372,
      "grad_norm": 0.18752747662093455,
      "learning_rate": 5e-06,
      "loss": 0.7139,
      "step": 5030
    },
    {
      "epoch": 2.575370464997445,
      "grad_norm": 0.18531393933159526,
      "learning_rate": 5e-06,
      "loss": 0.7152,
      "step": 5040
    },
    {
      "epoch": 2.5804803270311703,
      "grad_norm": 0.2020545252464496,
      "learning_rate": 5e-06,
      "loss": 0.724,
      "step": 5050
    },
    {
      "epoch": 2.585590189064895,
      "grad_norm": 0.1817661308445167,
      "learning_rate": 5e-06,
      "loss": 0.7099,
      "step": 5060
    },
    {
      "epoch": 2.5907000510986204,
      "grad_norm": 0.1874710204305083,
      "learning_rate": 5e-06,
      "loss": 0.727,
      "step": 5070
    },
    {
      "epoch": 2.5958099131323453,
      "grad_norm": 0.19697140144619885,
      "learning_rate": 5e-06,
      "loss": 0.7184,
      "step": 5080
    },
    {
      "epoch": 2.6009197751660706,
      "grad_norm": 0.1889733854232041,
      "learning_rate": 5e-06,
      "loss": 0.7169,
      "step": 5090
    },
    {
      "epoch": 2.6060296371997955,
      "grad_norm": 0.20000474912796498,
      "learning_rate": 5e-06,
      "loss": 0.7089,
      "step": 5100
    },
    {
      "epoch": 2.611139499233521,
      "grad_norm": 0.18020813610110156,
      "learning_rate": 5e-06,
      "loss": 0.7264,
      "step": 5110
    },
    {
      "epoch": 2.6162493612672457,
      "grad_norm": 0.17876620180082428,
      "learning_rate": 5e-06,
      "loss": 0.7007,
      "step": 5120
    },
    {
      "epoch": 2.6213592233009706,
      "grad_norm": 0.186937820916383,
      "learning_rate": 5e-06,
      "loss": 0.7219,
      "step": 5130
    },
    {
      "epoch": 2.626469085334696,
      "grad_norm": 0.19293145357331443,
      "learning_rate": 5e-06,
      "loss": 0.7116,
      "step": 5140
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.18779972078705487,
      "learning_rate": 5e-06,
      "loss": 0.7147,
      "step": 5150
    },
    {
      "epoch": 2.636688809402146,
      "grad_norm": 0.2004320087337195,
      "learning_rate": 5e-06,
      "loss": 0.7046,
      "step": 5160
    },
    {
      "epoch": 2.641798671435871,
      "grad_norm": 0.18155864298582336,
      "learning_rate": 5e-06,
      "loss": 0.7024,
      "step": 5170
    },
    {
      "epoch": 2.6469085334695963,
      "grad_norm": 0.18766505517066065,
      "learning_rate": 5e-06,
      "loss": 0.7092,
      "step": 5180
    },
    {
      "epoch": 2.6520183955033216,
      "grad_norm": 0.18241808648535346,
      "learning_rate": 5e-06,
      "loss": 0.7231,
      "step": 5190
    },
    {
      "epoch": 2.6571282575370465,
      "grad_norm": 0.1956290063678602,
      "learning_rate": 5e-06,
      "loss": 0.7071,
      "step": 5200
    },
    {
      "epoch": 2.6622381195707714,
      "grad_norm": 0.19687281837098963,
      "learning_rate": 5e-06,
      "loss": 0.7237,
      "step": 5210
    },
    {
      "epoch": 2.6673479816044967,
      "grad_norm": 0.18253071868952309,
      "learning_rate": 5e-06,
      "loss": 0.7085,
      "step": 5220
    },
    {
      "epoch": 2.672457843638222,
      "grad_norm": 0.18783632566318903,
      "learning_rate": 5e-06,
      "loss": 0.717,
      "step": 5230
    },
    {
      "epoch": 2.677567705671947,
      "grad_norm": 0.1856712541693123,
      "learning_rate": 5e-06,
      "loss": 0.7121,
      "step": 5240
    },
    {
      "epoch": 2.682677567705672,
      "grad_norm": 0.17913630024093868,
      "learning_rate": 5e-06,
      "loss": 0.7141,
      "step": 5250
    },
    {
      "epoch": 2.687787429739397,
      "grad_norm": 0.1814607328558563,
      "learning_rate": 5e-06,
      "loss": 0.7143,
      "step": 5260
    },
    {
      "epoch": 2.692897291773122,
      "grad_norm": 0.20516860190923697,
      "learning_rate": 5e-06,
      "loss": 0.7153,
      "step": 5270
    },
    {
      "epoch": 2.6980071538068473,
      "grad_norm": 0.18691014833866346,
      "learning_rate": 5e-06,
      "loss": 0.7021,
      "step": 5280
    },
    {
      "epoch": 2.703117015840572,
      "grad_norm": 0.19243627032023453,
      "learning_rate": 5e-06,
      "loss": 0.7198,
      "step": 5290
    },
    {
      "epoch": 2.7082268778742975,
      "grad_norm": 0.1782810408332171,
      "learning_rate": 5e-06,
      "loss": 0.7144,
      "step": 5300
    },
    {
      "epoch": 2.7133367399080224,
      "grad_norm": 0.18246834780709842,
      "learning_rate": 5e-06,
      "loss": 0.7073,
      "step": 5310
    },
    {
      "epoch": 2.7184466019417477,
      "grad_norm": 0.19188458416675994,
      "learning_rate": 5e-06,
      "loss": 0.7033,
      "step": 5320
    },
    {
      "epoch": 2.7235564639754726,
      "grad_norm": 0.19431224425803312,
      "learning_rate": 5e-06,
      "loss": 0.7121,
      "step": 5330
    },
    {
      "epoch": 2.728666326009198,
      "grad_norm": 0.19472680477435397,
      "learning_rate": 5e-06,
      "loss": 0.7204,
      "step": 5340
    },
    {
      "epoch": 2.7337761880429228,
      "grad_norm": 0.17416333272267362,
      "learning_rate": 5e-06,
      "loss": 0.722,
      "step": 5350
    },
    {
      "epoch": 2.738886050076648,
      "grad_norm": 0.17997245175531543,
      "learning_rate": 5e-06,
      "loss": 0.7209,
      "step": 5360
    },
    {
      "epoch": 2.743995912110373,
      "grad_norm": 0.1818254672253798,
      "learning_rate": 5e-06,
      "loss": 0.7064,
      "step": 5370
    },
    {
      "epoch": 2.7491057741440983,
      "grad_norm": 0.19143052614657435,
      "learning_rate": 5e-06,
      "loss": 0.7251,
      "step": 5380
    },
    {
      "epoch": 2.754215636177823,
      "grad_norm": 0.1849449249230007,
      "learning_rate": 5e-06,
      "loss": 0.7134,
      "step": 5390
    },
    {
      "epoch": 2.759325498211548,
      "grad_norm": 0.18960800644635575,
      "learning_rate": 5e-06,
      "loss": 0.7132,
      "step": 5400
    },
    {
      "epoch": 2.7644353602452734,
      "grad_norm": 0.18450711551330742,
      "learning_rate": 5e-06,
      "loss": 0.7218,
      "step": 5410
    },
    {
      "epoch": 2.7695452222789987,
      "grad_norm": 0.1775798100347251,
      "learning_rate": 5e-06,
      "loss": 0.7246,
      "step": 5420
    },
    {
      "epoch": 2.7746550843127236,
      "grad_norm": 0.19385869953809406,
      "learning_rate": 5e-06,
      "loss": 0.7154,
      "step": 5430
    },
    {
      "epoch": 2.7797649463464484,
      "grad_norm": 0.18947315882216123,
      "learning_rate": 5e-06,
      "loss": 0.7146,
      "step": 5440
    },
    {
      "epoch": 2.7848748083801738,
      "grad_norm": 0.19874490028355687,
      "learning_rate": 5e-06,
      "loss": 0.7098,
      "step": 5450
    },
    {
      "epoch": 2.789984670413899,
      "grad_norm": 0.19216650910265412,
      "learning_rate": 5e-06,
      "loss": 0.7157,
      "step": 5460
    },
    {
      "epoch": 2.795094532447624,
      "grad_norm": 0.1884532290700664,
      "learning_rate": 5e-06,
      "loss": 0.73,
      "step": 5470
    },
    {
      "epoch": 2.800204394481349,
      "grad_norm": 0.178944069456414,
      "learning_rate": 5e-06,
      "loss": 0.7023,
      "step": 5480
    },
    {
      "epoch": 2.805314256515074,
      "grad_norm": 0.19897724638804584,
      "learning_rate": 5e-06,
      "loss": 0.7266,
      "step": 5490
    },
    {
      "epoch": 2.810424118548799,
      "grad_norm": 0.19676207860709408,
      "learning_rate": 5e-06,
      "loss": 0.7042,
      "step": 5500
    },
    {
      "epoch": 2.8155339805825244,
      "grad_norm": 0.17311366117436,
      "learning_rate": 5e-06,
      "loss": 0.7059,
      "step": 5510
    },
    {
      "epoch": 2.8206438426162492,
      "grad_norm": 0.1862348663291006,
      "learning_rate": 5e-06,
      "loss": 0.7012,
      "step": 5520
    },
    {
      "epoch": 2.8257537046499746,
      "grad_norm": 0.19532852701482903,
      "learning_rate": 5e-06,
      "loss": 0.7105,
      "step": 5530
    },
    {
      "epoch": 2.8308635666836994,
      "grad_norm": 0.18410412264007187,
      "learning_rate": 5e-06,
      "loss": 0.7163,
      "step": 5540
    },
    {
      "epoch": 2.8359734287174247,
      "grad_norm": 0.21025902958863738,
      "learning_rate": 5e-06,
      "loss": 0.7359,
      "step": 5550
    },
    {
      "epoch": 2.8410832907511496,
      "grad_norm": 0.17818540055082727,
      "learning_rate": 5e-06,
      "loss": 0.7111,
      "step": 5560
    },
    {
      "epoch": 2.846193152784875,
      "grad_norm": 0.1842948913149892,
      "learning_rate": 5e-06,
      "loss": 0.7167,
      "step": 5570
    },
    {
      "epoch": 2.8513030148186,
      "grad_norm": 0.18686037650976978,
      "learning_rate": 5e-06,
      "loss": 0.7138,
      "step": 5580
    },
    {
      "epoch": 2.856412876852325,
      "grad_norm": 0.203403946822995,
      "learning_rate": 5e-06,
      "loss": 0.6996,
      "step": 5590
    },
    {
      "epoch": 2.86152273888605,
      "grad_norm": 0.19304193259535427,
      "learning_rate": 5e-06,
      "loss": 0.7184,
      "step": 5600
    },
    {
      "epoch": 2.8666326009197753,
      "grad_norm": 0.17636221017029402,
      "learning_rate": 5e-06,
      "loss": 0.6989,
      "step": 5610
    },
    {
      "epoch": 2.8717424629535,
      "grad_norm": 0.2037376472703345,
      "learning_rate": 5e-06,
      "loss": 0.7179,
      "step": 5620
    },
    {
      "epoch": 2.876852324987225,
      "grad_norm": 0.19200332931870842,
      "learning_rate": 5e-06,
      "loss": 0.7139,
      "step": 5630
    },
    {
      "epoch": 2.8819621870209504,
      "grad_norm": 0.18491595431761487,
      "learning_rate": 5e-06,
      "loss": 0.7128,
      "step": 5640
    },
    {
      "epoch": 2.8870720490546757,
      "grad_norm": 0.1838785582117868,
      "learning_rate": 5e-06,
      "loss": 0.6955,
      "step": 5650
    },
    {
      "epoch": 2.8921819110884006,
      "grad_norm": 0.19428058651152774,
      "learning_rate": 5e-06,
      "loss": 0.7102,
      "step": 5660
    },
    {
      "epoch": 2.8972917731221255,
      "grad_norm": 0.17936626194327465,
      "learning_rate": 5e-06,
      "loss": 0.7078,
      "step": 5670
    },
    {
      "epoch": 2.902401635155851,
      "grad_norm": 0.17631468596491473,
      "learning_rate": 5e-06,
      "loss": 0.7162,
      "step": 5680
    },
    {
      "epoch": 2.907511497189576,
      "grad_norm": 0.18415745158446212,
      "learning_rate": 5e-06,
      "loss": 0.7233,
      "step": 5690
    },
    {
      "epoch": 2.912621359223301,
      "grad_norm": 0.17279113129182289,
      "learning_rate": 5e-06,
      "loss": 0.6992,
      "step": 5700
    },
    {
      "epoch": 2.917731221257026,
      "grad_norm": 0.19005862366918605,
      "learning_rate": 5e-06,
      "loss": 0.7015,
      "step": 5710
    },
    {
      "epoch": 2.922841083290751,
      "grad_norm": 0.1865410089900959,
      "learning_rate": 5e-06,
      "loss": 0.7043,
      "step": 5720
    },
    {
      "epoch": 2.927950945324476,
      "grad_norm": 0.19166084388796084,
      "learning_rate": 5e-06,
      "loss": 0.7212,
      "step": 5730
    },
    {
      "epoch": 2.9330608073582014,
      "grad_norm": 0.19183886093470404,
      "learning_rate": 5e-06,
      "loss": 0.7117,
      "step": 5740
    },
    {
      "epoch": 2.9381706693919263,
      "grad_norm": 0.18542747500934698,
      "learning_rate": 5e-06,
      "loss": 0.7031,
      "step": 5750
    },
    {
      "epoch": 2.9432805314256516,
      "grad_norm": 0.19800100325472195,
      "learning_rate": 5e-06,
      "loss": 0.7003,
      "step": 5760
    },
    {
      "epoch": 2.9483903934593765,
      "grad_norm": 0.19068782203027565,
      "learning_rate": 5e-06,
      "loss": 0.7157,
      "step": 5770
    },
    {
      "epoch": 2.953500255493102,
      "grad_norm": 0.1893011994562663,
      "learning_rate": 5e-06,
      "loss": 0.7049,
      "step": 5780
    },
    {
      "epoch": 2.9586101175268267,
      "grad_norm": 0.17894632421886825,
      "learning_rate": 5e-06,
      "loss": 0.721,
      "step": 5790
    },
    {
      "epoch": 2.963719979560552,
      "grad_norm": 0.18749644947518349,
      "learning_rate": 5e-06,
      "loss": 0.7245,
      "step": 5800
    },
    {
      "epoch": 2.968829841594277,
      "grad_norm": 0.18920042895955116,
      "learning_rate": 5e-06,
      "loss": 0.7086,
      "step": 5810
    },
    {
      "epoch": 2.973939703628002,
      "grad_norm": 0.18597523756708192,
      "learning_rate": 5e-06,
      "loss": 0.7143,
      "step": 5820
    },
    {
      "epoch": 2.979049565661727,
      "grad_norm": 0.18741696462849414,
      "learning_rate": 5e-06,
      "loss": 0.7195,
      "step": 5830
    },
    {
      "epoch": 2.9841594276954524,
      "grad_norm": 0.18985455877786367,
      "learning_rate": 5e-06,
      "loss": 0.7091,
      "step": 5840
    },
    {
      "epoch": 2.9892692897291773,
      "grad_norm": 0.18003618910403507,
      "learning_rate": 5e-06,
      "loss": 0.7111,
      "step": 5850
    },
    {
      "epoch": 2.994379151762902,
      "grad_norm": 0.19511416723954572,
      "learning_rate": 5e-06,
      "loss": 0.7105,
      "step": 5860
    },
    {
      "epoch": 2.9994890137966275,
      "grad_norm": 0.18652462173874393,
      "learning_rate": 5e-06,
      "loss": 0.7294,
      "step": 5870
    },
    {
      "epoch": 3.0,
      "step": 5871,
      "total_flos": 2506240622592000.0,
      "train_loss": 0.754048796695819,
      "train_runtime": 27383.3094,
      "train_samples_per_second": 109.726,
      "train_steps_per_second": 0.214
    }
  ],
  "logging_steps": 10,
  "max_steps": 5871,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2506240622592000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}