{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 1.9750077149218717,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8462,
      "step": 5
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.4320818536235405,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8235,
      "step": 10
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 1.1296543108414194,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.7682,
      "step": 15
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.6804177314185508,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.7273,
      "step": 20
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5114220473629801,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.6958,
      "step": 25
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.4776075911866677,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6736,
      "step": 30
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.42771970924840125,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6447,
      "step": 35
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.4477521517865052,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6379,
      "step": 40
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.4136530731052078,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6334,
      "step": 45
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.4136692403680581,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6149,
      "step": 50
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.44362487561505304,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.6342,
      "step": 55
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.4820513641866406,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6273,
      "step": 60
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.4389919560166169,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6061,
      "step": 65
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.4639488632458146,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6351,
      "step": 70
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.5802537471900081,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.6125,
      "step": 75
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.5313712773119875,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6081,
      "step": 80
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.6217373859670787,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.6005,
      "step": 85
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.4803371554553108,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.5897,
      "step": 90
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.5375998658362728,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.5917,
      "step": 95
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.5661967250341784,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.6046,
      "step": 100
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.4531271843109738,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.6101,
      "step": 105
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.5104394692649273,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5971,
      "step": 110
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.4916843277201706,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.5876,
      "step": 115
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.5272068863337249,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.5949,
      "step": 120
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.5093640172515701,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5922,
      "step": 125
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.45447887964282774,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.585,
      "step": 130
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.4001143050434082,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5846,
      "step": 135
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.5577082012607348,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5868,
      "step": 140
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.5476231926493538,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5893,
      "step": 145
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.46507776087097996,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5732,
      "step": 150
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.4554325042497221,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5776,
      "step": 155
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.4693614005345544,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5853,
      "step": 160
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.48210638314020626,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.5813,
      "step": 165
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.40918672155147295,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5916,
      "step": 170
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.40104404458379694,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.5857,
      "step": 175
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.4363659278961314,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5814,
      "step": 180
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.44403368992242354,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5743,
      "step": 185
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.4510875347188779,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5681,
      "step": 190
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.4496061645025098,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.5683,
      "step": 195
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.3995171453892101,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5681,
      "step": 200
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.49476062390659087,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5818,
      "step": 205
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.42828769292928387,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5737,
      "step": 210
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.6953178358038983,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5714,
      "step": 215
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.5573543272065618,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5672,
      "step": 220
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.4991321031350928,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.5599,
      "step": 225
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.4979276045514997,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5688,
      "step": 230
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.44150041814390334,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.5758,
      "step": 235
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.43778559930387784,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5749,
      "step": 240
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.46769554029009164,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5601,
      "step": 245
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.4999686970886678,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5604,
      "step": 250
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.3976067100762551,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.5578,
      "step": 255
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.4110596325571904,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.567,
      "step": 260
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.38286531363520704,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5556,
      "step": 265
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.39790335264166216,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5543,
      "step": 270
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.45253398550514967,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5586,
      "step": 275
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.5081936287841872,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5621,
      "step": 280
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.4337740773506668,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.5673,
      "step": 285
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.4008240226697529,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5514,
      "step": 290
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.36096047071687043,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5454,
      "step": 295
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.4613958212475722,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5614,
      "step": 300
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.40966372209924423,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5458,
      "step": 305
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.3587520084929092,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5539,
      "step": 310
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.4205722817346792,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5567,
      "step": 315
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.4364503166332872,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5682,
      "step": 320
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.5407958495399552,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.5527,
      "step": 325
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.49572285110933245,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5594,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.4572780520378008,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.555,
      "step": 335
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.49574398911606543,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5411,
      "step": 340
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.39551925905047247,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5556,
      "step": 345
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.46195208494399076,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5521,
      "step": 350
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.4161079327413066,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.546,
      "step": 355
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.42625484461446933,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5635,
      "step": 360
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.4176716961209062,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5576,
      "step": 365
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3575941764883018,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5597,
      "step": 370
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.4176002570317524,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5469,
      "step": 375
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.4400273027474589,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5461,
      "step": 380
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.34916157516720675,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.5447,
      "step": 385
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.37071494530496585,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5481,
      "step": 390
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.3875564230104429,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5546,
      "step": 395
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.377377501563779,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5464,
      "step": 400
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.396122939242755,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.5488,
      "step": 405
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.4912829346966443,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5344,
      "step": 410
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.37146001640483844,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5459,
      "step": 415
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.4001029124019289,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5509,
      "step": 420
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.4032895368862278,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5542,
      "step": 425
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.42916343143626445,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5511,
      "step": 430
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.40722469487970786,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5547,
      "step": 435
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.4068592075733166,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5471,
      "step": 440
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.33587271715668876,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5454,
      "step": 445
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.4552018121411592,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5481,
      "step": 450
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.41025001926757526,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5408,
      "step": 455
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.37559299835698057,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5368,
      "step": 460
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.32956652581967366,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5276,
      "step": 465
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.41980838082285094,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5516,
      "step": 470
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.36529498236397145,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5511,
      "step": 475
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.4069384844792981,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5431,
      "step": 480
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.33914734341400626,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5379,
      "step": 485
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.34003189361025093,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5524,
      "step": 490
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.3612968977854989,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5365,
      "step": 495
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.4146432031798141,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5324,
      "step": 500
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.3883712691652435,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5448,
      "step": 505
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.32921311640370177,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5395,
      "step": 510
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.31351506549571895,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5383,
      "step": 515
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.3260094537514537,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5298,
      "step": 520
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.36968161395431975,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5331,
      "step": 525
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.3518812846796185,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5231,
      "step": 530
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.3446586065633445,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5448,
      "step": 535
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.3244767706271663,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5408,
      "step": 540
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.35460428127125937,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.5438,
      "step": 545
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.36062597357178056,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.539,
      "step": 550
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.3240564200933119,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5369,
      "step": 555
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.3294455486450902,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5404,
      "step": 560
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.3546329847122573,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5312,
      "step": 565
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.32852819814856365,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5216,
      "step": 570
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.3212442318874353,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5365,
      "step": 575
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.3450607174489223,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5411,
      "step": 580
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.30946288323123555,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.529,
      "step": 585
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.3012443551581362,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5266,
      "step": 590
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.36152500589188347,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.5365,
      "step": 595
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.3184664977196892,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5276,
      "step": 600
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.29640705405648343,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5198,
      "step": 605
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.32588181630029356,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5342,
      "step": 610
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.33033324530738756,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5408,
      "step": 615
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3589730155787974,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5255,
      "step": 620
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.3258150814986452,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5292,
      "step": 625
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.35576554794037185,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5307,
      "step": 630
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.3419104850810916,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5339,
      "step": 635
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.3191173873553689,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5184,
      "step": 640
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.30741023796232286,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.5304,
      "step": 645
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.297628096921709,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5181,
      "step": 650
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.36151662976014337,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5269,
      "step": 655
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.3051778567350286,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5307,
      "step": 660
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.31871784370224476,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.5268,
      "step": 665
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.333108381519437,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.524,
      "step": 670
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.3082362879080899,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.5314,
      "step": 675
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.3104148195644461,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.524,
      "step": 680
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.31999382004894306,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5237,
      "step": 685
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.29098513225948724,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.5177,
      "step": 690
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.30363273295609583,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5219,
      "step": 695
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.3092910268334778,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5386,
      "step": 700
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.2926377812753093,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.5216,
      "step": 705
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.3524567874129882,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5135,
      "step": 710
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.3171716884579878,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5171,
      "step": 715
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.2988258180846129,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5194,
      "step": 720
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.3217760098800343,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5232,
      "step": 725
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.300160560624556,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5194,
      "step": 730
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.3089928808380071,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.5207,
      "step": 735
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.27657515702454555,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5192,
      "step": 740
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.2793879876725504,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5146,
      "step": 745
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.3120045609062447,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5311,
      "step": 750
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.30192200166107913,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5318,
      "step": 755
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.30679686235298426,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5182,
      "step": 760
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.3092446606554235,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.526,
      "step": 765
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.31326688912108297,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.529,
      "step": 770
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.3001327981363857,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.5358,
      "step": 775
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.28356506354955435,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5151,
      "step": 780
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.28639919565797317,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.5182,
      "step": 785
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.28571313805500137,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5269,
      "step": 790
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.2912910658822969,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.5151,
      "step": 795
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.2982324936963473,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5213,
      "step": 800
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.26692133348534597,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5356,
      "step": 805
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.2988564566061472,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5142,
      "step": 810
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.28127158773732697,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.519,
      "step": 815
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.29971711675190804,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5036,
      "step": 820
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.2716855772334307,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5246,
      "step": 825
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.277232403879636,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5288,
      "step": 830
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.27040909679168806,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5188,
      "step": 835
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.2946913564827007,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.5177,
      "step": 840
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.27444187622708766,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.5138,
      "step": 845
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.26298524484599706,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5197,
      "step": 850
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.2577022474279286,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.5079,
      "step": 855
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.2832801101449736,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.527,
      "step": 860
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.2942428377508324,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5282,
      "step": 865
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.2609215538008877,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.5103,
      "step": 870
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.2672043757726758,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.5148,
      "step": 875
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.297000685198304,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5128,
      "step": 880
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.275296265606091,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5228,
      "step": 885
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.25704178255931553,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5251,
      "step": 890
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.27398722372405143,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5193,
      "step": 895
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.2516834451733719,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5221,
      "step": 900
    },
    {
      "epoch": 0.8442164179104478,
      "grad_norm": 0.2662834635954678,
      "learning_rate": 7.922507959181673e-06,
      "loss": 0.5104,
      "step": 905
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.2564159812075332,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5166,
      "step": 910
    },
    {
      "epoch": 0.8535447761194029,
      "grad_norm": 0.317298990431241,
      "learning_rate": 7.58966441993719e-06,
      "loss": 0.5132,
      "step": 915
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.2854200167114319,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5342,
      "step": 920
    },
    {
      "epoch": 0.8628731343283582,
      "grad_norm": 0.2989252117784558,
      "learning_rate": 7.275781315903374e-06,
      "loss": 0.5328,
      "step": 925
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.26420429180404253,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5239,
      "step": 930
    },
    {
      "epoch": 0.8722014925373134,
      "grad_norm": 0.25605782788591613,
      "learning_rate": 6.9811575551641224e-06,
      "loss": 0.512,
      "step": 935
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.2691470223345099,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.5058,
      "step": 940
    },
    {
      "epoch": 0.8815298507462687,
      "grad_norm": 0.26948743349958904,
      "learning_rate": 6.706073705302254e-06,
      "loss": 0.5174,
      "step": 945
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.27108247792036266,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.5075,
      "step": 950
    },
    {
      "epoch": 0.8908582089552238,
      "grad_norm": 0.2579773143862865,
      "learning_rate": 6.450791726217538e-06,
      "loss": 0.5151,
      "step": 955
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.25145771917807047,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.5188,
      "step": 960
    },
    {
      "epoch": 0.9001865671641791,
      "grad_norm": 0.2535827191009452,
      "learning_rate": 6.215554720664598e-06,
      "loss": 0.5249,
      "step": 965
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.2708392506019422,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5247,
      "step": 970
    },
    {
      "epoch": 0.9095149253731343,
      "grad_norm": 0.25771890812792486,
      "learning_rate": 6.000586702748301e-06,
      "loss": 0.5103,
      "step": 975
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.2501403892481585,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.5124,
      "step": 980
    },
    {
      "epoch": 0.9188432835820896,
      "grad_norm": 0.26774429808991634,
      "learning_rate": 5.8060923845971825e-06,
      "loss": 0.5206,
      "step": 985
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.2460471855503398,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.5062,
      "step": 990
    },
    {
      "epoch": 0.9281716417910447,
      "grad_norm": 0.2644019337153552,
      "learning_rate": 5.632256981417845e-06,
      "loss": 0.5144,
      "step": 995
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.25859447689759785,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5166,
      "step": 1000
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.24883418482798325,
      "learning_rate": 5.479246035116201e-06,
      "loss": 0.5094,
      "step": 1005
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.253586747488515,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5117,
      "step": 1010
    },
    {
      "epoch": 0.9468283582089553,
      "grad_norm": 0.2721339619373832,
      "learning_rate": 5.347205256653387e-06,
      "loss": 0.5144,
      "step": 1015
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.2802449749636702,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5075,
      "step": 1020
    },
    {
      "epoch": 0.9561567164179104,
      "grad_norm": 0.25322551665468956,
      "learning_rate": 5.236260387286509e-06,
      "loss": 0.5138,
      "step": 1025
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.2546704125209255,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5104,
      "step": 1030
    },
    {
      "epoch": 0.9654850746268657,
      "grad_norm": 0.24795592848823428,
      "learning_rate": 5.1465170788263595e-06,
      "loss": 0.5184,
      "step": 1035
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.26082536016005153,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5141,
      "step": 1040
    },
    {
      "epoch": 0.9748134328358209,
      "grad_norm": 0.2583468665669609,
      "learning_rate": 5.07806079302615e-06,
      "loss": 0.5212,
      "step": 1045
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.2512890918345273,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5158,
      "step": 1050
    },
    {
      "epoch": 0.9841417910447762,
      "grad_norm": 0.2646142546440819,
      "learning_rate": 5.030956720197035e-06,
      "loss": 0.5083,
      "step": 1055
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.2511048148212438,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.526,
      "step": 1060
    },
    {
      "epoch": 0.9934701492537313,
      "grad_norm": 0.2698500767525977,
      "learning_rate": 5.005249717127964e-06,
      "loss": 0.5171,
      "step": 1065
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.2662909375749429,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5175,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "step": 1072,
      "total_flos": 488621249396736.0,
      "train_loss": 0.5498761394019447,
      "train_runtime": 14566.6023,
      "train_samples_per_second": 2.355,
      "train_steps_per_second": 0.074
    }
  ],
  "logging_steps": 5,
  "max_steps": 1072,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 488621249396736.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}