{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1004,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00796812749003984,
      "grad_norm": 0.9957464337348938,
      "learning_rate": 3.921568627450981e-07,
      "loss": 2.092358350753784,
      "step": 2
    },
    {
      "epoch": 0.01593625498007968,
      "grad_norm": 0.8930569887161255,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 2.0088067054748535,
      "step": 4
    },
    {
      "epoch": 0.02390438247011952,
      "grad_norm": 2.4471495151519775,
      "learning_rate": 1.96078431372549e-06,
      "loss": 2.108173370361328,
      "step": 6
    },
    {
      "epoch": 0.03187250996015936,
      "grad_norm": 0.4212417006492615,
      "learning_rate": 2.7450980392156867e-06,
      "loss": 1.815500020980835,
      "step": 8
    },
    {
      "epoch": 0.0398406374501992,
      "grad_norm": 0.9693068861961365,
      "learning_rate": 3.529411764705883e-06,
      "loss": 1.2837271690368652,
      "step": 10
    },
    {
      "epoch": 0.04780876494023904,
      "grad_norm": 0.7192004919052124,
      "learning_rate": 4.313725490196079e-06,
      "loss": 1.5911014080047607,
      "step": 12
    },
    {
      "epoch": 0.055776892430278883,
      "grad_norm": 0.4175472557544708,
      "learning_rate": 5.098039215686274e-06,
      "loss": 1.6157869100570679,
      "step": 14
    },
    {
      "epoch": 0.06374501992031872,
      "grad_norm": 3.3379127979278564,
      "learning_rate": 5.882352941176471e-06,
      "loss": 1.7002140283584595,
      "step": 16
    },
    {
      "epoch": 0.07171314741035857,
      "grad_norm": 0.7876597046852112,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.5408495664596558,
      "step": 18
    },
    {
      "epoch": 0.0796812749003984,
      "grad_norm": 0.46036335825920105,
      "learning_rate": 7.450980392156863e-06,
      "loss": 1.4705753326416016,
      "step": 20
    },
    {
      "epoch": 0.08764940239043825,
      "grad_norm": 0.673358678817749,
      "learning_rate": 8.23529411764706e-06,
      "loss": 1.1017413139343262,
      "step": 22
    },
    {
      "epoch": 0.09561752988047809,
      "grad_norm": 0.7209763526916504,
      "learning_rate": 9.019607843137256e-06,
      "loss": 1.294268250465393,
      "step": 24
    },
    {
      "epoch": 0.10358565737051793,
      "grad_norm": 0.4403802752494812,
      "learning_rate": 9.803921568627451e-06,
      "loss": 1.395767331123352,
      "step": 26
    },
    {
      "epoch": 0.11155378486055777,
      "grad_norm": 0.4559030830860138,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 1.6222474575042725,
      "step": 28
    },
    {
      "epoch": 0.11952191235059761,
      "grad_norm": 0.8342038989067078,
      "learning_rate": 1.1372549019607844e-05,
      "loss": 1.3161066770553589,
      "step": 30
    },
    {
      "epoch": 0.12749003984063745,
      "grad_norm": 0.5698347091674805,
      "learning_rate": 1.215686274509804e-05,
      "loss": 1.1639069318771362,
      "step": 32
    },
    {
      "epoch": 0.13545816733067728,
      "grad_norm": 0.6199248433113098,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 1.24606192111969,
      "step": 34
    },
    {
      "epoch": 0.14342629482071714,
      "grad_norm": 0.288705974817276,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 1.3773506879806519,
      "step": 36
    },
    {
      "epoch": 0.15139442231075698,
      "grad_norm": 0.6997499465942383,
      "learning_rate": 1.4509803921568629e-05,
      "loss": 1.175160527229309,
      "step": 38
    },
    {
      "epoch": 0.1593625498007968,
      "grad_norm": 0.5133037567138672,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 1.2939589023590088,
      "step": 40
    },
    {
      "epoch": 0.16733067729083664,
      "grad_norm": 1.701066017150879,
      "learning_rate": 1.607843137254902e-05,
      "loss": 0.8205271363258362,
      "step": 42
    },
    {
      "epoch": 0.1752988047808765,
      "grad_norm": 0.4490733742713928,
      "learning_rate": 1.686274509803922e-05,
      "loss": 0.9646032452583313,
      "step": 44
    },
    {
      "epoch": 0.18326693227091634,
      "grad_norm": 0.7051479816436768,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.9549788236618042,
      "step": 46
    },
    {
      "epoch": 0.19123505976095617,
      "grad_norm": 0.8199656009674072,
      "learning_rate": 1.843137254901961e-05,
      "loss": 1.1781502962112427,
      "step": 48
    },
    {
      "epoch": 0.199203187250996,
      "grad_norm": 0.45470499992370605,
      "learning_rate": 1.9215686274509807e-05,
      "loss": 1.3290269374847412,
      "step": 50
    },
    {
      "epoch": 0.20717131474103587,
      "grad_norm": 0.850440502166748,
      "learning_rate": 2e-05,
      "loss": 1.2819088697433472,
      "step": 52
    },
    {
      "epoch": 0.2151394422310757,
      "grad_norm": 2.6981277465820312,
      "learning_rate": 1.9999804392782173e-05,
      "loss": 1.1220871210098267,
      "step": 54
    },
    {
      "epoch": 0.22310756972111553,
      "grad_norm": 0.5424857139587402,
      "learning_rate": 1.9999217579631398e-05,
      "loss": 1.534443974494934,
      "step": 56
    },
    {
      "epoch": 0.23107569721115537,
      "grad_norm": 0.3990076780319214,
      "learning_rate": 1.9998239586055426e-05,
      "loss": 1.3426909446716309,
      "step": 58
    },
    {
      "epoch": 0.23904382470119523,
      "grad_norm": 0.36483845114707947,
      "learning_rate": 1.999687045456595e-05,
      "loss": 0.7846556305885315,
      "step": 60
    },
    {
      "epoch": 0.24701195219123506,
      "grad_norm": 0.3828938603401184,
      "learning_rate": 1.9995110244676744e-05,
      "loss": 1.3494534492492676,
      "step": 62
    },
    {
      "epoch": 0.2549800796812749,
      "grad_norm": 1.3368655443191528,
      "learning_rate": 1.999295903290109e-05,
      "loss": 1.1254225969314575,
      "step": 64
    },
    {
      "epoch": 0.26294820717131473,
      "grad_norm": 0.5956477522850037,
      "learning_rate": 1.999041691274844e-05,
      "loss": 1.283703327178955,
      "step": 66
    },
    {
      "epoch": 0.27091633466135456,
      "grad_norm": 1.2891621589660645,
      "learning_rate": 1.998748399472037e-05,
      "loss": 0.9298644661903381,
      "step": 68
    },
    {
      "epoch": 0.2788844621513944,
      "grad_norm": 0.5659030675888062,
      "learning_rate": 1.9984160406305745e-05,
      "loss": 1.5809705257415771,
      "step": 70
    },
    {
      "epoch": 0.2868525896414343,
      "grad_norm": 0.6271833181381226,
      "learning_rate": 1.9980446291975217e-05,
      "loss": 0.895917534828186,
      "step": 72
    },
    {
      "epoch": 0.2948207171314741,
      "grad_norm": 1.1333332061767578,
      "learning_rate": 1.99763418131749e-05,
      "loss": 1.1716889142990112,
      "step": 74
    },
    {
      "epoch": 0.30278884462151395,
      "grad_norm": 0.9091482758522034,
      "learning_rate": 1.9971847148319392e-05,
      "loss": 1.4918345212936401,
      "step": 76
    },
    {
      "epoch": 0.3107569721115538,
      "grad_norm": 1.295212745666504,
      "learning_rate": 1.9966962492784006e-05,
      "loss": 1.0331875085830688,
      "step": 78
    },
    {
      "epoch": 0.3187250996015936,
      "grad_norm": 0.2440294772386551,
      "learning_rate": 1.996168805889627e-05,
      "loss": 0.924660861492157,
      "step": 80
    },
    {
      "epoch": 0.32669322709163345,
      "grad_norm": 0.28573158383369446,
      "learning_rate": 1.99560240759267e-05,
      "loss": 1.375991702079773,
      "step": 82
    },
    {
      "epoch": 0.3346613545816733,
      "grad_norm": 0.45677199959754944,
      "learning_rate": 1.9949970790078847e-05,
      "loss": 0.9936637878417969,
      "step": 84
    },
    {
      "epoch": 0.3426294820717131,
      "grad_norm": 0.3065952956676483,
      "learning_rate": 1.9943528464478573e-05,
      "loss": 1.3540968894958496,
      "step": 86
    },
    {
      "epoch": 0.350597609561753,
      "grad_norm": 1.1819368600845337,
      "learning_rate": 1.993669737916263e-05,
      "loss": 1.0978449583053589,
      "step": 88
    },
    {
      "epoch": 0.35856573705179284,
      "grad_norm": 0.39413952827453613,
      "learning_rate": 1.9929477831066485e-05,
      "loss": 1.3455610275268555,
      "step": 90
    },
    {
      "epoch": 0.3665338645418327,
      "grad_norm": 0.586239755153656,
      "learning_rate": 1.992187013401141e-05,
      "loss": 1.158692717552185,
      "step": 92
    },
    {
      "epoch": 0.3745019920318725,
      "grad_norm": 0.2929425537586212,
      "learning_rate": 1.9913874618690837e-05,
      "loss": 1.2582529783248901,
      "step": 94
    },
    {
      "epoch": 0.38247011952191234,
      "grad_norm": 0.4790647327899933,
      "learning_rate": 1.9905491632655996e-05,
      "loss": 1.1448304653167725,
      "step": 96
    },
    {
      "epoch": 0.3904382470119522,
      "grad_norm": 0.31453412771224976,
      "learning_rate": 1.989672154030078e-05,
      "loss": 1.277166485786438,
      "step": 98
    },
    {
      "epoch": 0.398406374501992,
      "grad_norm": 0.27740514278411865,
      "learning_rate": 1.988756472284595e-05,
      "loss": 0.9082943201065063,
      "step": 100
    },
    {
      "epoch": 0.4063745019920319,
      "grad_norm": 0.516435444355011,
      "learning_rate": 1.9878021578322518e-05,
      "loss": 1.2856113910675049,
      "step": 102
    },
    {
      "epoch": 0.41434262948207173,
      "grad_norm": 0.38994523882865906,
      "learning_rate": 1.9868092521554478e-05,
      "loss": 0.8686521053314209,
      "step": 104
    },
    {
      "epoch": 0.42231075697211157,
      "grad_norm": 1.3962106704711914,
      "learning_rate": 1.9857777984140746e-05,
      "loss": 0.766525387763977,
      "step": 106
    },
    {
      "epoch": 0.4302788844621514,
      "grad_norm": 0.48875540494918823,
      "learning_rate": 1.9847078414436438e-05,
      "loss": 1.0129103660583496,
      "step": 108
    },
    {
      "epoch": 0.43824701195219123,
      "grad_norm": 0.679126501083374,
      "learning_rate": 1.983599427753334e-05,
      "loss": 1.319523572921753,
      "step": 110
    },
    {
      "epoch": 0.44621513944223107,
      "grad_norm": 0.6783590912818909,
      "learning_rate": 1.9824526055239712e-05,
      "loss": 0.8960539698600769,
      "step": 112
    },
    {
      "epoch": 0.4541832669322709,
      "grad_norm": 0.6845623254776001,
      "learning_rate": 1.981267424605935e-05,
      "loss": 1.1875278949737549,
      "step": 114
    },
    {
      "epoch": 0.46215139442231074,
      "grad_norm": 0.3849475085735321,
      "learning_rate": 1.98004393651699e-05,
      "loss": 1.332448959350586,
      "step": 116
    },
    {
      "epoch": 0.4701195219123506,
      "grad_norm": 0.3897473216056824,
      "learning_rate": 1.9787821944400477e-05,
      "loss": 1.3839960098266602,
      "step": 118
    },
    {
      "epoch": 0.47808764940239046,
      "grad_norm": 0.3927645981311798,
      "learning_rate": 1.9774822532208537e-05,
      "loss": 1.2699512243270874,
      "step": 120
    },
    {
      "epoch": 0.4860557768924303,
      "grad_norm": 0.4918311834335327,
      "learning_rate": 1.9761441693656052e-05,
      "loss": 1.0304157733917236,
      "step": 122
    },
    {
      "epoch": 0.4940239043824701,
      "grad_norm": 1.8012769222259521,
      "learning_rate": 1.974768001038493e-05,
      "loss": 1.088665246963501,
      "step": 124
    },
    {
      "epoch": 0.50199203187251,
      "grad_norm": 0.7356415390968323,
      "learning_rate": 1.973353808059175e-05,
      "loss": 0.9787855744361877,
      "step": 126
    },
    {
      "epoch": 0.5099601593625498,
      "grad_norm": 0.4110127091407776,
      "learning_rate": 1.9719016519001738e-05,
      "loss": 1.3301374912261963,
      "step": 128
    },
    {
      "epoch": 0.5179282868525896,
      "grad_norm": 0.3917061984539032,
      "learning_rate": 1.9704115956842066e-05,
      "loss": 0.9724149703979492,
      "step": 130
    },
    {
      "epoch": 0.5258964143426295,
      "grad_norm": 0.41339319944381714,
      "learning_rate": 1.9688837041814396e-05,
      "loss": 0.9403734803199768,
      "step": 132
    },
    {
      "epoch": 0.5338645418326693,
      "grad_norm": 0.4448390603065491,
      "learning_rate": 1.967318043806675e-05,
      "loss": 0.9430965781211853,
      "step": 134
    },
    {
      "epoch": 0.5418326693227091,
      "grad_norm": 1.0087687969207764,
      "learning_rate": 1.965714682616461e-05,
      "loss": 1.3703218698501587,
      "step": 136
    },
    {
      "epoch": 0.549800796812749,
      "grad_norm": 0.2762286067008972,
      "learning_rate": 1.9640736903061363e-05,
      "loss": 1.0063937902450562,
      "step": 138
    },
    {
      "epoch": 0.5577689243027888,
      "grad_norm": 0.6342159509658813,
      "learning_rate": 1.9623951382067983e-05,
      "loss": 1.1634577512741089,
      "step": 140
    },
    {
      "epoch": 0.5657370517928287,
      "grad_norm": 0.4940018951892853,
      "learning_rate": 1.960679099282204e-05,
      "loss": 1.080363392829895,
      "step": 142
    },
    {
      "epoch": 0.5737051792828686,
      "grad_norm": 0.361111044883728,
      "learning_rate": 1.9589256481255984e-05,
      "loss": 1.2627440690994263,
      "step": 144
    },
    {
      "epoch": 0.5816733067729084,
      "grad_norm": 0.6875961422920227,
      "learning_rate": 1.9571348609564697e-05,
      "loss": 1.2056587934494019,
      "step": 146
    },
    {
      "epoch": 0.5896414342629482,
      "grad_norm": 0.543710470199585,
      "learning_rate": 1.9553068156172404e-05,
      "loss": 1.059579849243164,
      "step": 148
    },
    {
      "epoch": 0.5976095617529881,
      "grad_norm": 0.5071445107460022,
      "learning_rate": 1.9534415915698793e-05,
      "loss": 1.0381158590316772,
      "step": 150
    },
    {
      "epoch": 0.6055776892430279,
      "grad_norm": 0.36371827125549316,
      "learning_rate": 1.9515392698924504e-05,
      "loss": 1.0027271509170532,
      "step": 152
    },
    {
      "epoch": 0.6135458167330677,
      "grad_norm": 0.6223253011703491,
      "learning_rate": 1.949599933275587e-05,
      "loss": 1.1761221885681152,
      "step": 154
    },
    {
      "epoch": 0.6215139442231076,
      "grad_norm": 2.302541494369507,
      "learning_rate": 1.9476236660188982e-05,
      "loss": 1.217422366142273,
      "step": 156
    },
    {
      "epoch": 0.6294820717131474,
      "grad_norm": 0.9959671497344971,
      "learning_rate": 1.9456105540273035e-05,
      "loss": 1.2247614860534668,
      "step": 158
    },
    {
      "epoch": 0.6374501992031872,
      "grad_norm": 0.4775502681732178,
      "learning_rate": 1.9435606848073003e-05,
      "loss": 1.0775845050811768,
      "step": 160
    },
    {
      "epoch": 0.6454183266932271,
      "grad_norm": 0.26653873920440674,
      "learning_rate": 1.9414741474631586e-05,
      "loss": 0.8565905690193176,
      "step": 162
    },
    {
      "epoch": 0.6533864541832669,
      "grad_norm": 0.22560259699821472,
      "learning_rate": 1.939351032693048e-05,
      "loss": 1.2054938077926636,
      "step": 164
    },
    {
      "epoch": 0.6613545816733067,
      "grad_norm": 1.6585121154785156,
      "learning_rate": 1.9371914327850958e-05,
      "loss": 0.9322201013565063,
      "step": 166
    },
    {
      "epoch": 0.6693227091633466,
      "grad_norm": 0.2651222050189972,
      "learning_rate": 1.934995441613376e-05,
      "loss": 1.2844748497009277,
      "step": 168
    },
    {
      "epoch": 0.6772908366533864,
      "grad_norm": 1.817061185836792,
      "learning_rate": 1.9327631546338263e-05,
      "loss": 1.0478943586349487,
      "step": 170
    },
    {
      "epoch": 0.6852589641434262,
      "grad_norm": 0.36144477128982544,
      "learning_rate": 1.9304946688801014e-05,
      "loss": 0.9816558957099915,
      "step": 172
    },
    {
      "epoch": 0.6932270916334662,
      "grad_norm": 0.4334961771965027,
      "learning_rate": 1.9281900829593544e-05,
      "loss": 1.272782564163208,
      "step": 174
    },
    {
      "epoch": 0.701195219123506,
      "grad_norm": 1.2739726305007935,
      "learning_rate": 1.9258494970479494e-05,
      "loss": 0.8628472089767456,
      "step": 176
    },
    {
      "epoch": 0.7091633466135459,
      "grad_norm": 0.8010619878768921,
      "learning_rate": 1.923473012887109e-05,
      "loss": 0.9980822205543518,
      "step": 178
    },
    {
      "epoch": 0.7171314741035857,
      "grad_norm": 0.6696943640708923,
      "learning_rate": 1.9210607337784885e-05,
      "loss": 1.1843255758285522,
      "step": 180
    },
    {
      "epoch": 0.7250996015936255,
      "grad_norm": 0.20754286646842957,
      "learning_rate": 1.9186127645796902e-05,
      "loss": 0.5432711839675903,
      "step": 182
    },
    {
      "epoch": 0.7330677290836654,
      "grad_norm": 0.3109559714794159,
      "learning_rate": 1.916129211699701e-05,
      "loss": 1.4254323244094849,
      "step": 184
    },
    {
      "epoch": 0.7410358565737052,
      "grad_norm": 0.27430951595306396,
      "learning_rate": 1.91361018309427e-05,
      "loss": 1.3617608547210693,
      "step": 186
    },
    {
      "epoch": 0.749003984063745,
      "grad_norm": 0.2770460247993469,
      "learning_rate": 1.911055788261214e-05,
      "loss": 1.1949325799942017,
      "step": 188
    },
    {
      "epoch": 0.7569721115537849,
      "grad_norm": 0.5085469484329224,
      "learning_rate": 1.9084661382356592e-05,
      "loss": 0.9465351104736328,
      "step": 190
    },
    {
      "epoch": 0.7649402390438247,
      "grad_norm": 0.6416473984718323,
      "learning_rate": 1.9058413455852142e-05,
      "loss": 1.2965277433395386,
      "step": 192
    },
    {
      "epoch": 0.7729083665338645,
      "grad_norm": 0.3408448398113251,
      "learning_rate": 1.9031815244050752e-05,
      "loss": 0.9588617086410522,
      "step": 194
    },
    {
      "epoch": 0.7808764940239044,
      "grad_norm": 0.5440645217895508,
      "learning_rate": 1.9004867903130694e-05,
      "loss": 0.7339329719543457,
      "step": 196
    },
    {
      "epoch": 0.7888446215139442,
      "grad_norm": 1.9453508853912354,
      "learning_rate": 1.897757260444628e-05,
      "loss": 0.825846254825592,
      "step": 198
    },
    {
      "epoch": 0.796812749003984,
      "grad_norm": 0.4652504026889801,
      "learning_rate": 1.8949930534476938e-05,
      "loss": 1.123836636543274,
      "step": 200
    },
    {
      "epoch": 0.8047808764940239,
      "grad_norm": 0.2496296465396881,
      "learning_rate": 1.892194289477565e-05,
      "loss": 1.1040383577346802,
      "step": 202
    },
    {
      "epoch": 0.8127490039840638,
      "grad_norm": 0.4298226237297058,
      "learning_rate": 1.8893610901916707e-05,
      "loss": 1.6137018203735352,
      "step": 204
    },
    {
      "epoch": 0.8207171314741036,
      "grad_norm": 0.4001586437225342,
      "learning_rate": 1.886493578744284e-05,
      "loss": 1.3078550100326538,
      "step": 206
    },
    {
      "epoch": 0.8286852589641435,
      "grad_norm": 0.8008004426956177,
      "learning_rate": 1.8835918797811693e-05,
      "loss": 0.6984291672706604,
      "step": 208
    },
    {
      "epoch": 0.8366533864541833,
      "grad_norm": 0.42881664633750916,
      "learning_rate": 1.880656119434162e-05,
      "loss": 1.335817813873291,
      "step": 210
    },
    {
      "epoch": 0.8446215139442231,
      "grad_norm": 1.1458430290222168,
      "learning_rate": 1.8776864253156875e-05,
      "loss": 1.2731460332870483,
      "step": 212
    },
    {
      "epoch": 0.852589641434263,
      "grad_norm": 0.5224696397781372,
      "learning_rate": 1.874682926513213e-05,
      "loss": 0.9890910387039185,
      "step": 214
    },
    {
      "epoch": 0.8605577689243028,
      "grad_norm": 0.8723546862602234,
      "learning_rate": 1.871645753583638e-05,
      "loss": 1.504272222518921,
      "step": 216
    },
    {
      "epoch": 0.8685258964143426,
      "grad_norm": 0.5456506609916687,
      "learning_rate": 1.8685750385476166e-05,
      "loss": 1.0292469263076782,
      "step": 218
    },
    {
      "epoch": 0.8764940239043825,
      "grad_norm": 0.30583837628364563,
      "learning_rate": 1.8654709148838218e-05,
      "loss": 1.2776914834976196,
      "step": 220
    },
    {
      "epoch": 0.8844621513944223,
      "grad_norm": 0.5236691236495972,
      "learning_rate": 1.8623335175231402e-05,
      "loss": 1.5422356128692627,
      "step": 222
    },
    {
      "epoch": 0.8924302788844621,
      "grad_norm": 0.6241909861564636,
      "learning_rate": 1.85916298284281e-05,
      "loss": 0.7620981335639954,
      "step": 224
    },
    {
      "epoch": 0.900398406374502,
      "grad_norm": 0.48723360896110535,
      "learning_rate": 1.8559594486604905e-05,
      "loss": 1.0757262706756592,
      "step": 226
    },
    {
      "epoch": 0.9083665338645418,
      "grad_norm": 1.2864985466003418,
      "learning_rate": 1.8527230542282724e-05,
      "loss": 1.1468511819839478,
      "step": 228
    },
    {
      "epoch": 0.9163346613545816,
      "grad_norm": 1.7492055892944336,
      "learning_rate": 1.849453940226625e-05,
      "loss": 1.213545799255371,
      "step": 230
    },
    {
      "epoch": 0.9243027888446215,
      "grad_norm": 0.40618953108787537,
      "learning_rate": 1.8461522487582803e-05,
      "loss": 1.3688737154006958,
      "step": 232
    },
    {
      "epoch": 0.9322709163346613,
      "grad_norm": 0.36780276894569397,
      "learning_rate": 1.8428181233420565e-05,
      "loss": 0.8684775233268738,
      "step": 234
    },
    {
      "epoch": 0.9402390438247012,
      "grad_norm": 0.3596510589122772,
      "learning_rate": 1.8394517089066197e-05,
      "loss": 0.8614349365234375,
      "step": 236
    },
    {
      "epoch": 0.9482071713147411,
      "grad_norm": 1.7488377094268799,
      "learning_rate": 1.8360531517841832e-05,
      "loss": 0.5599083304405212,
      "step": 238
    },
    {
      "epoch": 0.9561752988047809,
      "grad_norm": 0.9555627703666687,
      "learning_rate": 1.8326225997041482e-05,
      "loss": 0.9110929369926453,
      "step": 240
    },
    {
      "epoch": 0.9641434262948207,
      "grad_norm": 0.3523239493370056,
      "learning_rate": 1.8291602017866806e-05,
      "loss": 1.1641830205917358,
      "step": 242
    },
    {
      "epoch": 0.9721115537848606,
      "grad_norm": 0.43567129969596863,
      "learning_rate": 1.8256661085362308e-05,
      "loss": 1.2559393644332886,
      "step": 244
    },
    {
      "epoch": 0.9800796812749004,
      "grad_norm": 0.40791869163513184,
      "learning_rate": 1.8221404718349888e-05,
      "loss": 1.0120309591293335,
      "step": 246
    },
    {
      "epoch": 0.9880478087649402,
      "grad_norm": 0.24796508252620697,
      "learning_rate": 1.8185834449362855e-05,
      "loss": 1.0682169198989868,
      "step": 248
    },
    {
      "epoch": 0.9960159362549801,
      "grad_norm": 0.41540786623954773,
      "learning_rate": 1.8149951824579283e-05,
      "loss": 1.0582711696624756,
      "step": 250
    },
    {
      "epoch": 1.00398406374502,
      "grad_norm": 1.0092355012893677,
      "learning_rate": 1.8113758403754823e-05,
      "loss": 0.979528546333313,
      "step": 252
    },
    {
      "epoch": 1.0119521912350598,
      "grad_norm": 0.6038091778755188,
      "learning_rate": 1.8077255760154883e-05,
      "loss": 0.7742294073104858,
      "step": 254
    },
    {
      "epoch": 1.0199203187250996,
      "grad_norm": 0.3472102880477905,
      "learning_rate": 1.8040445480486254e-05,
      "loss": 1.1254103183746338,
      "step": 256
    },
    {
      "epoch": 1.0278884462151394,
      "grad_norm": 0.38630589842796326,
      "learning_rate": 1.8003329164828133e-05,
      "loss": 1.1129672527313232,
      "step": 258
    },
    {
      "epoch": 1.0358565737051793,
      "grad_norm": 0.4505392014980316,
      "learning_rate": 1.7965908426562582e-05,
      "loss": 1.004437804222107,
      "step": 260
    },
    {
      "epoch": 1.043824701195219,
      "grad_norm": 0.3152492046356201,
      "learning_rate": 1.7928184892304366e-05,
      "loss": 0.9400702118873596,
      "step": 262
    },
    {
      "epoch": 1.051792828685259,
      "grad_norm": 0.6728148460388184,
      "learning_rate": 1.789016020183029e-05,
      "loss": 1.0473450422286987,
      "step": 264
    },
    {
      "epoch": 1.0597609561752988,
      "grad_norm": 0.7942842245101929,
      "learning_rate": 1.7851836008007883e-05,
      "loss": 0.7050523161888123,
      "step": 266
    },
    {
      "epoch": 1.0677290836653386,
      "grad_norm": 0.24827615916728973,
      "learning_rate": 1.781321397672358e-05,
      "loss": 0.8870258927345276,
      "step": 268
    },
    {
      "epoch": 1.0756972111553784,
      "grad_norm": 0.3862256109714508,
      "learning_rate": 1.777429578681029e-05,
      "loss": 0.5399430394172668,
      "step": 270
    },
    {
      "epoch": 1.0836653386454183,
      "grad_norm": 0.3700801134109497,
      "learning_rate": 1.7735083129974423e-05,
      "loss": 1.2206511497497559,
      "step": 272
    },
    {
      "epoch": 1.091633466135458,
      "grad_norm": 0.7308570742607117,
      "learning_rate": 1.769557771072236e-05,
      "loss": 0.680720865726471,
      "step": 274
    },
    {
      "epoch": 1.099601593625498,
      "grad_norm": 0.2208869606256485,
      "learning_rate": 1.7655781246286345e-05,
      "loss": 0.8195149302482605,
      "step": 276
    },
    {
      "epoch": 1.1075697211155378,
      "grad_norm": 0.31552159786224365,
      "learning_rate": 1.761569546654988e-05,
      "loss": 0.7425749897956848,
      "step": 278
    },
    {
      "epoch": 1.1155378486055776,
      "grad_norm": 0.5200814008712769,
      "learning_rate": 1.7575322113972472e-05,
      "loss": 0.9490090608596802,
      "step": 280
    },
    {
      "epoch": 1.1235059760956174,
      "grad_norm": 0.3083236813545227,
      "learning_rate": 1.753466294351395e-05,
      "loss": 1.076003074645996,
      "step": 282
    },
    {
      "epoch": 1.1314741035856573,
      "grad_norm": 0.36041611433029175,
      "learning_rate": 1.7493719722558135e-05,
      "loss": 1.1486386060714722,
      "step": 284
    },
    {
      "epoch": 1.139442231075697,
      "grad_norm": 1.6613099575042725,
      "learning_rate": 1.7452494230836034e-05,
      "loss": 0.3132348954677582,
      "step": 286
    },
    {
      "epoch": 1.1474103585657371,
      "grad_norm": 0.2219257354736328,
      "learning_rate": 1.741098826034848e-05,
      "loss": 0.6475589275360107,
      "step": 288
    },
    {
      "epoch": 1.155378486055777,
      "grad_norm": 0.9050571918487549,
      "learning_rate": 1.7369203615288227e-05,
      "loss": 0.9298643469810486,
      "step": 290
    },
    {
      "epoch": 1.1633466135458168,
      "grad_norm": 0.651750922203064,
      "learning_rate": 1.7327142111961537e-05,
      "loss": 0.6063528656959534,
      "step": 292
    },
    {
      "epoch": 1.1713147410358566,
      "grad_norm": 0.19915056228637695,
      "learning_rate": 1.7284805578709218e-05,
      "loss": 1.1846861839294434,
      "step": 294
    },
    {
      "epoch": 1.1792828685258965,
      "grad_norm": 0.21233604848384857,
      "learning_rate": 1.724219585582716e-05,
      "loss": 1.264991044998169,
      "step": 296
    },
    {
      "epoch": 1.1872509960159363,
      "grad_norm": 0.27689361572265625,
      "learning_rate": 1.7199314795486325e-05,
      "loss": 1.2833882570266724,
      "step": 298
    },
    {
      "epoch": 1.1952191235059761,
      "grad_norm": 0.454823762178421,
      "learning_rate": 1.7156164261652247e-05,
      "loss": 0.58225017786026,
      "step": 300
    },
    {
      "epoch": 1.203187250996016,
      "grad_norm": 0.2573089003562927,
      "learning_rate": 1.711274613000401e-05,
      "loss": 0.9149743318557739,
      "step": 302
    },
    {
      "epoch": 1.2111553784860558,
      "grad_norm": 0.26930296421051025,
      "learning_rate": 1.7069062287852717e-05,
      "loss": 0.5347046852111816,
      "step": 304
    },
    {
      "epoch": 1.2191235059760956,
      "grad_norm": 0.22766169905662537,
      "learning_rate": 1.702511463405943e-05,
      "loss": 1.1858634948730469,
      "step": 306
    },
    {
      "epoch": 1.2270916334661355,
      "grad_norm": 1.8281174898147583,
      "learning_rate": 1.698090507895266e-05,
      "loss": 0.923149824142456,
      "step": 308
    },
    {
      "epoch": 1.2350597609561753,
      "grad_norm": 0.3246043622493744,
      "learning_rate": 1.693643554424532e-05,
      "loss": 1.1471027135849,
      "step": 310
    },
    {
      "epoch": 1.2430278884462151,
      "grad_norm": 0.5812107920646667,
      "learning_rate": 1.6891707962951173e-05,
      "loss": 1.0638307332992554,
      "step": 312
    },
    {
      "epoch": 1.250996015936255,
      "grad_norm": 0.2739630341529846,
      "learning_rate": 1.684672427930083e-05,
      "loss": 0.8980826735496521,
      "step": 314
    },
    {
      "epoch": 1.2589641434262948,
      "grad_norm": 0.31895291805267334,
      "learning_rate": 1.680148644865722e-05,
      "loss": 0.7298654317855835,
      "step": 316
    },
    {
      "epoch": 1.2669322709163346,
      "grad_norm": 0.2569485604763031,
      "learning_rate": 1.6755996437430622e-05,
      "loss": 0.6580713987350464,
      "step": 318
    },
    {
      "epoch": 1.2749003984063745,
      "grad_norm": 0.3961490988731384,
      "learning_rate": 1.6710256222993137e-05,
      "loss": 0.489969938993454,
      "step": 320
    },
    {
      "epoch": 1.2828685258964143,
      "grad_norm": 0.30034857988357544,
      "learning_rate": 1.6664267793592797e-05,
      "loss": 0.9578895568847656,
      "step": 322
    },
    {
      "epoch": 1.2908366533864541,
      "grad_norm": 0.36320650577545166,
      "learning_rate": 1.661803314826709e-05,
      "loss": 0.9098944067955017,
      "step": 324
    },
    {
      "epoch": 1.298804780876494,
      "grad_norm": 0.37193965911865234,
      "learning_rate": 1.65715542967561e-05,
      "loss": 1.2378405332565308,
      "step": 326
    },
    {
      "epoch": 1.3067729083665338,
      "grad_norm": 0.24212397634983063,
      "learning_rate": 1.652483325941511e-05,
      "loss": 0.6748937368392944,
      "step": 328
    },
    {
      "epoch": 1.3147410358565736,
      "grad_norm": 0.23975537717342377,
      "learning_rate": 1.647787206712683e-05,
      "loss": 0.7632936239242554,
      "step": 330
    },
    {
      "epoch": 1.3227091633466135,
      "grad_norm": 0.5843632221221924,
      "learning_rate": 1.6430672761213065e-05,
      "loss": 0.9058688282966614,
      "step": 332
    },
    {
      "epoch": 1.3306772908366533,
      "grad_norm": 1.3426114320755005,
      "learning_rate": 1.6383237393346025e-05,
      "loss": 0.6352471113204956,
      "step": 334
    },
    {
      "epoch": 1.3386454183266931,
      "grad_norm": 0.40683165192604065,
      "learning_rate": 1.633556802545911e-05,
      "loss": 0.8502663373947144,
      "step": 336
    },
    {
      "epoch": 1.3466135458167332,
      "grad_norm": 0.37066444754600525,
      "learning_rate": 1.628766672965731e-05,
      "loss": 1.084623098373413,
      "step": 338
    },
    {
      "epoch": 1.354581673306773,
      "grad_norm": 0.19215187430381775,
      "learning_rate": 1.6239535588127107e-05,
      "loss": 1.0811511278152466,
      "step": 340
    },
    {
      "epoch": 1.3625498007968129,
      "grad_norm": 0.20627997815608978,
      "learning_rate": 1.6191176693046e-05,
      "loss": 1.156653642654419,
      "step": 342
    },
    {
      "epoch": 1.3705179282868527,
      "grad_norm": 0.24132803082466125,
      "learning_rate": 1.6142592146491517e-05,
      "loss": 1.0989466905593872,
      "step": 344
    },
    {
      "epoch": 1.3784860557768925,
      "grad_norm": 1.311899185180664,
      "learning_rate": 1.6093784060349876e-05,
      "loss": 0.7021924257278442,
      "step": 346
    },
    {
      "epoch": 1.3864541832669324,
      "grad_norm": 0.2994781732559204,
      "learning_rate": 1.6044754556224178e-05,
      "loss": 1.1349058151245117,
      "step": 348
    },
    {
      "epoch": 1.3944223107569722,
      "grad_norm": 0.7709036469459534,
      "learning_rate": 1.5995505765342176e-05,
      "loss": 0.6554039716720581,
      "step": 350
    },
    {
      "epoch": 1.402390438247012,
      "grad_norm": 0.2403658926486969,
      "learning_rate": 1.594603982846364e-05,
      "loss": 1.129141926765442,
      "step": 352
    },
    {
      "epoch": 1.4103585657370519,
      "grad_norm": 0.38826340436935425,
      "learning_rate": 1.5896358895787304e-05,
      "loss": 1.1050136089324951,
      "step": 354
    },
    {
      "epoch": 1.4183266932270917,
      "grad_norm": 0.3303016722202301,
      "learning_rate": 1.5846465126857386e-05,
      "loss": 0.8948855400085449,
      "step": 356
    },
    {
      "epoch": 1.4262948207171315,
      "grad_norm": 0.25158822536468506,
      "learning_rate": 1.579636069046975e-05,
      "loss": 1.1352990865707397,
      "step": 358
    },
    {
      "epoch": 1.4342629482071714,
      "grad_norm": 0.3274855613708496,
      "learning_rate": 1.5746047764577586e-05,
      "loss": 0.5751365423202515,
      "step": 360
    },
    {
      "epoch": 1.4422310756972112,
      "grad_norm": 0.1873192936182022,
      "learning_rate": 1.5695528536196774e-05,
      "loss": 0.654511034488678,
      "step": 362
    },
    {
      "epoch": 1.450199203187251,
      "grad_norm": 0.2384817749261856,
      "learning_rate": 1.564480520131081e-05,
      "loss": 1.0301330089569092,
      "step": 364
    },
    {
      "epoch": 1.4581673306772909,
      "grad_norm": 0.18577991425991058,
      "learning_rate": 1.559387996477534e-05,
      "loss": 1.0699549913406372,
      "step": 366
    },
    {
      "epoch": 1.4661354581673307,
      "grad_norm": 0.20976194739341736,
      "learning_rate": 1.5542755040222325e-05,
      "loss": 0.5935020446777344,
      "step": 368
    },
    {
      "epoch": 1.4741035856573705,
      "grad_norm": 0.6415178179740906,
      "learning_rate": 1.5491432649963823e-05,
      "loss": 0.6804059147834778,
      "step": 370
    },
    {
      "epoch": 1.4820717131474104,
      "grad_norm": 0.6343056559562683,
      "learning_rate": 1.543991502489538e-05,
      "loss": 1.2095876932144165,
      "step": 372
    },
    {
      "epoch": 1.4900398406374502,
      "grad_norm": 0.19021844863891602,
      "learning_rate": 1.5388204404399073e-05,
      "loss": 1.059179663658142,
      "step": 374
    },
    {
      "epoch": 1.49800796812749,
      "grad_norm": 0.1728297770023346,
      "learning_rate": 1.533630303624614e-05,
      "loss": 1.0693382024765015,
      "step": 376
    },
    {
      "epoch": 1.5059760956175299,
      "grad_norm": 0.376954585313797,
      "learning_rate": 1.5284213176499302e-05,
      "loss": 0.8699037432670593,
      "step": 378
    },
    {
      "epoch": 1.5139442231075697,
      "grad_norm": 0.22751152515411377,
      "learning_rate": 1.5231937089414675e-05,
      "loss": 1.1054149866104126,
      "step": 380
    },
    {
      "epoch": 1.5219123505976095,
      "grad_norm": 0.23990006744861603,
      "learning_rate": 1.517947704734337e-05,
      "loss": 0.6814401745796204,
      "step": 382
    },
    {
      "epoch": 1.5298804780876494,
      "grad_norm": 0.4457078278064728,
      "learning_rate": 1.5126835330632677e-05,
      "loss": 0.7228714227676392,
      "step": 384
    },
    {
      "epoch": 1.5378486055776892,
      "grad_norm": 0.40967053174972534,
      "learning_rate": 1.5074014227526997e-05,
      "loss": 0.9597483277320862,
      "step": 386
    },
    {
      "epoch": 1.545816733067729,
      "grad_norm": 0.31324154138565063,
      "learning_rate": 1.502101603406833e-05,
      "loss": 1.1154417991638184,
      "step": 388
    },
    {
      "epoch": 1.5537848605577689,
      "grad_norm": 0.9542575478553772,
      "learning_rate": 1.4967843053996494e-05,
      "loss": 0.5699301362037659,
      "step": 390
    },
    {
      "epoch": 1.5617529880478087,
      "grad_norm": 0.31226351857185364,
      "learning_rate": 1.4914497598648973e-05,
      "loss": 0.6108444333076477,
      "step": 392
    },
    {
      "epoch": 1.5697211155378485,
      "grad_norm": 6.123598098754883,
      "learning_rate": 1.4860981986860464e-05,
      "loss": 1.0147069692611694,
      "step": 394
    },
    {
      "epoch": 1.5776892430278884,
      "grad_norm": 0.5377170443534851,
      "learning_rate": 1.4807298544862054e-05,
      "loss": 1.0180771350860596,
      "step": 396
    },
    {
      "epoch": 1.5856573705179282,
      "grad_norm": 0.20692817866802216,
      "learning_rate": 1.4753449606180137e-05,
      "loss": 1.1011747121810913,
      "step": 398
    },
    {
      "epoch": 1.593625498007968,
      "grad_norm": 0.4750509560108185,
      "learning_rate": 1.4699437511534947e-05,
      "loss": 0.5985136032104492,
      "step": 400
    },
    {
      "epoch": 1.6015936254980079,
      "grad_norm": 0.3046919107437134,
      "learning_rate": 1.4645264608738832e-05,
      "loss": 1.0762031078338623,
      "step": 402
    },
    {
      "epoch": 1.6095617529880477,
      "grad_norm": 0.22464171051979065,
      "learning_rate": 1.4590933252594193e-05,
      "loss": 1.2067527770996094,
      "step": 404
    },
    {
      "epoch": 1.6175298804780875,
      "grad_norm": 2.2710113525390625,
      "learning_rate": 1.4536445804791125e-05,
      "loss": 0.7306970357894897,
      "step": 406
    },
    {
      "epoch": 1.6254980079681274,
      "grad_norm": 0.1687576174736023,
      "learning_rate": 1.4481804633804746e-05,
      "loss": 0.5554207563400269,
      "step": 408
    },
    {
      "epoch": 1.6334661354581672,
      "grad_norm": 0.39087173342704773,
      "learning_rate": 1.4427012114792269e-05,
      "loss": 0.7205751538276672,
      "step": 410
    },
    {
      "epoch": 1.641434262948207,
      "grad_norm": 0.2729649841785431,
      "learning_rate": 1.4372070629489744e-05,
      "loss": 0.4776345193386078,
      "step": 412
    },
    {
      "epoch": 1.6494023904382469,
      "grad_norm": 0.3640289306640625,
      "learning_rate": 1.4316982566108515e-05,
      "loss": 0.7648314237594604,
      "step": 414
    },
    {
      "epoch": 1.6573705179282867,
      "grad_norm": 0.6474630832672119,
      "learning_rate": 1.4261750319231432e-05,
      "loss": 0.7392333745956421,
      "step": 416
    },
    {
      "epoch": 1.6653386454183265,
      "grad_norm": 0.34522193670272827,
      "learning_rate": 1.4206376289708752e-05,
      "loss": 0.883467435836792,
      "step": 418
    },
    {
      "epoch": 1.6733067729083664,
      "grad_norm": 0.277730256319046,
      "learning_rate": 1.4150862884553776e-05,
      "loss": 1.131731390953064,
      "step": 420
    },
    {
      "epoch": 1.6812749003984062,
      "grad_norm": 0.3468133807182312,
      "learning_rate": 1.4095212516838216e-05,
      "loss": 0.9705685377120972,
      "step": 422
    },
    {
      "epoch": 1.6892430278884463,
      "grad_norm": 0.9739753603935242,
      "learning_rate": 1.4039427605587326e-05,
      "loss": 0.5433371663093567,
      "step": 424
    },
    {
      "epoch": 1.697211155378486,
      "grad_norm": 0.45867398381233215,
      "learning_rate": 1.398351057567472e-05,
      "loss": 1.1449828147888184,
      "step": 426
    },
    {
      "epoch": 1.705179282868526,
      "grad_norm": 0.25316959619522095,
      "learning_rate": 1.392746385771699e-05,
      "loss": 1.084163784980774,
      "step": 428
    },
    {
      "epoch": 1.7131474103585658,
      "grad_norm": 0.25396499037742615,
      "learning_rate": 1.3871289887968034e-05,
      "loss": 0.7046893239021301,
      "step": 430
    },
    {
      "epoch": 1.7211155378486056,
      "grad_norm": 0.6979869604110718,
      "learning_rate": 1.3814991108213173e-05,
      "loss": 0.8699672222137451,
      "step": 432
    },
    {
      "epoch": 1.7290836653386454,
      "grad_norm": 0.632816731929779,
      "learning_rate": 1.3758569965663011e-05,
      "loss": 0.6414164304733276,
      "step": 434
    },
    {
      "epoch": 1.7370517928286853,
      "grad_norm": 0.5996594429016113,
      "learning_rate": 1.370202891284703e-05,
      "loss": 0.4016932547092438,
      "step": 436
    },
    {
      "epoch": 1.745019920318725,
      "grad_norm": 0.374830037355423,
      "learning_rate": 1.3645370407507017e-05,
      "loss": 0.7375532388687134,
      "step": 438
    },
    {
      "epoch": 1.752988047808765,
      "grad_norm": 1.5595433712005615,
      "learning_rate": 1.3588596912490222e-05,
      "loss": 0.7329553365707397,
      "step": 440
    },
    {
      "epoch": 1.7609561752988048,
      "grad_norm": 0.526985228061676,
      "learning_rate": 1.3531710895642288e-05,
      "loss": 1.329801321029663,
      "step": 442
    },
    {
      "epoch": 1.7689243027888446,
      "grad_norm": 0.2958177328109741,
      "learning_rate": 1.3474714829699987e-05,
      "loss": 0.7087810039520264,
      "step": 444
    },
    {
      "epoch": 1.7768924302788844,
      "grad_norm": 0.5098376274108887,
      "learning_rate": 1.341761119218374e-05,
      "loss": 1.0286879539489746,
      "step": 446
    },
    {
      "epoch": 1.7848605577689243,
      "grad_norm": 0.7928891777992249,
      "learning_rate": 1.3360402465289916e-05,
      "loss": 1.0772615671157837,
      "step": 448
    },
    {
      "epoch": 1.792828685258964,
      "grad_norm": 0.26385796070098877,
      "learning_rate": 1.3303091135782942e-05,
      "loss": 0.8347827196121216,
      "step": 450
    },
    {
      "epoch": 1.800796812749004,
      "grad_norm": 0.28425443172454834,
      "learning_rate": 1.3245679694887192e-05,
      "loss": 0.44587215781211853,
      "step": 452
    },
    {
      "epoch": 1.8087649402390438,
      "grad_norm": 0.5673894882202148,
      "learning_rate": 1.318817063817872e-05,
      "loss": 0.9081963896751404,
      "step": 454
    },
    {
      "epoch": 1.8167330677290838,
      "grad_norm": 0.3212035298347473,
      "learning_rate": 1.3130566465476773e-05,
      "loss": 0.6184219717979431,
      "step": 456
    },
    {
      "epoch": 1.8247011952191237,
      "grad_norm": 0.30924293398857117,
      "learning_rate": 1.307286968073511e-05,
      "loss": 1.0215346813201904,
      "step": 458
    },
    {
      "epoch": 1.8326693227091635,
      "grad_norm": 0.28430241346359253,
      "learning_rate": 1.3015082791933182e-05,
      "loss": 0.7581181526184082,
      "step": 460
    },
    {
      "epoch": 1.8406374501992033,
      "grad_norm": 3.09944748878479,
      "learning_rate": 1.295720831096712e-05,
      "loss": 0.8157633543014526,
      "step": 462
    },
    {
      "epoch": 1.8486055776892432,
      "grad_norm": 0.1777675449848175,
      "learning_rate": 1.2899248753540518e-05,
      "loss": 0.13458193838596344,
      "step": 464
    },
    {
      "epoch": 1.856573705179283,
      "grad_norm": 0.6683889627456665,
      "learning_rate": 1.2841206639055108e-05,
      "loss": 1.1309783458709717,
      "step": 466
    },
    {
      "epoch": 1.8645418326693228,
      "grad_norm": 0.7414482831954956,
      "learning_rate": 1.2783084490501229e-05,
      "loss": 0.5139486193656921,
      "step": 468
    },
    {
      "epoch": 1.8725099601593627,
      "grad_norm": 0.4845532774925232,
      "learning_rate": 1.2724884834348163e-05,
      "loss": 1.1115937232971191,
      "step": 470
    },
    {
      "epoch": 1.8804780876494025,
      "grad_norm": 0.42305639386177063,
      "learning_rate": 1.266661020043432e-05,
      "loss": 0.9607354402542114,
      "step": 472
    },
    {
      "epoch": 1.8884462151394423,
      "grad_norm": 0.9543145895004272,
      "learning_rate": 1.2608263121857244e-05,
      "loss": 0.9997490048408508,
      "step": 474
    },
    {
      "epoch": 1.8964143426294822,
      "grad_norm": 0.227760910987854,
      "learning_rate": 1.254984613486355e-05,
      "loss": 0.958122730255127,
      "step": 476
    },
    {
      "epoch": 1.904382470119522,
      "grad_norm": 0.20254026353359222,
      "learning_rate": 1.2491361778738631e-05,
      "loss": 0.8319299817085266,
      "step": 478
    },
    {
      "epoch": 1.9123505976095618,
      "grad_norm": 1.022902250289917,
      "learning_rate": 1.243281259569631e-05,
      "loss": 0.5840953588485718,
      "step": 480
    },
    {
      "epoch": 1.9203187250996017,
      "grad_norm": 0.37438637018203735,
      "learning_rate": 1.2374201130768325e-05,
      "loss": 0.771386444568634,
      "step": 482
    },
    {
      "epoch": 1.9282868525896415,
      "grad_norm": 0.2661597728729248,
      "learning_rate": 1.2315529931693682e-05,
      "loss": 1.074930191040039,
      "step": 484
    },
    {
      "epoch": 1.9362549800796813,
      "grad_norm": 0.5989009141921997,
      "learning_rate": 1.2256801548807948e-05,
      "loss": 1.0152567625045776,
      "step": 486
    },
    {
      "epoch": 1.9442231075697212,
      "grad_norm": 0.23072665929794312,
      "learning_rate": 1.2198018534932367e-05,
      "loss": 1.0628975629806519,
      "step": 488
    },
    {
      "epoch": 1.952191235059761,
      "grad_norm": 0.281194806098938,
      "learning_rate": 1.2139183445262892e-05,
      "loss": 1.063922643661499,
      "step": 490
    },
    {
      "epoch": 1.9601593625498008,
      "grad_norm": 0.2512539327144623,
      "learning_rate": 1.2080298837259124e-05,
      "loss": 0.7215337157249451,
      "step": 492
    },
    {
      "epoch": 1.9681274900398407,
      "grad_norm": 1.4636787176132202,
      "learning_rate": 1.2021367270533138e-05,
      "loss": 0.753463625907898,
      "step": 494
    },
    {
      "epoch": 1.9760956175298805,
      "grad_norm": 0.20480957627296448,
      "learning_rate": 1.1962391306738226e-05,
      "loss": 1.0385065078735352,
      "step": 496
    },
    {
      "epoch": 1.9840637450199203,
      "grad_norm": 3.6724064350128174,
      "learning_rate": 1.190337350945754e-05,
      "loss": 0.8450489044189453,
      "step": 498
    },
    {
      "epoch": 1.9920318725099602,
      "grad_norm": 0.24277392029762268,
      "learning_rate": 1.1844316444092667e-05,
      "loss": 1.1586743593215942,
      "step": 500
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4689848721027374,
      "learning_rate": 1.1785222677752105e-05,
      "loss": 0.6194124221801758,
      "step": 502
    },
    {
      "epoch": 2.00796812749004,
      "grad_norm": 0.19164475798606873,
      "learning_rate": 1.1726094779139685e-05,
      "loss": 0.5169786810874939,
      "step": 504
    },
    {
      "epoch": 2.0159362549800797,
      "grad_norm": 0.3737673759460449,
      "learning_rate": 1.1666935318442905e-05,
      "loss": 0.4465612769126892,
      "step": 506
    },
    {
      "epoch": 2.0239043824701195,
      "grad_norm": 0.059804949909448624,
      "learning_rate": 1.1607746867221208e-05,
      "loss": 0.4815758168697357,
      "step": 508
    },
    {
      "epoch": 2.0318725099601593,
      "grad_norm": 0.19923484325408936,
      "learning_rate": 1.1548531998294217e-05,
      "loss": 0.7441997528076172,
      "step": 510
    },
    {
      "epoch": 2.039840637450199,
      "grad_norm": 0.39438050985336304,
      "learning_rate": 1.148929328562987e-05,
      "loss": 0.6462599635124207,
      "step": 512
    },
    {
      "epoch": 2.047808764940239,
      "grad_norm": 0.6492677927017212,
      "learning_rate": 1.1430033304232569e-05,
      "loss": 0.3487642705440521,
      "step": 514
    },
    {
      "epoch": 2.055776892430279,
      "grad_norm": 0.21399064362049103,
      "learning_rate": 1.137075463003122e-05,
      "loss": 0.7897251844406128,
      "step": 516
    },
    {
      "epoch": 2.0637450199203187,
      "grad_norm": 0.45779934525489807,
      "learning_rate": 1.1311459839767278e-05,
      "loss": 0.47086820006370544,
      "step": 518
    },
    {
      "epoch": 2.0717131474103585,
      "grad_norm": 0.45006412267684937,
      "learning_rate": 1.1252151510882731e-05,
      "loss": 0.6521967053413391,
      "step": 520
    },
    {
      "epoch": 2.0796812749003983,
      "grad_norm": 0.5039223432540894,
      "learning_rate": 1.1192832221408068e-05,
      "loss": 0.8312692046165466,
      "step": 522
    },
    {
      "epoch": 2.087649402390438,
      "grad_norm": 0.5422983765602112,
      "learning_rate": 1.1133504549850218e-05,
      "loss": 0.744563639163971,
      "step": 524
    },
    {
      "epoch": 2.095617529880478,
      "grad_norm": 0.4640710949897766,
      "learning_rate": 1.1074171075080469e-05,
      "loss": 0.3822844624519348,
      "step": 526
    },
    {
      "epoch": 2.103585657370518,
      "grad_norm": 0.2076382040977478,
      "learning_rate": 1.1014834376222351e-05,
      "loss": 0.6223887205123901,
      "step": 528
    },
    {
      "epoch": 2.1115537848605577,
      "grad_norm": 0.12689504027366638,
      "learning_rate": 1.0955497032539557e-05,
      "loss": 0.47389209270477295,
      "step": 530
    },
    {
      "epoch": 2.1195219123505975,
      "grad_norm": 0.184941366314888,
      "learning_rate": 1.0896161623323795e-05,
      "loss": 0.22031724452972412,
      "step": 532
    },
    {
      "epoch": 2.1274900398406373,
      "grad_norm": 0.3087546229362488,
      "learning_rate": 1.0836830727782692e-05,
      "loss": 0.6414751410484314,
      "step": 534
    },
    {
      "epoch": 2.135458167330677,
      "grad_norm": 0.6677108407020569,
      "learning_rate": 1.0777506924927667e-05,
      "loss": 0.596595823764801,
      "step": 536
    },
    {
      "epoch": 2.143426294820717,
      "grad_norm": 0.6424462795257568,
      "learning_rate": 1.0718192793461849e-05,
      "loss": 0.6871726512908936,
      "step": 538
    },
    {
      "epoch": 2.151394422310757,
      "grad_norm": 0.19472549855709076,
      "learning_rate": 1.065889091166795e-05,
      "loss": 0.7834715247154236,
      "step": 540
    },
    {
      "epoch": 2.1593625498007967,
      "grad_norm": 0.26970356702804565,
      "learning_rate": 1.059960385729621e-05,
      "loss": 0.7224296927452087,
      "step": 542
    },
    {
      "epoch": 2.1673306772908365,
      "grad_norm": 0.19753654301166534,
      "learning_rate": 1.0540334207452363e-05,
      "loss": 0.5219651460647583,
      "step": 544
    },
    {
      "epoch": 2.1752988047808763,
      "grad_norm": 0.23047387599945068,
      "learning_rate": 1.0481084538485589e-05,
      "loss": 0.5418217182159424,
      "step": 546
    },
    {
      "epoch": 2.183266932270916,
      "grad_norm": 0.3223879635334015,
      "learning_rate": 1.0421857425876524e-05,
      "loss": 0.802264928817749,
      "step": 548
    },
    {
      "epoch": 2.191235059760956,
      "grad_norm": 0.27228158712387085,
      "learning_rate": 1.0362655444125334e-05,
      "loss": 0.8337580561637878,
      "step": 550
    },
    {
      "epoch": 2.199203187250996,
      "grad_norm": 0.37938058376312256,
      "learning_rate": 1.0303481166639784e-05,
      "loss": 0.5192627310752869,
      "step": 552
    },
    {
      "epoch": 2.2071713147410357,
      "grad_norm": 0.5020936727523804,
      "learning_rate": 1.0244337165623378e-05,
      "loss": 0.4618823230266571,
      "step": 554
    },
    {
      "epoch": 2.2151394422310755,
      "grad_norm": 0.2157645970582962,
      "learning_rate": 1.0185226011963563e-05,
      "loss": 0.6217008233070374,
      "step": 556
    },
    {
      "epoch": 2.2231075697211153,
      "grad_norm": 0.4750109612941742,
      "learning_rate": 1.0126150275119958e-05,
      "loss": 0.7997477054595947,
      "step": 558
    },
    {
      "epoch": 2.231075697211155,
      "grad_norm": 0.24116151034832,
      "learning_rate": 1.0067112523012686e-05,
      "loss": 0.7890470027923584,
      "step": 560
    },
    {
      "epoch": 2.239043824701195,
      "grad_norm": 0.3130785822868347,
      "learning_rate": 1.0008115321910729e-05,
      "loss": 0.9172634482383728,
      "step": 562
    },
    {
      "epoch": 2.247011952191235,
      "grad_norm": 0.43472668528556824,
      "learning_rate": 9.949161236320388e-06,
      "loss": 0.6723137497901917,
      "step": 564
    },
    {
      "epoch": 2.2549800796812747,
      "grad_norm": 0.491053968667984,
      "learning_rate": 9.890252828873816e-06,
      "loss": 0.4990261197090149,
      "step": 566
    },
    {
      "epoch": 2.2629482071713145,
      "grad_norm": 0.19463039934635162,
      "learning_rate": 9.831392660217608e-06,
      "loss": 0.6660031080245972,
      "step": 568
    },
    {
      "epoch": 2.2709163346613543,
      "grad_norm": 0.1894519180059433,
      "learning_rate": 9.772583288901502e-06,
      "loss": 0.804633617401123,
      "step": 570
    },
    {
      "epoch": 2.278884462151394,
      "grad_norm": 0.6049039363861084,
      "learning_rate": 9.713827271267169e-06,
      "loss": 0.5522835850715637,
      "step": 572
    },
    {
      "epoch": 2.2868525896414345,
      "grad_norm": 0.21983656287193298,
      "learning_rate": 9.655127161337088e-06,
      "loss": 0.6103823184967041,
      "step": 574
    },
    {
      "epoch": 2.2948207171314743,
      "grad_norm": 0.6579908728599548,
      "learning_rate": 9.59648551070351e-06,
      "loss": 0.24219289422035217,
      "step": 576
    },
    {
      "epoch": 2.302788844621514,
      "grad_norm": 0.5361920595169067,
      "learning_rate": 9.53790486841759e-06,
      "loss": 0.5663895010948181,
      "step": 578
    },
    {
      "epoch": 2.310756972111554,
      "grad_norm": 0.3519666790962219,
      "learning_rate": 9.479387780878531e-06,
      "loss": 0.5547711849212646,
      "step": 580
    },
    {
      "epoch": 2.318725099601594,
      "grad_norm": 3.0020523071289062,
      "learning_rate": 9.42093679172294e-06,
      "loss": 0.5223433375358582,
      "step": 582
    },
    {
      "epoch": 2.3266932270916336,
      "grad_norm": 0.20067761838436127,
      "learning_rate": 9.362554441714216e-06,
      "loss": 0.6483029127120972,
      "step": 584
    },
    {
      "epoch": 2.3346613545816735,
      "grad_norm": 0.6246410012245178,
      "learning_rate": 9.304243268632165e-06,
      "loss": 0.5529161691665649,
      "step": 586
    },
    {
      "epoch": 2.3426294820717133,
      "grad_norm": 0.4139845669269562,
      "learning_rate": 9.246005807162632e-06,
      "loss": 0.7177547812461853,
      "step": 588
    },
    {
      "epoch": 2.350597609561753,
      "grad_norm": 0.4397997260093689,
      "learning_rate": 9.187844588787354e-06,
      "loss": 0.8279111981391907,
      "step": 590
    },
    {
      "epoch": 2.358565737051793,
      "grad_norm": 0.22359026968479156,
      "learning_rate": 9.12976214167391e-06,
      "loss": 0.8503004908561707,
      "step": 592
    },
    {
      "epoch": 2.366533864541833,
      "grad_norm": 0.3120749890804291,
      "learning_rate": 9.071760990565832e-06,
      "loss": 0.5646694898605347,
      "step": 594
    },
    {
      "epoch": 2.3745019920318726,
      "grad_norm": 0.1713973730802536,
      "learning_rate": 9.013843656672854e-06,
      "loss": 0.7497953772544861,
      "step": 596
    },
    {
      "epoch": 2.3824701195219125,
      "grad_norm": 0.13825774192810059,
      "learning_rate": 8.956012657561314e-06,
      "loss": 0.5747182965278625,
      "step": 598
    },
    {
      "epoch": 2.3904382470119523,
      "grad_norm": 0.6295970678329468,
      "learning_rate": 8.89827050704474e-06,
      "loss": 0.6441735625267029,
      "step": 600
    },
    {
      "epoch": 2.398406374501992,
      "grad_norm": 0.3059506416320801,
      "learning_rate": 8.840619715074553e-06,
      "loss": 0.8171231746673584,
      "step": 602
    },
    {
      "epoch": 2.406374501992032,
      "grad_norm": 0.416616290807724,
      "learning_rate": 8.78306278763098e-06,
      "loss": 0.49859338998794556,
      "step": 604
    },
    {
      "epoch": 2.414342629482072,
      "grad_norm": 0.6188241243362427,
      "learning_rate": 8.725602226614121e-06,
      "loss": 0.23148350417613983,
      "step": 606
    },
    {
      "epoch": 2.4223107569721116,
      "grad_norm": 0.17639189958572388,
      "learning_rate": 8.668240529735192e-06,
      "loss": 0.3683927357196808,
      "step": 608
    },
    {
      "epoch": 2.4302788844621515,
      "grad_norm": 0.14717182517051697,
      "learning_rate": 8.610980190407958e-06,
      "loss": 0.6629661321640015,
      "step": 610
    },
    {
      "epoch": 2.4382470119521913,
      "grad_norm": 0.472764790058136,
      "learning_rate": 8.55382369764034e-06,
      "loss": 0.6623281836509705,
      "step": 612
    },
    {
      "epoch": 2.446215139442231,
      "grad_norm": 0.28185757994651794,
      "learning_rate": 8.496773535926242e-06,
      "loss": 0.4608023464679718,
      "step": 614
    },
    {
      "epoch": 2.454183266932271,
      "grad_norm": 0.246009960770607,
      "learning_rate": 8.439832185137529e-06,
      "loss": 0.7289039492607117,
      "step": 616
    },
    {
      "epoch": 2.462151394422311,
      "grad_norm": 0.3146935701370239,
      "learning_rate": 8.383002120416241e-06,
      "loss": 0.595883309841156,
      "step": 618
    },
    {
      "epoch": 2.4701195219123506,
      "grad_norm": 0.46129530668258667,
      "learning_rate": 8.326285812067023e-06,
      "loss": 0.6848819851875305,
      "step": 620
    },
    {
      "epoch": 2.4780876494023905,
      "grad_norm": 0.39660993218421936,
      "learning_rate": 8.269685725449708e-06,
      "loss": 0.6775418519973755,
      "step": 622
    },
    {
      "epoch": 2.4860557768924303,
      "grad_norm": 0.4998105764389038,
      "learning_rate": 8.213204320872176e-06,
      "loss": 0.668332040309906,
      "step": 624
    },
    {
      "epoch": 2.49402390438247,
      "grad_norm": 0.3267073333263397,
      "learning_rate": 8.156844053483412e-06,
      "loss": 0.4088161289691925,
      "step": 626
    },
    {
      "epoch": 2.50199203187251,
      "grad_norm": 0.6505297422409058,
      "learning_rate": 8.100607373166773e-06,
      "loss": 0.8275009393692017,
      "step": 628
    },
    {
      "epoch": 2.50996015936255,
      "grad_norm": 0.40992996096611023,
      "learning_rate": 8.044496724433493e-06,
      "loss": 0.3234272003173828,
      "step": 630
    },
    {
      "epoch": 2.5179282868525896,
      "grad_norm": 0.5893640518188477,
      "learning_rate": 7.988514546316437e-06,
      "loss": 0.4355270564556122,
      "step": 632
    },
    {
      "epoch": 2.5258964143426295,
      "grad_norm": 0.2000420093536377,
      "learning_rate": 7.932663272264079e-06,
      "loss": 0.8469869494438171,
      "step": 634
    },
    {
      "epoch": 2.5338645418326693,
      "grad_norm": 2.3752691745758057,
      "learning_rate": 7.87694533003471e-06,
      "loss": 0.7660741209983826,
      "step": 636
    },
    {
      "epoch": 2.541832669322709,
      "grad_norm": 0.38962018489837646,
      "learning_rate": 7.82136314159092e-06,
      "loss": 0.8690819144248962,
      "step": 638
    },
    {
      "epoch": 2.549800796812749,
      "grad_norm": 0.12522974610328674,
      "learning_rate": 7.765919122994319e-06,
      "loss": 0.30734166502952576,
      "step": 640
    },
    {
      "epoch": 2.557768924302789,
      "grad_norm": 0.2867269217967987,
      "learning_rate": 7.710615684300518e-06,
      "loss": 0.8029553890228271,
      "step": 642
    },
    {
      "epoch": 2.5657370517928286,
      "grad_norm": 0.19536034762859344,
      "learning_rate": 7.655455229454354e-06,
      "loss": 0.7299827337265015,
      "step": 644
    },
    {
      "epoch": 2.5737051792828685,
      "grad_norm": 0.38802143931388855,
      "learning_rate": 7.6004401561854e-06,
      "loss": 0.5647068023681641,
      "step": 646
    },
    {
      "epoch": 2.5816733067729083,
      "grad_norm": 0.23780430853366852,
      "learning_rate": 7.545572855903756e-06,
      "loss": 0.8187406659126282,
      "step": 648
    },
    {
      "epoch": 2.589641434262948,
      "grad_norm": 0.16200827062129974,
      "learning_rate": 7.490855713596081e-06,
      "loss": 0.6111049056053162,
      "step": 650
    },
    {
      "epoch": 2.597609561752988,
      "grad_norm": 0.3508452773094177,
      "learning_rate": 7.4362911077219155e-06,
      "loss": 0.5618999600410461,
      "step": 652
    },
    {
      "epoch": 2.605577689243028,
      "grad_norm": 0.632539689540863,
      "learning_rate": 7.381881410110326e-06,
      "loss": 0.5881422758102417,
      "step": 654
    },
    {
      "epoch": 2.6135458167330676,
      "grad_norm": 0.3363528251647949,
      "learning_rate": 7.327628985856765e-06,
      "loss": 0.27832284569740295,
      "step": 656
    },
    {
      "epoch": 2.6215139442231075,
      "grad_norm": 0.3875587582588196,
      "learning_rate": 7.2735361932202965e-06,
      "loss": 0.48132815957069397,
      "step": 658
    },
    {
      "epoch": 2.6294820717131473,
      "grad_norm": 0.17317701876163483,
      "learning_rate": 7.219605383521067e-06,
      "loss": 0.7134672999382019,
      "step": 660
    },
    {
      "epoch": 2.637450199203187,
      "grad_norm": 0.7390320897102356,
      "learning_rate": 7.165838901038107e-06,
      "loss": 0.6096608638763428,
      "step": 662
    },
    {
      "epoch": 2.645418326693227,
      "grad_norm": 0.4265226423740387,
      "learning_rate": 7.112239082907433e-06,
      "loss": 0.4687889814376831,
      "step": 664
    },
    {
      "epoch": 2.653386454183267,
      "grad_norm": 0.3511049151420593,
      "learning_rate": 7.058808259020442e-06,
      "loss": 0.7662704586982727,
      "step": 666
    },
    {
      "epoch": 2.6613545816733066,
      "grad_norm": 0.14830322563648224,
      "learning_rate": 7.005548751922642e-06,
      "loss": 0.58231520652771,
      "step": 668
    },
    {
      "epoch": 2.6693227091633465,
      "grad_norm": 1.2010635137557983,
      "learning_rate": 6.952462876712707e-06,
      "loss": 0.5792977213859558,
      "step": 670
    },
    {
      "epoch": 2.6772908366533863,
      "grad_norm": 0.09235794842243195,
      "learning_rate": 6.899552940941829e-06,
      "loss": 0.25369662046432495,
      "step": 672
    },
    {
      "epoch": 2.685258964143426,
      "grad_norm": 0.2072373628616333,
      "learning_rate": 6.846821244513411e-06,
      "loss": 0.8279775381088257,
      "step": 674
    },
    {
      "epoch": 2.6932270916334664,
      "grad_norm": 1.2984479665756226,
      "learning_rate": 6.794270079583101e-06,
      "loss": 0.5745002031326294,
      "step": 676
    },
    {
      "epoch": 2.7011952191235062,
      "grad_norm": 0.3513108193874359,
      "learning_rate": 6.741901730459166e-06,
      "loss": 0.5112911462783813,
      "step": 678
    },
    {
      "epoch": 2.709163346613546,
      "grad_norm": 0.3224731385707855,
      "learning_rate": 6.68971847350317e-06,
      "loss": 0.8467031717300415,
      "step": 680
    },
    {
      "epoch": 2.717131474103586,
      "grad_norm": 0.1862608939409256,
      "learning_rate": 6.6377225770310514e-06,
      "loss": 0.8552311062812805,
      "step": 682
    },
    {
      "epoch": 2.7250996015936257,
      "grad_norm": 0.2126304805278778,
      "learning_rate": 6.585916301214519e-06,
      "loss": 0.6002206206321716,
      "step": 684
    },
    {
      "epoch": 2.7330677290836656,
      "grad_norm": 0.24358995258808136,
      "learning_rate": 6.534301897982774e-06,
      "loss": 0.3887878954410553,
      "step": 686
    },
    {
      "epoch": 2.7410358565737054,
      "grad_norm": 0.19447919726371765,
      "learning_rate": 6.482881610924674e-06,
      "loss": 0.83692467212677,
      "step": 688
    },
    {
      "epoch": 2.7490039840637452,
      "grad_norm": 0.4111102223396301,
      "learning_rate": 6.43165767519117e-06,
      "loss": 0.4429361820220947,
      "step": 690
    },
    {
      "epoch": 2.756972111553785,
      "grad_norm": 0.24252432584762573,
      "learning_rate": 6.380632317398168e-06,
      "loss": 0.9488629102706909,
      "step": 692
    },
    {
      "epoch": 2.764940239043825,
      "grad_norm": 1.053657054901123,
      "learning_rate": 6.3298077555297245e-06,
      "loss": 0.6902448534965515,
      "step": 694
    },
    {
      "epoch": 2.7729083665338647,
      "grad_norm": 0.5043600797653198,
      "learning_rate": 6.279186198841647e-06,
      "loss": 0.3843526840209961,
      "step": 696
    },
    {
      "epoch": 2.7808764940239046,
      "grad_norm": 0.2123395949602127,
      "learning_rate": 6.228769847765468e-06,
      "loss": 0.6492659449577332,
      "step": 698
    },
    {
      "epoch": 2.7888446215139444,
      "grad_norm": 0.2770644426345825,
      "learning_rate": 6.178560893812774e-06,
      "loss": 0.8826568126678467,
      "step": 700
    },
    {
      "epoch": 2.7968127490039842,
      "grad_norm": 0.36284762620925903,
      "learning_rate": 6.1285615194799695e-06,
      "loss": 0.49446773529052734,
      "step": 702
    },
    {
      "epoch": 2.804780876494024,
      "grad_norm": 0.42989182472229004,
      "learning_rate": 6.0787738981533825e-06,
      "loss": 0.8925838470458984,
      "step": 704
    },
    {
      "epoch": 2.812749003984064,
      "grad_norm": 0.19607830047607422,
      "learning_rate": 6.0292001940148174e-06,
      "loss": 0.7888954281806946,
      "step": 706
    },
    {
      "epoch": 2.8207171314741037,
      "grad_norm": 0.3013284504413605,
      "learning_rate": 5.979842561947455e-06,
      "loss": 0.9048675298690796,
      "step": 708
    },
    {
      "epoch": 2.8286852589641436,
      "grad_norm": 0.1352807581424713,
      "learning_rate": 5.9307031474422074e-06,
      "loss": 0.45338767766952515,
      "step": 710
    },
    {
      "epoch": 2.8366533864541834,
      "grad_norm": 0.2785419225692749,
      "learning_rate": 5.88178408650445e-06,
      "loss": 0.8890714645385742,
      "step": 712
    },
    {
      "epoch": 2.8446215139442232,
      "grad_norm": 0.29292890429496765,
      "learning_rate": 5.833087505561148e-06,
      "loss": 0.5466417074203491,
      "step": 714
    },
    {
      "epoch": 2.852589641434263,
      "grad_norm": 0.37993285059928894,
      "learning_rate": 5.784615521368468e-06,
      "loss": 0.8255132436752319,
      "step": 716
    },
    {
      "epoch": 2.860557768924303,
      "grad_norm": 1.8282309770584106,
      "learning_rate": 5.736370240919735e-06,
      "loss": 0.5976371765136719,
      "step": 718
    },
    {
      "epoch": 2.8685258964143427,
      "grad_norm": 0.3135332465171814,
      "learning_rate": 5.688353761353862e-06,
      "loss": 0.5985286831855774,
      "step": 720
    },
    {
      "epoch": 2.8764940239043826,
      "grad_norm": 0.4926091730594635,
      "learning_rate": 5.640568169864173e-06,
      "loss": 0.5094777941703796,
      "step": 722
    },
    {
      "epoch": 2.8844621513944224,
      "grad_norm": 0.28557467460632324,
      "learning_rate": 5.5930155436076875e-06,
      "loss": 0.5855981111526489,
      "step": 724
    },
    {
      "epoch": 2.8924302788844622,
      "grad_norm": 0.1873752772808075,
      "learning_rate": 5.5456979496148385e-06,
      "loss": 0.8087204694747925,
      "step": 726
    },
    {
      "epoch": 2.900398406374502,
      "grad_norm": 0.17823685705661774,
      "learning_rate": 5.498617444699603e-06,
      "loss": 0.48339203000068665,
      "step": 728
    },
    {
      "epoch": 2.908366533864542,
      "grad_norm": 0.18616022169589996,
      "learning_rate": 5.451776075370114e-06,
      "loss": 0.8099565505981445,
      "step": 730
    },
    {
      "epoch": 2.9163346613545817,
      "grad_norm": 1.2350081205368042,
      "learning_rate": 5.405175877739684e-06,
      "loss": 0.23912093043327332,
      "step": 732
    },
    {
      "epoch": 2.9243027888446216,
      "grad_norm": 0.18930251896381378,
      "learning_rate": 5.3588188774383235e-06,
      "loss": 0.8681591153144836,
      "step": 734
    },
    {
      "epoch": 2.9322709163346614,
      "grad_norm": 0.6436243057250977,
      "learning_rate": 5.3127070895246604e-06,
      "loss": 0.37712156772613525,
      "step": 736
    },
    {
      "epoch": 2.9402390438247012,
      "grad_norm": 0.5814879536628723,
      "learning_rate": 5.266842518398369e-06,
      "loss": 0.5876435041427612,
      "step": 738
    },
    {
      "epoch": 2.948207171314741,
      "grad_norm": 0.19967974722385406,
      "learning_rate": 5.2212271577130495e-06,
      "loss": 0.8929345607757568,
      "step": 740
    },
    {
      "epoch": 2.956175298804781,
      "grad_norm": 0.5970357656478882,
      "learning_rate": 5.1758629902895375e-06,
      "loss": 0.24971675872802734,
      "step": 742
    },
    {
      "epoch": 2.9641434262948207,
      "grad_norm": 0.1709216833114624,
      "learning_rate": 5.1307519880297385e-06,
      "loss": 0.6873310804367065,
      "step": 744
    },
    {
      "epoch": 2.9721115537848606,
      "grad_norm": 0.32703879475593567,
      "learning_rate": 5.085896111830908e-06,
      "loss": 0.7358705997467041,
      "step": 746
    },
    {
      "epoch": 2.9800796812749004,
      "grad_norm": 0.6799337267875671,
      "learning_rate": 5.041297311500417e-06,
      "loss": 0.40029191970825195,
      "step": 748
    },
    {
      "epoch": 2.9880478087649402,
      "grad_norm": 0.2632702589035034,
      "learning_rate": 4.996957525670984e-06,
      "loss": 0.11657044291496277,
      "step": 750
    },
    {
      "epoch": 2.99601593625498,
      "grad_norm": 0.18080343306064606,
      "learning_rate": 4.952878681716416e-06,
      "loss": 0.832659125328064,
      "step": 752
    },
    {
      "epoch": 3.00398406374502,
      "grad_norm": 0.18914125859737396,
      "learning_rate": 4.9090626956678355e-06,
      "loss": 0.4053405225276947,
      "step": 754
    },
    {
      "epoch": 3.0119521912350598,
      "grad_norm": 0.27092087268829346,
      "learning_rate": 4.865511472130379e-06,
      "loss": 0.37303829193115234,
      "step": 756
    },
    {
      "epoch": 3.0199203187250996,
      "grad_norm": 0.450639545917511,
      "learning_rate": 4.822226904200422e-06,
      "loss": 0.16986803710460663,
      "step": 758
    },
    {
      "epoch": 3.0278884462151394,
      "grad_norm": 0.3367942273616791,
      "learning_rate": 4.779210873383273e-06,
      "loss": 0.3704196810722351,
      "step": 760
    },
    {
      "epoch": 3.0358565737051793,
      "grad_norm": 1.1672158241271973,
      "learning_rate": 4.7364652495114e-06,
      "loss": 0.2740946114063263,
      "step": 762
    },
    {
      "epoch": 3.043824701195219,
      "grad_norm": 0.2160903364419937,
      "learning_rate": 4.693991890663152e-06,
      "loss": 0.5514514446258545,
      "step": 764
    },
    {
      "epoch": 3.051792828685259,
      "grad_norm": 0.40776246786117554,
      "learning_rate": 4.651792643081986e-06,
      "loss": 0.18550026416778564,
      "step": 766
    },
    {
      "epoch": 3.0597609561752988,
      "grad_norm": 1.529194951057434,
      "learning_rate": 4.609869341096217e-06,
      "loss": 0.3695829212665558,
      "step": 768
    },
    {
      "epoch": 3.0677290836653386,
      "grad_norm": 0.2934879660606384,
      "learning_rate": 4.568223807039282e-06,
      "loss": 0.40844109654426575,
      "step": 770
    },
    {
      "epoch": 3.0756972111553784,
      "grad_norm": 0.3209177553653717,
      "learning_rate": 4.52685785117052e-06,
      "loss": 0.6010529398918152,
      "step": 772
    },
    {
      "epoch": 3.0836653386454183,
      "grad_norm": 0.4266740083694458,
      "learning_rate": 4.485773271596503e-06,
      "loss": 0.5307098627090454,
      "step": 774
    },
    {
      "epoch": 3.091633466135458,
      "grad_norm": 0.05216313526034355,
      "learning_rate": 4.444971854192848e-06,
      "loss": 0.31485700607299805,
      "step": 776
    },
    {
      "epoch": 3.099601593625498,
      "grad_norm": 0.25069746375083923,
      "learning_rate": 4.404455372526615e-06,
      "loss": 0.3358692228794098,
      "step": 778
    },
    {
      "epoch": 3.1075697211155378,
      "grad_norm": 0.3168417513370514,
      "learning_rate": 4.3642255877791876e-06,
      "loss": 0.5717326402664185,
      "step": 780
    },
    {
      "epoch": 3.1155378486055776,
      "grad_norm": 0.08099119365215302,
      "learning_rate": 4.324284248669742e-06,
      "loss": 0.13734716176986694,
      "step": 782
    },
    {
      "epoch": 3.1235059760956174,
      "grad_norm": 0.1908828169107437,
      "learning_rate": 4.2846330913792136e-06,
      "loss": 0.6576658487319946,
      "step": 784
    },
    {
      "epoch": 3.1314741035856573,
      "grad_norm": 0.2333727329969406,
      "learning_rate": 4.245273839474835e-06,
      "loss": 0.4396320879459381,
      "step": 786
    },
    {
      "epoch": 3.139442231075697,
      "grad_norm": 2.3938891887664795,
      "learning_rate": 4.20620820383523e-06,
      "loss": 0.3137848973274231,
      "step": 788
    },
    {
      "epoch": 3.147410358565737,
      "grad_norm": 0.1568251997232437,
      "learning_rate": 4.167437882576009e-06,
      "loss": 0.2973101735115051,
      "step": 790
    },
    {
      "epoch": 3.1553784860557768,
      "grad_norm": 0.4758032560348511,
      "learning_rate": 4.128964560975993e-06,
      "loss": 0.36776888370513916,
      "step": 792
    },
    {
      "epoch": 3.1633466135458166,
      "grad_norm": 0.07684729248285294,
      "learning_rate": 4.090789911403938e-06,
      "loss": 0.16028432548046112,
      "step": 794
    },
    {
      "epoch": 3.1713147410358564,
      "grad_norm": 0.38295745849609375,
      "learning_rate": 4.0529155932458444e-06,
      "loss": 0.35760873556137085,
      "step": 796
    },
    {
      "epoch": 3.1792828685258963,
      "grad_norm": 0.2896295189857483,
      "learning_rate": 4.015343252832824e-06,
      "loss": 0.41373881697654724,
      "step": 798
    },
    {
      "epoch": 3.187250996015936,
      "grad_norm": 0.3329697847366333,
      "learning_rate": 3.978074523369533e-06,
      "loss": 0.5053466558456421,
      "step": 800
    },
    {
      "epoch": 3.195219123505976,
      "grad_norm": 0.15466375648975372,
      "learning_rate": 3.941111024863193e-06,
      "loss": 0.42649704217910767,
      "step": 802
    },
    {
      "epoch": 3.2031872509960158,
      "grad_norm": 1.1632299423217773,
      "learning_rate": 3.90445436405316e-06,
      "loss": 0.28943392634391785,
      "step": 804
    },
    {
      "epoch": 3.2111553784860556,
      "grad_norm": 0.9543743133544922,
      "learning_rate": 3.86810613434109e-06,
      "loss": 0.5052630305290222,
      "step": 806
    },
    {
      "epoch": 3.2191235059760954,
      "grad_norm": 0.3861115574836731,
      "learning_rate": 3.832067915721661e-06,
      "loss": 0.06477329879999161,
      "step": 808
    },
    {
      "epoch": 3.2270916334661353,
      "grad_norm": 0.3546229898929596,
      "learning_rate": 3.7963412747139204e-06,
      "loss": 0.38985639810562134,
      "step": 810
    },
    {
      "epoch": 3.235059760956175,
      "grad_norm": 0.37589749693870544,
      "learning_rate": 3.7609277642931642e-06,
      "loss": 0.2961961328983307,
      "step": 812
    },
    {
      "epoch": 3.243027888446215,
      "grad_norm": 0.21967440843582153,
      "learning_rate": 3.725828923823447e-06,
      "loss": 0.477074533700943,
      "step": 814
    },
    {
      "epoch": 3.2509960159362548,
      "grad_norm": 0.5088498592376709,
      "learning_rate": 3.6910462789906697e-06,
      "loss": 0.09672769904136658,
      "step": 816
    },
    {
      "epoch": 3.2589641434262946,
      "grad_norm": 3.130514621734619,
      "learning_rate": 3.6565813417362476e-06,
      "loss": 0.5637337565422058,
      "step": 818
    },
    {
      "epoch": 3.2669322709163344,
      "grad_norm": 1.0084922313690186,
      "learning_rate": 3.622435610191398e-06,
      "loss": 0.22835613787174225,
      "step": 820
    },
    {
      "epoch": 3.2749003984063743,
      "grad_norm": 0.7883324027061462,
      "learning_rate": 3.5886105686120267e-06,
      "loss": 0.23144333064556122,
      "step": 822
    },
    {
      "epoch": 3.2828685258964145,
      "grad_norm": 1.4939353466033936,
      "learning_rate": 3.5551076873141875e-06,
      "loss": 0.32171422243118286,
      "step": 824
    },
    {
      "epoch": 3.2908366533864544,
      "grad_norm": 0.250313937664032,
      "learning_rate": 3.5219284226101992e-06,
      "loss": 0.5594013333320618,
      "step": 826
    },
    {
      "epoch": 3.298804780876494,
      "grad_norm": 0.3546219766139984,
      "learning_rate": 3.4890742167453134e-06,
      "loss": 0.6331716775894165,
      "step": 828
    },
    {
      "epoch": 3.306772908366534,
      "grad_norm": 0.39389073848724365,
      "learning_rate": 3.4565464978350447e-06,
      "loss": 0.46244314312934875,
      "step": 830
    },
    {
      "epoch": 3.314741035856574,
      "grad_norm": 0.3902747333049774,
      "learning_rate": 3.4243466798030813e-06,
      "loss": 0.24344006180763245,
      "step": 832
    },
    {
      "epoch": 3.3227091633466137,
      "grad_norm": 0.6764956712722778,
      "learning_rate": 3.3924761623198276e-06,
      "loss": 0.21558792889118195,
      "step": 834
    },
    {
      "epoch": 3.3306772908366535,
      "grad_norm": 0.38500118255615234,
      "learning_rate": 3.360936330741563e-06,
      "loss": 0.3872177302837372,
      "step": 836
    },
    {
      "epoch": 3.3386454183266934,
      "grad_norm": 2.351408004760742,
      "learning_rate": 3.329728556050219e-06,
      "loss": 0.45764562487602234,
      "step": 838
    },
    {
      "epoch": 3.346613545816733,
      "grad_norm": 2.8567023277282715,
      "learning_rate": 3.2988541947937948e-06,
      "loss": 0.0879729688167572,
      "step": 840
    },
    {
      "epoch": 3.354581673306773,
      "grad_norm": 0.3541049063205719,
      "learning_rate": 3.2683145890273803e-06,
      "loss": 0.3018583655357361,
      "step": 842
    },
    {
      "epoch": 3.362549800796813,
      "grad_norm": 0.19183960556983948,
      "learning_rate": 3.2381110662548244e-06,
      "loss": 0.36365094780921936,
      "step": 844
    },
    {
      "epoch": 3.3705179282868527,
      "grad_norm": 0.2597419023513794,
      "learning_rate": 3.208244939371029e-06,
      "loss": 0.6369303464889526,
      "step": 846
    },
    {
      "epoch": 3.3784860557768925,
      "grad_norm": 0.261401891708374,
      "learning_rate": 3.178717506604878e-06,
      "loss": 0.579558789730072,
      "step": 848
    },
    {
      "epoch": 3.3864541832669324,
      "grad_norm": 0.9293565154075623,
      "learning_rate": 3.1495300514628137e-06,
      "loss": 0.4159211218357086,
      "step": 850
    },
    {
      "epoch": 3.394422310756972,
      "grad_norm": 0.2699755132198334,
      "learning_rate": 3.120683842673029e-06,
      "loss": 0.5360977053642273,
      "step": 852
    },
    {
      "epoch": 3.402390438247012,
      "grad_norm": 0.2359551638364792,
      "learning_rate": 3.092180134130341e-06,
      "loss": 0.3280014395713806,
      "step": 854
    },
    {
      "epoch": 3.410358565737052,
      "grad_norm": 0.270657479763031,
      "learning_rate": 3.064020164841661e-06,
      "loss": 0.4834939241409302,
      "step": 856
    },
    {
      "epoch": 3.4183266932270917,
      "grad_norm": 0.22715269029140472,
      "learning_rate": 3.03620515887216e-06,
      "loss": 0.5365287065505981,
      "step": 858
    },
    {
      "epoch": 3.4262948207171315,
      "grad_norm": 0.26945826411247253,
      "learning_rate": 3.0087363252920392e-06,
      "loss": 0.6132234930992126,
      "step": 860
    },
    {
      "epoch": 3.4342629482071714,
      "grad_norm": 0.3072214722633362,
      "learning_rate": 2.9816148581239945e-06,
      "loss": 0.3467521667480469,
      "step": 862
    },
    {
      "epoch": 3.442231075697211,
      "grad_norm": 0.43960896134376526,
      "learning_rate": 2.954841936291304e-06,
      "loss": 0.2627929449081421,
      "step": 864
    },
    {
      "epoch": 3.450199203187251,
      "grad_norm": 0.283093124628067,
      "learning_rate": 2.9284187235665706e-06,
      "loss": 0.48483937978744507,
      "step": 866
    },
    {
      "epoch": 3.458167330677291,
      "grad_norm": 0.25170329213142395,
      "learning_rate": 2.902346368521161e-06,
      "loss": 0.3998684883117676,
      "step": 868
    },
    {
      "epoch": 3.4661354581673307,
      "grad_norm": 0.6622265577316284,
      "learning_rate": 2.8766260044752596e-06,
      "loss": 0.36085110902786255,
      "step": 870
    },
    {
      "epoch": 3.4741035856573705,
      "grad_norm": 0.042979441583156586,
      "learning_rate": 2.851258749448613e-06,
      "loss": 0.3579687774181366,
      "step": 872
    },
    {
      "epoch": 3.4820717131474104,
      "grad_norm": 0.07233504951000214,
      "learning_rate": 2.8262457061119275e-06,
      "loss": 0.3310646712779999,
      "step": 874
    },
    {
      "epoch": 3.49003984063745,
      "grad_norm": 0.19331711530685425,
      "learning_rate": 2.801587961738936e-06,
      "loss": 0.28486010432243347,
      "step": 876
    },
    {
      "epoch": 3.49800796812749,
      "grad_norm": 0.24624978005886078,
      "learning_rate": 2.7772865881591487e-06,
      "loss": 0.35191646218299866,
      "step": 878
    },
    {
      "epoch": 3.50597609561753,
      "grad_norm": 0.07663383334875107,
      "learning_rate": 2.7533426417112463e-06,
      "loss": 0.23816527426242828,
      "step": 880
    },
    {
      "epoch": 3.5139442231075697,
      "grad_norm": 0.1692771017551422,
      "learning_rate": 2.729757163197175e-06,
      "loss": 0.20403024554252625,
      "step": 882
    },
    {
      "epoch": 3.5219123505976095,
      "grad_norm": 0.4337323009967804,
      "learning_rate": 2.706531177836896e-06,
      "loss": 0.04598721116781235,
      "step": 884
    },
    {
      "epoch": 3.5298804780876494,
      "grad_norm": 0.34074339270591736,
      "learning_rate": 2.683665695223827e-06,
      "loss": 0.4786261022090912,
      "step": 886
    },
    {
      "epoch": 3.537848605577689,
      "grad_norm": 0.2945248782634735,
      "learning_rate": 2.661161709280954e-06,
      "loss": 0.5180566310882568,
      "step": 888
    },
    {
      "epoch": 3.545816733067729,
      "grad_norm": 0.2568756937980652,
      "learning_rate": 2.6390201982176284e-06,
      "loss": 0.3125327527523041,
      "step": 890
    },
    {
      "epoch": 3.553784860557769,
      "grad_norm": 1.1016383171081543,
      "learning_rate": 2.617242124487048e-06,
      "loss": 0.3341105580329895,
      "step": 892
    },
    {
      "epoch": 3.5617529880478087,
      "grad_norm": 0.25893574953079224,
      "learning_rate": 2.595828434744411e-06,
      "loss": 0.7315629124641418,
      "step": 894
    },
    {
      "epoch": 3.5697211155378485,
      "grad_norm": 0.10278267413377762,
      "learning_rate": 2.574780059805779e-06,
      "loss": 0.09938185662031174,
      "step": 896
    },
    {
      "epoch": 3.5776892430278884,
      "grad_norm": 0.2548079788684845,
      "learning_rate": 2.5540979146076145e-06,
      "loss": 0.3255918323993683,
      "step": 898
    },
    {
      "epoch": 3.585657370517928,
      "grad_norm": 0.3227812349796295,
      "learning_rate": 2.5337828981669966e-06,
      "loss": 0.13555879890918732,
      "step": 900
    },
    {
      "epoch": 3.593625498007968,
      "grad_norm": 0.5672808289527893,
      "learning_rate": 2.513835893542561e-06,
      "loss": 0.1049320250749588,
      "step": 902
    },
    {
      "epoch": 3.601593625498008,
      "grad_norm": 0.25081929564476013,
      "learning_rate": 2.494257767796101e-06,
      "loss": 0.3797585070133209,
      "step": 904
    },
    {
      "epoch": 3.6095617529880477,
      "grad_norm": 0.4710356891155243,
      "learning_rate": 2.475049371954888e-06,
      "loss": 0.3027661144733429,
      "step": 906
    },
    {
      "epoch": 3.6175298804780875,
      "grad_norm": 0.2510360777378082,
      "learning_rate": 2.4562115409746662e-06,
      "loss": 0.5094070434570312,
      "step": 908
    },
    {
      "epoch": 3.6254980079681274,
      "grad_norm": 0.43042221665382385,
      "learning_rate": 2.4377450937033753e-06,
      "loss": 0.4744018614292145,
      "step": 910
    },
    {
      "epoch": 3.633466135458167,
      "grad_norm": 0.06483194977045059,
      "learning_rate": 2.419650832845541e-06,
      "loss": 0.04786789417266846,
      "step": 912
    },
    {
      "epoch": 3.641434262948207,
      "grad_norm": 0.25760993361473083,
      "learning_rate": 2.40192954492739e-06,
      "loss": 0.40470847487449646,
      "step": 914
    },
    {
      "epoch": 3.649402390438247,
      "grad_norm": 0.7198867201805115,
      "learning_rate": 2.3845820002626623e-06,
      "loss": 0.5723772048950195,
      "step": 916
    },
    {
      "epoch": 3.6573705179282867,
      "grad_norm": 0.5992567539215088,
      "learning_rate": 2.3676089529191234e-06,
      "loss": 0.06981807202100754,
      "step": 918
    },
    {
      "epoch": 3.6653386454183265,
      "grad_norm": 0.2739182412624359,
      "learning_rate": 2.3510111406857915e-06,
      "loss": 0.533017098903656,
      "step": 920
    },
    {
      "epoch": 3.6733067729083664,
      "grad_norm": 0.16629727184772491,
      "learning_rate": 2.3347892850408575e-06,
      "loss": 0.07635924965143204,
      "step": 922
    },
    {
      "epoch": 3.681274900398406,
      "grad_norm": 0.5157669186592102,
      "learning_rate": 2.3189440911203328e-06,
      "loss": 0.3664593994617462,
      "step": 924
    },
    {
      "epoch": 3.6892430278884465,
      "grad_norm": 0.27757561206817627,
      "learning_rate": 2.3034762476873973e-06,
      "loss": 0.1329723596572876,
      "step": 926
    },
    {
      "epoch": 3.6972111553784863,
      "grad_norm": 0.4265131950378418,
      "learning_rate": 2.2883864271024524e-06,
      "loss": 0.31522467732429504,
      "step": 928
    },
    {
      "epoch": 3.705179282868526,
      "grad_norm": 0.10874529927968979,
      "learning_rate": 2.273675285293905e-06,
      "loss": 0.42426741123199463,
      "step": 930
    },
    {
      "epoch": 3.713147410358566,
      "grad_norm": 0.35626259446144104,
      "learning_rate": 2.2593434617296474e-06,
      "loss": 0.5419244170188904,
      "step": 932
    },
    {
      "epoch": 3.721115537848606,
      "grad_norm": 1.1875288486480713,
      "learning_rate": 2.2453915793892665e-06,
      "loss": 0.23740825057029724,
      "step": 934
    },
    {
      "epoch": 3.7290836653386457,
      "grad_norm": 0.21155528724193573,
      "learning_rate": 2.231820244736958e-06,
      "loss": 0.5010062456130981,
      "step": 936
    },
    {
      "epoch": 3.7370517928286855,
      "grad_norm": 0.35504090785980225,
      "learning_rate": 2.2186300476951693e-06,
      "loss": 0.16797490417957306,
      "step": 938
    },
    {
      "epoch": 3.7450199203187253,
      "grad_norm": 0.3238517940044403,
      "learning_rate": 2.2058215616189578e-06,
      "loss": 0.5096255540847778,
      "step": 940
    },
    {
      "epoch": 3.752988047808765,
      "grad_norm": 0.36483195424079895,
      "learning_rate": 2.1933953432710588e-06,
      "loss": 0.2782249450683594,
      "step": 942
    },
    {
      "epoch": 3.760956175298805,
      "grad_norm": 0.46273383498191833,
      "learning_rate": 2.181351932797695e-06,
      "loss": 0.31418055295944214,
      "step": 944
    },
    {
      "epoch": 3.768924302788845,
      "grad_norm": 0.4258814752101898,
      "learning_rate": 2.169691853705094e-06,
      "loss": 0.6382843255996704,
      "step": 946
    },
    {
      "epoch": 3.7768924302788847,
      "grad_norm": 0.2860477864742279,
      "learning_rate": 2.1584156128367303e-06,
      "loss": 0.32270297408103943,
      "step": 948
    },
    {
      "epoch": 3.7848605577689245,
      "grad_norm": 0.21522466838359833,
      "learning_rate": 2.147523700351291e-06,
      "loss": 0.5712357759475708,
      "step": 950
    },
    {
      "epoch": 3.7928286852589643,
      "grad_norm": 0.2516200840473175,
      "learning_rate": 2.137016589701375e-06,
      "loss": 0.5555206537246704,
      "step": 952
    },
    {
      "epoch": 3.800796812749004,
      "grad_norm": 0.23942607641220093,
      "learning_rate": 2.126894737612914e-06,
      "loss": 0.2843998968601227,
      "step": 954
    },
    {
      "epoch": 3.808764940239044,
      "grad_norm": 1.4363948106765747,
      "learning_rate": 2.117158584065313e-06,
      "loss": 0.4706723392009735,
      "step": 956
    },
    {
      "epoch": 3.816733067729084,
      "grad_norm": 0.670280396938324,
      "learning_rate": 2.107808552272329e-06,
      "loss": 0.27754148840904236,
      "step": 958
    },
    {
      "epoch": 3.8247011952191237,
      "grad_norm": 0.40577781200408936,
      "learning_rate": 2.0988450486636745e-06,
      "loss": 0.1861335188150406,
      "step": 960
    },
    {
      "epoch": 3.8326693227091635,
      "grad_norm": 0.12758664786815643,
      "learning_rate": 2.090268462867351e-06,
      "loss": 0.34570229053497314,
      "step": 962
    },
    {
      "epoch": 3.8406374501992033,
      "grad_norm": 0.3902248740196228,
      "learning_rate": 2.0820791676927093e-06,
      "loss": 0.6225243210792542,
      "step": 964
    },
    {
      "epoch": 3.848605577689243,
      "grad_norm": 0.42479851841926575,
      "learning_rate": 2.0742775191142496e-06,
      "loss": 0.1883630007505417,
      "step": 966
    },
    {
      "epoch": 3.856573705179283,
      "grad_norm": 0.32185542583465576,
      "learning_rate": 2.066863856256143e-06,
      "loss": 0.39622795581817627,
      "step": 968
    },
    {
      "epoch": 3.864541832669323,
      "grad_norm": 0.315428227186203,
      "learning_rate": 2.0598385013774933e-06,
      "loss": 0.3555518686771393,
      "step": 970
    },
    {
      "epoch": 3.8725099601593627,
      "grad_norm": 0.10775071382522583,
      "learning_rate": 2.053201759858327e-06,
      "loss": 0.3210831582546234,
      "step": 972
    },
    {
      "epoch": 3.8804780876494025,
      "grad_norm": 0.11887846142053604,
      "learning_rate": 2.0469539201863224e-06,
      "loss": 0.2502799928188324,
      "step": 974
    },
    {
      "epoch": 3.8884462151394423,
      "grad_norm": 0.31070947647094727,
      "learning_rate": 2.0410952539442624e-06,
      "loss": 0.5481354594230652,
      "step": 976
    },
    {
      "epoch": 3.896414342629482,
      "grad_norm": 0.31816530227661133,
      "learning_rate": 2.0356260157982393e-06,
      "loss": 0.2996366322040558,
      "step": 978
    },
    {
      "epoch": 3.904382470119522,
      "grad_norm": 0.18345490097999573,
      "learning_rate": 2.030546443486575e-06,
      "loss": 0.24936345219612122,
      "step": 980
    },
    {
      "epoch": 3.912350597609562,
      "grad_norm": 0.26529887318611145,
      "learning_rate": 2.0258567578094947e-06,
      "loss": 0.19975776970386505,
      "step": 982
    },
    {
      "epoch": 3.9203187250996017,
      "grad_norm": 0.25069504976272583,
      "learning_rate": 2.0215571626195242e-06,
      "loss": 0.5472447872161865,
      "step": 984
    },
    {
      "epoch": 3.9282868525896415,
      "grad_norm": 0.33956313133239746,
      "learning_rate": 2.0176478448126306e-06,
      "loss": 0.14271996915340424,
      "step": 986
    },
    {
      "epoch": 3.9362549800796813,
      "grad_norm": 0.8762012124061584,
      "learning_rate": 2.014128974320099e-06,
      "loss": 0.494120717048645,
      "step": 988
    },
    {
      "epoch": 3.944223107569721,
      "grad_norm": 0.06045879051089287,
      "learning_rate": 2.011000704101143e-06,
      "loss": 0.3118327856063843,
      "step": 990
    },
    {
      "epoch": 3.952191235059761,
      "grad_norm": 0.22002360224723816,
      "learning_rate": 2.00826317013626e-06,
      "loss": 0.53458571434021,
      "step": 992
    },
    {
      "epoch": 3.960159362549801,
      "grad_norm": 0.08198000490665436,
      "learning_rate": 2.0059164914213175e-06,
      "loss": 0.1380147933959961,
      "step": 994
    },
    {
      "epoch": 3.9681274900398407,
      "grad_norm": 0.6921637058258057,
      "learning_rate": 2.00396076996238e-06,
      "loss": 0.5101975798606873,
      "step": 996
    },
    {
      "epoch": 3.9760956175298805,
      "grad_norm": 0.09050004929304123,
      "learning_rate": 2.0023960907712782e-06,
      "loss": 0.3632965683937073,
      "step": 998
    },
    {
      "epoch": 3.9840637450199203,
      "grad_norm": 0.2718322277069092,
      "learning_rate": 2.0012225218619097e-06,
      "loss": 0.24679109454154968,
      "step": 1000
    },
    {
      "epoch": 3.99203187250996,
      "grad_norm": 0.2601509988307953,
      "learning_rate": 2.0004401142472857e-06,
      "loss": 0.5495860576629639,
      "step": 1002
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0840764045715332,
      "learning_rate": 2.000048901937313e-06,
      "loss": 0.12128544598817825,
      "step": 1004
    },
    {
      "epoch": 4.0,
      "step": 1004,
      "total_flos": 3.519329211045118e+18,
      "train_loss": 0.7596988798197047,
      "train_runtime": 8661.0758,
      "train_samples_per_second": 6.955,
      "train_steps_per_second": 0.116
    }
  ],
  "logging_steps": 2,
  "max_steps": 1004,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 99999,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.519329211045118e+18,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}