{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1056,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014204545454545454,
      "grad_norm": 16.233426918647123,
      "learning_rate": 3.7735849056603773e-06,
      "loss": 1.9367,
      "step": 5
    },
    {
      "epoch": 0.028409090909090908,
      "grad_norm": 3.700378147836118,
      "learning_rate": 8.49056603773585e-06,
      "loss": 1.5623,
      "step": 10
    },
    {
      "epoch": 0.04261363636363636,
      "grad_norm": 1.411828049687858,
      "learning_rate": 1.320754716981132e-05,
      "loss": 1.0659,
      "step": 15
    },
    {
      "epoch": 0.056818181818181816,
      "grad_norm": 0.5502005679393165,
      "learning_rate": 1.7924528301886792e-05,
      "loss": 0.876,
      "step": 20
    },
    {
      "epoch": 0.07102272727272728,
      "grad_norm": 0.39964518009977995,
      "learning_rate": 2.2641509433962265e-05,
      "loss": 0.7709,
      "step": 25
    },
    {
      "epoch": 0.08522727272727272,
      "grad_norm": 0.2819938310171901,
      "learning_rate": 2.7358490566037738e-05,
      "loss": 0.7212,
      "step": 30
    },
    {
      "epoch": 0.09943181818181818,
      "grad_norm": 0.22596397421278658,
      "learning_rate": 3.207547169811321e-05,
      "loss": 0.6941,
      "step": 35
    },
    {
      "epoch": 0.11363636363636363,
      "grad_norm": 0.19821070636934998,
      "learning_rate": 3.679245283018868e-05,
      "loss": 0.6626,
      "step": 40
    },
    {
      "epoch": 0.1278409090909091,
      "grad_norm": 0.1776825768871472,
      "learning_rate": 4.150943396226415e-05,
      "loss": 0.6458,
      "step": 45
    },
    {
      "epoch": 0.14204545454545456,
      "grad_norm": 0.15959969460809642,
      "learning_rate": 4.6226415094339625e-05,
      "loss": 0.6317,
      "step": 50
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.16520239332693426,
      "learning_rate": 4.999988963025308e-05,
      "loss": 0.6111,
      "step": 55
    },
    {
      "epoch": 0.17045454545454544,
      "grad_norm": 0.14604811373242252,
      "learning_rate": 4.999602680280325e-05,
      "loss": 0.6049,
      "step": 60
    },
    {
      "epoch": 0.1846590909090909,
      "grad_norm": 0.153405093129135,
      "learning_rate": 4.998664657075762e-05,
      "loss": 0.5878,
      "step": 65
    },
    {
      "epoch": 0.19886363636363635,
      "grad_norm": 0.13290633271306487,
      "learning_rate": 4.9971751234724005e-05,
      "loss": 0.5872,
      "step": 70
    },
    {
      "epoch": 0.21306818181818182,
      "grad_norm": 0.14464611604457972,
      "learning_rate": 4.995134444795177e-05,
      "loss": 0.5809,
      "step": 75
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.1539967921458611,
      "learning_rate": 4.992543121543581e-05,
      "loss": 0.5687,
      "step": 80
    },
    {
      "epoch": 0.24147727272727273,
      "grad_norm": 0.1872261556450498,
      "learning_rate": 4.9894017892689e-05,
      "loss": 0.5736,
      "step": 85
    },
    {
      "epoch": 0.2556818181818182,
      "grad_norm": 0.1461192468496281,
      "learning_rate": 4.985711218418346e-05,
      "loss": 0.5729,
      "step": 90
    },
    {
      "epoch": 0.26988636363636365,
      "grad_norm": 0.15510991989675932,
      "learning_rate": 4.981472314146093e-05,
      "loss": 0.5651,
      "step": 95
    },
    {
      "epoch": 0.2840909090909091,
      "grad_norm": 0.15383826812713372,
      "learning_rate": 4.976686116091281e-05,
      "loss": 0.5669,
      "step": 100
    },
    {
      "epoch": 0.29829545454545453,
      "grad_norm": 0.12937769599328974,
      "learning_rate": 4.971353798123026e-05,
      "loss": 0.5608,
      "step": 105
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.14866942632641453,
      "learning_rate": 4.965476668052522e-05,
      "loss": 0.5547,
      "step": 110
    },
    {
      "epoch": 0.32670454545454547,
      "grad_norm": 0.15767211601593334,
      "learning_rate": 4.959056167312283e-05,
      "loss": 0.5572,
      "step": 115
    },
    {
      "epoch": 0.3409090909090909,
      "grad_norm": 0.16235663995148145,
      "learning_rate": 4.952093870602615e-05,
      "loss": 0.5613,
      "step": 120
    },
    {
      "epoch": 0.35511363636363635,
      "grad_norm": 0.15329109428934323,
      "learning_rate": 4.944591485505405e-05,
      "loss": 0.5515,
      "step": 125
    },
    {
      "epoch": 0.3693181818181818,
      "grad_norm": 0.16475626406284527,
      "learning_rate": 4.93655085206531e-05,
      "loss": 0.5538,
      "step": 130
    },
    {
      "epoch": 0.3835227272727273,
      "grad_norm": 0.17396494811540575,
      "learning_rate": 4.92797394233848e-05,
      "loss": 0.5455,
      "step": 135
    },
    {
      "epoch": 0.3977272727272727,
      "grad_norm": 0.1684931659310488,
      "learning_rate": 4.918862859908871e-05,
      "loss": 0.5439,
      "step": 140
    },
    {
      "epoch": 0.4119318181818182,
      "grad_norm": 0.15970431189153267,
      "learning_rate": 4.909219839372333e-05,
      "loss": 0.5431,
      "step": 145
    },
    {
      "epoch": 0.42613636363636365,
      "grad_norm": 0.1333165751370488,
      "learning_rate": 4.8990472457885386e-05,
      "loss": 0.5367,
      "step": 150
    },
    {
      "epoch": 0.4403409090909091,
      "grad_norm": 0.15417885622445218,
      "learning_rate": 4.8883475741009324e-05,
      "loss": 0.5381,
      "step": 155
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.1416173613974234,
      "learning_rate": 4.877123448524815e-05,
      "loss": 0.5356,
      "step": 160
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.13087882893136601,
      "learning_rate": 4.865377621903726e-05,
      "loss": 0.5288,
      "step": 165
    },
    {
      "epoch": 0.48295454545454547,
      "grad_norm": 0.14764636629826566,
      "learning_rate": 4.853112975034276e-05,
      "loss": 0.5347,
      "step": 170
    },
    {
      "epoch": 0.4971590909090909,
      "grad_norm": 0.16758747023050016,
      "learning_rate": 4.840332515959601e-05,
      "loss": 0.5372,
      "step": 175
    },
    {
      "epoch": 0.5113636363636364,
      "grad_norm": 0.18224306322957698,
      "learning_rate": 4.827039379231604e-05,
      "loss": 0.5275,
      "step": 180
    },
    {
      "epoch": 0.5255681818181818,
      "grad_norm": 0.1662865774101808,
      "learning_rate": 4.813236825142173e-05,
      "loss": 0.5408,
      "step": 185
    },
    {
      "epoch": 0.5397727272727273,
      "grad_norm": 0.17763881404547607,
      "learning_rate": 4.798928238923555e-05,
      "loss": 0.5289,
      "step": 190
    },
    {
      "epoch": 0.5539772727272727,
      "grad_norm": 0.16091215915603133,
      "learning_rate": 4.7841171299180906e-05,
      "loss": 0.5397,
      "step": 195
    },
    {
      "epoch": 0.5681818181818182,
      "grad_norm": 0.1404358353075446,
      "learning_rate": 4.768807130717508e-05,
      "loss": 0.5256,
      "step": 200
    },
    {
      "epoch": 0.5823863636363636,
      "grad_norm": 0.166459644318507,
      "learning_rate": 4.753001996271991e-05,
      "loss": 0.5235,
      "step": 205
    },
    {
      "epoch": 0.5965909090909091,
      "grad_norm": 0.13606607208723026,
      "learning_rate": 4.73670560296923e-05,
      "loss": 0.5291,
      "step": 210
    },
    {
      "epoch": 0.6107954545454546,
      "grad_norm": 0.1370837785726949,
      "learning_rate": 4.7199219476837034e-05,
      "loss": 0.5254,
      "step": 215
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.130041325528082,
      "learning_rate": 4.702655146796385e-05,
      "loss": 0.5281,
      "step": 220
    },
    {
      "epoch": 0.6392045454545454,
      "grad_norm": 0.13838008337473828,
      "learning_rate": 4.684909435185169e-05,
      "loss": 0.5163,
      "step": 225
    },
    {
      "epoch": 0.6534090909090909,
      "grad_norm": 0.1249145523375738,
      "learning_rate": 4.6666891651862096e-05,
      "loss": 0.5164,
      "step": 230
    },
    {
      "epoch": 0.6676136363636364,
      "grad_norm": 0.11918127255464575,
      "learning_rate": 4.6479988055264654e-05,
      "loss": 0.5156,
      "step": 235
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.13410033100059227,
      "learning_rate": 4.628842940227693e-05,
      "loss": 0.5267,
      "step": 240
    },
    {
      "epoch": 0.6960227272727273,
      "grad_norm": 0.14513137581304156,
      "learning_rate": 4.609226267482162e-05,
      "loss": 0.516,
      "step": 245
    },
    {
      "epoch": 0.7102272727272727,
      "grad_norm": 0.17472387433508568,
      "learning_rate": 4.589153598500372e-05,
      "loss": 0.5103,
      "step": 250
    },
    {
      "epoch": 0.7244318181818182,
      "grad_norm": 0.13807513906855917,
      "learning_rate": 4.568629856331049e-05,
      "loss": 0.5111,
      "step": 255
    },
    {
      "epoch": 0.7386363636363636,
      "grad_norm": 0.1606554123531758,
      "learning_rate": 4.547660074653713e-05,
      "loss": 0.5106,
      "step": 260
    },
    {
      "epoch": 0.7528409090909091,
      "grad_norm": 0.13512067917257073,
      "learning_rate": 4.526249396544107e-05,
      "loss": 0.52,
      "step": 265
    },
    {
      "epoch": 0.7670454545454546,
      "grad_norm": 0.12966854507444983,
      "learning_rate": 4.504403073212805e-05,
      "loss": 0.5137,
      "step": 270
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.14417839979685987,
      "learning_rate": 4.482126462717289e-05,
      "loss": 0.5131,
      "step": 275
    },
    {
      "epoch": 0.7954545454545454,
      "grad_norm": 0.14967027976448785,
      "learning_rate": 4.45942502864783e-05,
      "loss": 0.5164,
      "step": 280
    },
    {
      "epoch": 0.8096590909090909,
      "grad_norm": 0.13862324342167873,
      "learning_rate": 4.436304338787472e-05,
      "loss": 0.5083,
      "step": 285
    },
    {
      "epoch": 0.8238636363636364,
      "grad_norm": 0.1285629705362091,
      "learning_rate": 4.4127700637464834e-05,
      "loss": 0.5196,
      "step": 290
    },
    {
      "epoch": 0.8380681818181818,
      "grad_norm": 0.12976169635746898,
      "learning_rate": 4.388827975571561e-05,
      "loss": 0.5125,
      "step": 295
    },
    {
      "epoch": 0.8522727272727273,
      "grad_norm": 0.14210550405367858,
      "learning_rate": 4.364483946330182e-05,
      "loss": 0.5051,
      "step": 300
    },
    {
      "epoch": 0.8664772727272727,
      "grad_norm": 0.16907430743029442,
      "learning_rate": 4.3397439466704074e-05,
      "loss": 0.5135,
      "step": 305
    },
    {
      "epoch": 0.8806818181818182,
      "grad_norm": 0.16584741007419773,
      "learning_rate": 4.314614044356517e-05,
      "loss": 0.5104,
      "step": 310
    },
    {
      "epoch": 0.8948863636363636,
      "grad_norm": 0.15722645631531904,
      "learning_rate": 4.289100402780813e-05,
      "loss": 0.5044,
      "step": 315
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.13402395269180703,
      "learning_rate": 4.2632092794519926e-05,
      "loss": 0.5126,
      "step": 320
    },
    {
      "epoch": 0.9232954545454546,
      "grad_norm": 0.12897493665476578,
      "learning_rate": 4.236947024460409e-05,
      "loss": 0.5073,
      "step": 325
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.12307110209575203,
      "learning_rate": 4.2103200789206496e-05,
      "loss": 0.5064,
      "step": 330
    },
    {
      "epoch": 0.9517045454545454,
      "grad_norm": 0.13103687604123465,
      "learning_rate": 4.1833349733917774e-05,
      "loss": 0.5018,
      "step": 335
    },
    {
      "epoch": 0.9659090909090909,
      "grad_norm": 0.1441008725503634,
      "learning_rate": 4.155998326275636e-05,
      "loss": 0.5016,
      "step": 340
    },
    {
      "epoch": 0.9801136363636364,
      "grad_norm": 0.14676840920070153,
      "learning_rate": 4.128316842193612e-05,
      "loss": 0.5017,
      "step": 345
    },
    {
      "epoch": 0.9943181818181818,
      "grad_norm": 0.1299772490214553,
      "learning_rate": 4.100297310342252e-05,
      "loss": 0.5084,
      "step": 350
    },
    {
      "epoch": 1.0085227272727273,
      "grad_norm": 0.11841362182036683,
      "learning_rate": 4.071946602828136e-05,
      "loss": 0.484,
      "step": 355
    },
    {
      "epoch": 1.0227272727272727,
      "grad_norm": 0.1320677520863094,
      "learning_rate": 4.043271672982416e-05,
      "loss": 0.4757,
      "step": 360
    },
    {
      "epoch": 1.0369318181818181,
      "grad_norm": 0.1525515212145726,
      "learning_rate": 4.014279553655429e-05,
      "loss": 0.4712,
      "step": 365
    },
    {
      "epoch": 1.0511363636363635,
      "grad_norm": 0.12388717475970465,
      "learning_rate": 3.984977355491816e-05,
      "loss": 0.4741,
      "step": 370
    },
    {
      "epoch": 1.0653409090909092,
      "grad_norm": 0.12285428485981521,
      "learning_rate": 3.9553722651865525e-05,
      "loss": 0.4805,
      "step": 375
    },
    {
      "epoch": 1.0795454545454546,
      "grad_norm": 0.14068486142094294,
      "learning_rate": 3.925471543722334e-05,
      "loss": 0.476,
      "step": 380
    },
    {
      "epoch": 1.09375,
      "grad_norm": 0.11956188719769793,
      "learning_rate": 3.895282524588736e-05,
      "loss": 0.4744,
      "step": 385
    },
    {
      "epoch": 1.1079545454545454,
      "grad_norm": 0.12566341647227572,
      "learning_rate": 3.8648126119835895e-05,
      "loss": 0.4728,
      "step": 390
    },
    {
      "epoch": 1.1221590909090908,
      "grad_norm": 0.11910968307671285,
      "learning_rate": 3.834069278997025e-05,
      "loss": 0.4672,
      "step": 395
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.1357524801191257,
      "learning_rate": 3.803060065778607e-05,
      "loss": 0.4765,
      "step": 400
    },
    {
      "epoch": 1.1505681818181819,
      "grad_norm": 0.12366787866403461,
      "learning_rate": 3.77179257768803e-05,
      "loss": 0.4769,
      "step": 405
    },
    {
      "epoch": 1.1647727272727273,
      "grad_norm": 0.1322634035357931,
      "learning_rate": 3.740274483429815e-05,
      "loss": 0.4808,
      "step": 410
    },
    {
      "epoch": 1.1789772727272727,
      "grad_norm": 0.13258784372135785,
      "learning_rate": 3.708513513172481e-05,
      "loss": 0.4763,
      "step": 415
    },
    {
      "epoch": 1.1931818181818181,
      "grad_norm": 0.1356892941793857,
      "learning_rate": 3.67651745665262e-05,
      "loss": 0.4772,
      "step": 420
    },
    {
      "epoch": 1.2073863636363638,
      "grad_norm": 0.12523099567966772,
      "learning_rate": 3.644294161264394e-05,
      "loss": 0.4691,
      "step": 425
    },
    {
      "epoch": 1.2215909090909092,
      "grad_norm": 0.12166675336360953,
      "learning_rate": 3.611851530134856e-05,
      "loss": 0.4763,
      "step": 430
    },
    {
      "epoch": 1.2357954545454546,
      "grad_norm": 0.11908896334134955,
      "learning_rate": 3.579197520185628e-05,
      "loss": 0.4772,
      "step": 435
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.1273632443581102,
      "learning_rate": 3.546340140181376e-05,
      "loss": 0.4887,
      "step": 440
    },
    {
      "epoch": 1.2642045454545454,
      "grad_norm": 0.11548733790030413,
      "learning_rate": 3.5132874487655685e-05,
      "loss": 0.4688,
      "step": 445
    },
    {
      "epoch": 1.2784090909090908,
      "grad_norm": 0.1251859404517232,
      "learning_rate": 3.480047552484001e-05,
      "loss": 0.4695,
      "step": 450
    },
    {
      "epoch": 1.2926136363636362,
      "grad_norm": 0.11622578447126647,
      "learning_rate": 3.4466286037965786e-05,
      "loss": 0.476,
      "step": 455
    },
    {
      "epoch": 1.3068181818181819,
      "grad_norm": 0.12517918383447438,
      "learning_rate": 3.4130387990778286e-05,
      "loss": 0.4805,
      "step": 460
    },
    {
      "epoch": 1.3210227272727273,
      "grad_norm": 0.12611864483533994,
      "learning_rate": 3.379286376606651e-05,
      "loss": 0.4754,
      "step": 465
    },
    {
      "epoch": 1.3352272727272727,
      "grad_norm": 0.13719373565556317,
      "learning_rate": 3.3453796145457835e-05,
      "loss": 0.4659,
      "step": 470
    },
    {
      "epoch": 1.3494318181818181,
      "grad_norm": 0.11875949906730647,
      "learning_rate": 3.311326828911494e-05,
      "loss": 0.4769,
      "step": 475
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.13353175627047656,
      "learning_rate": 3.277136371533981e-05,
      "loss": 0.4759,
      "step": 480
    },
    {
      "epoch": 1.3778409090909092,
      "grad_norm": 0.1417715943995476,
      "learning_rate": 3.2428166280089885e-05,
      "loss": 0.4744,
      "step": 485
    },
    {
      "epoch": 1.3920454545454546,
      "grad_norm": 0.14007028648406247,
      "learning_rate": 3.20837601564115e-05,
      "loss": 0.4734,
      "step": 490
    },
    {
      "epoch": 1.40625,
      "grad_norm": 0.12629456979175258,
      "learning_rate": 3.173822981379548e-05,
      "loss": 0.4724,
      "step": 495
    },
    {
      "epoch": 1.4204545454545454,
      "grad_norm": 0.12349607254647936,
      "learning_rate": 3.1391659997460064e-05,
      "loss": 0.4713,
      "step": 500
    },
    {
      "epoch": 1.4346590909090908,
      "grad_norm": 0.11464965836513225,
      "learning_rate": 3.104413570756612e-05,
      "loss": 0.4708,
      "step": 505
    },
    {
      "epoch": 1.4488636363636362,
      "grad_norm": 0.1208312662969904,
      "learning_rate": 3.069574217836993e-05,
      "loss": 0.4803,
      "step": 510
    },
    {
      "epoch": 1.4630681818181819,
      "grad_norm": 0.10698876943855158,
      "learning_rate": 3.0346564857318567e-05,
      "loss": 0.4633,
      "step": 515
    },
    {
      "epoch": 1.4772727272727273,
      "grad_norm": 0.11384636090749484,
      "learning_rate": 2.9996689384092874e-05,
      "loss": 0.4699,
      "step": 520
    },
    {
      "epoch": 1.4914772727272727,
      "grad_norm": 0.11145569232907346,
      "learning_rate": 2.964620156960342e-05,
      "loss": 0.4744,
      "step": 525
    },
    {
      "epoch": 1.5056818181818183,
      "grad_norm": 0.13346919699354048,
      "learning_rate": 2.929518737494437e-05,
      "loss": 0.4747,
      "step": 530
    },
    {
      "epoch": 1.5198863636363638,
      "grad_norm": 0.11742663104280518,
      "learning_rate": 2.8943732890310577e-05,
      "loss": 0.4713,
      "step": 535
    },
    {
      "epoch": 1.5340909090909092,
      "grad_norm": 0.12233693509061423,
      "learning_rate": 2.859192431388298e-05,
      "loss": 0.4747,
      "step": 540
    },
    {
      "epoch": 1.5482954545454546,
      "grad_norm": 0.11567941988020804,
      "learning_rate": 2.8239847930687522e-05,
      "loss": 0.4776,
      "step": 545
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.12718334961619032,
      "learning_rate": 2.788759009143279e-05,
      "loss": 0.4686,
      "step": 550
    },
    {
      "epoch": 1.5767045454545454,
      "grad_norm": 0.10446651216785705,
      "learning_rate": 2.753523719133152e-05,
      "loss": 0.4712,
      "step": 555
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.11060336379300346,
      "learning_rate": 2.7182875648911177e-05,
      "loss": 0.4739,
      "step": 560
    },
    {
      "epoch": 1.6051136363636362,
      "grad_norm": 0.11390898204723482,
      "learning_rate": 2.683059188481887e-05,
      "loss": 0.4638,
      "step": 565
    },
    {
      "epoch": 1.6193181818181817,
      "grad_norm": 0.11190601188366513,
      "learning_rate": 2.647847230062567e-05,
      "loss": 0.4711,
      "step": 570
    },
    {
      "epoch": 1.6335227272727273,
      "grad_norm": 0.11305263482331018,
      "learning_rate": 2.6126603257635733e-05,
      "loss": 0.4691,
      "step": 575
    },
    {
      "epoch": 1.6477272727272727,
      "grad_norm": 0.1265534565322895,
      "learning_rate": 2.5775071055705075e-05,
      "loss": 0.4791,
      "step": 580
    },
    {
      "epoch": 1.6619318181818183,
      "grad_norm": 0.11028857191690714,
      "learning_rate": 2.5423961912075712e-05,
      "loss": 0.4741,
      "step": 585
    },
    {
      "epoch": 1.6761363636363638,
      "grad_norm": 0.11709279310297076,
      "learning_rate": 2.5073361940229773e-05,
      "loss": 0.4627,
      "step": 590
    },
    {
      "epoch": 1.6903409090909092,
      "grad_norm": 0.12689675181012736,
      "learning_rate": 2.4723357128769286e-05,
      "loss": 0.4716,
      "step": 595
    },
    {
      "epoch": 1.7045454545454546,
      "grad_norm": 0.12232567128887789,
      "learning_rate": 2.4374033320326506e-05,
      "loss": 0.4718,
      "step": 600
    },
    {
      "epoch": 1.71875,
      "grad_norm": 0.11216636447930772,
      "learning_rate": 2.4025476190509938e-05,
      "loss": 0.4703,
      "step": 605
    },
    {
      "epoch": 1.7329545454545454,
      "grad_norm": 0.11075120188097681,
      "learning_rate": 2.3677771226891564e-05,
      "loss": 0.4749,
      "step": 610
    },
    {
      "epoch": 1.7471590909090908,
      "grad_norm": 0.1296834181423037,
      "learning_rate": 2.3331003708039932e-05,
      "loss": 0.4756,
      "step": 615
    },
    {
      "epoch": 1.7613636363636362,
      "grad_norm": 0.10873440353655982,
      "learning_rate": 2.2985258682604706e-05,
      "loss": 0.4753,
      "step": 620
    },
    {
      "epoch": 1.7755681818181817,
      "grad_norm": 0.11054731896051188,
      "learning_rate": 2.264062094845749e-05,
      "loss": 0.4681,
      "step": 625
    },
    {
      "epoch": 1.7897727272727273,
      "grad_norm": 0.1077785017076799,
      "learning_rate": 2.229717503189418e-05,
      "loss": 0.4667,
      "step": 630
    },
    {
      "epoch": 1.8039772727272727,
      "grad_norm": 0.12146741261267094,
      "learning_rate": 2.195500516690398e-05,
      "loss": 0.4667,
      "step": 635
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.11013296161652279,
      "learning_rate": 2.1614195274510002e-05,
      "loss": 0.4692,
      "step": 640
    },
    {
      "epoch": 1.8323863636363638,
      "grad_norm": 0.1122431509566975,
      "learning_rate": 2.1274828942186743e-05,
      "loss": 0.4761,
      "step": 645
    },
    {
      "epoch": 1.8465909090909092,
      "grad_norm": 0.11152245824612014,
      "learning_rate": 2.09369894033593e-05,
      "loss": 0.4611,
      "step": 650
    },
    {
      "epoch": 1.8607954545454546,
      "grad_norm": 0.10540289603858785,
      "learning_rate": 2.0600759516989397e-05,
      "loss": 0.4659,
      "step": 655
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.12268974044452592,
      "learning_rate": 2.026622174725336e-05,
      "loss": 0.4659,
      "step": 660
    },
    {
      "epoch": 1.8892045454545454,
      "grad_norm": 0.11217674945544069,
      "learning_rate": 1.9933458143316748e-05,
      "loss": 0.4683,
      "step": 665
    },
    {
      "epoch": 1.9034090909090908,
      "grad_norm": 0.11593305886838406,
      "learning_rate": 1.9602550319210967e-05,
      "loss": 0.4655,
      "step": 670
    },
    {
      "epoch": 1.9176136363636362,
      "grad_norm": 0.11394452786176468,
      "learning_rate": 1.9273579433816452e-05,
      "loss": 0.4634,
      "step": 675
    },
    {
      "epoch": 1.9318181818181817,
      "grad_norm": 0.12344588418657035,
      "learning_rate": 1.894662617095753e-05,
      "loss": 0.4605,
      "step": 680
    },
    {
      "epoch": 1.9460227272727273,
      "grad_norm": 0.10588427958742584,
      "learning_rate": 1.8621770719613862e-05,
      "loss": 0.4636,
      "step": 685
    },
    {
      "epoch": 1.9602272727272727,
      "grad_norm": 0.10548340772355062,
      "learning_rate": 1.829909275425309e-05,
      "loss": 0.4694,
      "step": 690
    },
    {
      "epoch": 1.9744318181818183,
      "grad_norm": 0.10424071348650422,
      "learning_rate": 1.7978671415289878e-05,
      "loss": 0.4705,
      "step": 695
    },
    {
      "epoch": 1.9886363636363638,
      "grad_norm": 0.1105427740992237,
      "learning_rate": 1.7660585289675758e-05,
      "loss": 0.4631,
      "step": 700
    },
    {
      "epoch": 2.002840909090909,
      "grad_norm": 0.12132401905710084,
      "learning_rate": 1.7344912391624826e-05,
      "loss": 0.4573,
      "step": 705
    },
    {
      "epoch": 2.0170454545454546,
      "grad_norm": 0.11150850216312608,
      "learning_rate": 1.7031730143479873e-05,
      "loss": 0.4495,
      "step": 710
    },
    {
      "epoch": 2.03125,
      "grad_norm": 0.1062932021752816,
      "learning_rate": 1.672111535672367e-05,
      "loss": 0.4361,
      "step": 715
    },
    {
      "epoch": 2.0454545454545454,
      "grad_norm": 0.10247863399966832,
      "learning_rate": 1.64131442131401e-05,
      "loss": 0.4437,
      "step": 720
    },
    {
      "epoch": 2.059659090909091,
      "grad_norm": 0.10677014234090489,
      "learning_rate": 1.6107892246129692e-05,
      "loss": 0.4434,
      "step": 725
    },
    {
      "epoch": 2.0738636363636362,
      "grad_norm": 0.10650883791134519,
      "learning_rate": 1.5805434322184224e-05,
      "loss": 0.4445,
      "step": 730
    },
    {
      "epoch": 2.0880681818181817,
      "grad_norm": 0.10387373586373833,
      "learning_rate": 1.550584462252487e-05,
      "loss": 0.4509,
      "step": 735
    },
    {
      "epoch": 2.102272727272727,
      "grad_norm": 0.1030086032533747,
      "learning_rate": 1.520919662490839e-05,
      "loss": 0.4438,
      "step": 740
    },
    {
      "epoch": 2.116477272727273,
      "grad_norm": 0.10277657855159573,
      "learning_rate": 1.4915563085605922e-05,
      "loss": 0.4391,
      "step": 745
    },
    {
      "epoch": 2.1306818181818183,
      "grad_norm": 0.09814581198724093,
      "learning_rate": 1.4625016021558654e-05,
      "loss": 0.4458,
      "step": 750
    },
    {
      "epoch": 2.1448863636363638,
      "grad_norm": 0.11056576999353432,
      "learning_rate": 1.4337626692714862e-05,
      "loss": 0.4496,
      "step": 755
    },
    {
      "epoch": 2.159090909090909,
      "grad_norm": 0.09890533126932854,
      "learning_rate": 1.4053465584552613e-05,
      "loss": 0.4395,
      "step": 760
    },
    {
      "epoch": 2.1732954545454546,
      "grad_norm": 0.10516851231359878,
      "learning_rate": 1.3772602390792397e-05,
      "loss": 0.442,
      "step": 765
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.10009393657351306,
      "learning_rate": 1.3495105996303959e-05,
      "loss": 0.4384,
      "step": 770
    },
    {
      "epoch": 2.2017045454545454,
      "grad_norm": 0.0998627875200131,
      "learning_rate": 1.3221044460211496e-05,
      "loss": 0.4436,
      "step": 775
    },
    {
      "epoch": 2.215909090909091,
      "grad_norm": 0.09883322341572116,
      "learning_rate": 1.2950484999201407e-05,
      "loss": 0.4508,
      "step": 780
    },
    {
      "epoch": 2.2301136363636362,
      "grad_norm": 0.09827162855077891,
      "learning_rate": 1.2683493971036633e-05,
      "loss": 0.4394,
      "step": 785
    },
    {
      "epoch": 2.2443181818181817,
      "grad_norm": 0.1028260573381402,
      "learning_rate": 1.2420136858281663e-05,
      "loss": 0.4442,
      "step": 790
    },
    {
      "epoch": 2.2585227272727275,
      "grad_norm": 0.10185969414875563,
      "learning_rate": 1.2160478252242191e-05,
      "loss": 0.4522,
      "step": 795
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.09897503980351197,
      "learning_rate": 1.1904581837123386e-05,
      "loss": 0.4395,
      "step": 800
    },
    {
      "epoch": 2.2869318181818183,
      "grad_norm": 0.10179604173619207,
      "learning_rate": 1.1652510374410607e-05,
      "loss": 0.4476,
      "step": 805
    },
    {
      "epoch": 2.3011363636363638,
      "grad_norm": 0.10087044008989485,
      "learning_rate": 1.1404325687476435e-05,
      "loss": 0.455,
      "step": 810
    },
    {
      "epoch": 2.315340909090909,
      "grad_norm": 0.09796524933029822,
      "learning_rate": 1.1160088646417834e-05,
      "loss": 0.441,
      "step": 815
    },
    {
      "epoch": 2.3295454545454546,
      "grad_norm": 0.10278194962659604,
      "learning_rate": 1.0919859153127004e-05,
      "loss": 0.4415,
      "step": 820
    },
    {
      "epoch": 2.34375,
      "grad_norm": 0.10008666177717428,
      "learning_rate": 1.0683696126599816e-05,
      "loss": 0.4504,
      "step": 825
    },
    {
      "epoch": 2.3579545454545454,
      "grad_norm": 0.10127145328521546,
      "learning_rate": 1.045165748848525e-05,
      "loss": 0.4432,
      "step": 830
    },
    {
      "epoch": 2.372159090909091,
      "grad_norm": 0.09465727719419995,
      "learning_rate": 1.0223800148879456e-05,
      "loss": 0.4406,
      "step": 835
    },
    {
      "epoch": 2.3863636363636362,
      "grad_norm": 0.10223238212578617,
      "learning_rate": 1.0000179992367925e-05,
      "loss": 0.4417,
      "step": 840
    },
    {
      "epoch": 2.4005681818181817,
      "grad_norm": 0.10542315681224666,
      "learning_rate": 9.780851864319119e-06,
      "loss": 0.4447,
      "step": 845
    },
    {
      "epoch": 2.4147727272727275,
      "grad_norm": 0.0963765340314958,
      "learning_rate": 9.565869557433065e-06,
      "loss": 0.4412,
      "step": 850
    },
    {
      "epoch": 2.4289772727272725,
      "grad_norm": 0.10046217850351946,
      "learning_rate": 9.35528579854809e-06,
      "loss": 0.4484,
      "step": 855
    },
    {
      "epoch": 2.4431818181818183,
      "grad_norm": 0.10147180197503058,
      "learning_rate": 9.149152235708938e-06,
      "loss": 0.4397,
      "step": 860
    },
    {
      "epoch": 2.4573863636363638,
      "grad_norm": 0.11146703090708097,
      "learning_rate": 8.947519425499546e-06,
      "loss": 0.4457,
      "step": 865
    },
    {
      "epoch": 2.471590909090909,
      "grad_norm": 0.0945703122512324,
      "learning_rate": 8.750436820643412e-06,
      "loss": 0.4392,
      "step": 870
    },
    {
      "epoch": 2.4857954545454546,
      "grad_norm": 0.09908284212508756,
      "learning_rate": 8.557952757874824e-06,
      "loss": 0.442,
      "step": 875
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.09395206380523909,
      "learning_rate": 8.370114446083686e-06,
      "loss": 0.4354,
      "step": 880
    },
    {
      "epoch": 2.5142045454545454,
      "grad_norm": 0.1005576468311754,
      "learning_rate": 8.186967954736992e-06,
      "loss": 0.4533,
      "step": 885
    },
    {
      "epoch": 2.528409090909091,
      "grad_norm": 0.1020807372597039,
      "learning_rate": 8.008558202579839e-06,
      "loss": 0.443,
      "step": 890
    },
    {
      "epoch": 2.5426136363636362,
      "grad_norm": 0.09991703967673085,
      "learning_rate": 7.834928946618499e-06,
      "loss": 0.4475,
      "step": 895
    },
    {
      "epoch": 2.5568181818181817,
      "grad_norm": 0.10275386995373673,
      "learning_rate": 7.666122771388617e-06,
      "loss": 0.4484,
      "step": 900
    },
    {
      "epoch": 2.5710227272727275,
      "grad_norm": 0.10616220162227955,
      "learning_rate": 7.502181078510801e-06,
      "loss": 0.441,
      "step": 905
    },
    {
      "epoch": 2.5852272727272725,
      "grad_norm": 0.09678868754399234,
      "learning_rate": 7.343144076536432e-06,
      "loss": 0.4345,
      "step": 910
    },
    {
      "epoch": 2.5994318181818183,
      "grad_norm": 0.09985443664593775,
      "learning_rate": 7.189050771086084e-06,
      "loss": 0.4432,
      "step": 915
    },
    {
      "epoch": 2.6136363636363638,
      "grad_norm": 0.09670777253001571,
      "learning_rate": 7.039938955282907e-06,
      "loss": 0.4417,
      "step": 920
    },
    {
      "epoch": 2.627840909090909,
      "grad_norm": 0.09783216540452833,
      "learning_rate": 6.895845200483489e-06,
      "loss": 0.4424,
      "step": 925
    },
    {
      "epoch": 2.6420454545454546,
      "grad_norm": 0.09147945018928255,
      "learning_rate": 6.756804847308286e-06,
      "loss": 0.4434,
      "step": 930
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.09802106237190225,
      "learning_rate": 6.622851996973932e-06,
      "loss": 0.4407,
      "step": 935
    },
    {
      "epoch": 2.6704545454545454,
      "grad_norm": 0.09462692879617098,
      "learning_rate": 6.49401950292956e-06,
      "loss": 0.4433,
      "step": 940
    },
    {
      "epoch": 2.684659090909091,
      "grad_norm": 0.09624748034407322,
      "learning_rate": 6.3703389627990844e-06,
      "loss": 0.4387,
      "step": 945
    },
    {
      "epoch": 2.6988636363636362,
      "grad_norm": 0.09543251734275723,
      "learning_rate": 6.2518407106315695e-06,
      "loss": 0.4363,
      "step": 950
    },
    {
      "epoch": 2.7130681818181817,
      "grad_norm": 0.09883095226273896,
      "learning_rate": 6.138553809461422e-06,
      "loss": 0.4386,
      "step": 955
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.09756625437555602,
      "learning_rate": 6.030506044180375e-06,
      "loss": 0.4459,
      "step": 960
    },
    {
      "epoch": 2.7414772727272725,
      "grad_norm": 0.09607808880251172,
      "learning_rate": 5.92772391472293e-06,
      "loss": 0.4359,
      "step": 965
    },
    {
      "epoch": 2.7556818181818183,
      "grad_norm": 0.09216704090541984,
      "learning_rate": 5.830232629566935e-06,
      "loss": 0.45,
      "step": 970
    },
    {
      "epoch": 2.7698863636363638,
      "grad_norm": 0.09291163689146356,
      "learning_rate": 5.738056099550936e-06,
      "loss": 0.4363,
      "step": 975
    },
    {
      "epoch": 2.784090909090909,
      "grad_norm": 0.09345711945306576,
      "learning_rate": 5.6512169320097665e-06,
      "loss": 0.4497,
      "step": 980
    },
    {
      "epoch": 2.7982954545454546,
      "grad_norm": 0.09335066659653964,
      "learning_rate": 5.569736425229844e-06,
      "loss": 0.4411,
      "step": 985
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.09732221176443216,
      "learning_rate": 5.493634563225524e-06,
      "loss": 0.4453,
      "step": 990
    },
    {
      "epoch": 2.8267045454545454,
      "grad_norm": 0.0954814046395625,
      "learning_rate": 5.422930010837793e-06,
      "loss": 0.4433,
      "step": 995
    },
    {
      "epoch": 2.840909090909091,
      "grad_norm": 0.09445562287163194,
      "learning_rate": 5.357640109156516e-06,
      "loss": 0.4396,
      "step": 1000
    },
    {
      "epoch": 2.8551136363636362,
      "grad_norm": 0.09370114782801439,
      "learning_rate": 5.297780871267323e-06,
      "loss": 0.441,
      "step": 1005
    },
    {
      "epoch": 2.8693181818181817,
      "grad_norm": 0.09116071509216306,
      "learning_rate": 5.243366978324231e-06,
      "loss": 0.4422,
      "step": 1010
    },
    {
      "epoch": 2.8835227272727275,
      "grad_norm": 0.09885237977800734,
      "learning_rate": 5.1944117759489156e-06,
      "loss": 0.4404,
      "step": 1015
    },
    {
      "epoch": 2.8977272727272725,
      "grad_norm": 0.09995837383801799,
      "learning_rate": 5.1509272709575616e-06,
      "loss": 0.4456,
      "step": 1020
    },
    {
      "epoch": 2.9119318181818183,
      "grad_norm": 0.09596178401038803,
      "learning_rate": 5.112924128416038e-06,
      "loss": 0.4404,
      "step": 1025
    },
    {
      "epoch": 2.9261363636363638,
      "grad_norm": 0.09406970508523729,
      "learning_rate": 5.080411669024179e-06,
      "loss": 0.4391,
      "step": 1030
    },
    {
      "epoch": 2.940340909090909,
      "grad_norm": 0.0951093143907049,
      "learning_rate": 5.0533978668297775e-06,
      "loss": 0.4483,
      "step": 1035
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.09615998973664797,
      "learning_rate": 5.031889347272857e-06,
      "loss": 0.4408,
      "step": 1040
    },
    {
      "epoch": 2.96875,
      "grad_norm": 0.09428736632782057,
      "learning_rate": 5.015891385560713e-06,
      "loss": 0.4432,
      "step": 1045
    },
    {
      "epoch": 2.9829545454545454,
      "grad_norm": 0.09338757409389373,
      "learning_rate": 5.005407905374099e-06,
      "loss": 0.4372,
      "step": 1050
    },
    {
      "epoch": 2.997159090909091,
      "grad_norm": 0.09934638872755362,
      "learning_rate": 5.0004414779049116e-06,
      "loss": 0.4421,
      "step": 1055
    },
    {
      "epoch": 3.0,
      "step": 1056,
      "total_flos": 962656789856256.0,
      "train_loss": 0.5030564876329718,
      "train_runtime": 14725.7023,
      "train_samples_per_second": 2.293,
      "train_steps_per_second": 0.072
    }
  ],
  "logging_steps": 5,
  "max_steps": 1056,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 962656789856256.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}