{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 740,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06756756756756757,
      "grad_norm": 4.363862700534323,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 1.4367,
      "step": 5
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 2.705123580523887,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 1.2876,
      "step": 10
    },
    {
      "epoch": 0.20270270270270271,
      "grad_norm": 1.9103615735976536,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 1.1519,
      "step": 15
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 1.4290192342216463,
      "learning_rate": 2.702702702702703e-05,
      "loss": 1.0869,
      "step": 20
    },
    {
      "epoch": 0.33783783783783783,
      "grad_norm": 1.6272402064271356,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 1.0231,
      "step": 25
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 1.5586320099924926,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 1.0081,
      "step": 30
    },
    {
      "epoch": 0.47297297297297297,
      "grad_norm": 1.495486561607122,
      "learning_rate": 4.72972972972973e-05,
      "loss": 0.9979,
      "step": 35
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.5561925026806689,
      "learning_rate": 4.9997978016429605e-05,
      "loss": 1.0,
      "step": 40
    },
    {
      "epoch": 0.6081081081081081,
      "grad_norm": 1.2919640366032352,
      "learning_rate": 4.9985622766211935e-05,
      "loss": 0.9753,
      "step": 45
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 1.1989455363385786,
      "learning_rate": 4.996204175076325e-05,
      "loss": 0.962,
      "step": 50
    },
    {
      "epoch": 0.7432432432432432,
      "grad_norm": 1.3974962823237607,
      "learning_rate": 4.99272467427147e-05,
      "loss": 0.9677,
      "step": 55
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.0651268932789728,
      "learning_rate": 4.9881255113192526e-05,
      "loss": 0.9513,
      "step": 60
    },
    {
      "epoch": 0.8783783783783784,
      "grad_norm": 1.155576062356755,
      "learning_rate": 4.982408982314565e-05,
      "loss": 0.9619,
      "step": 65
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 1.153407826787955,
      "learning_rate": 4.975577941188258e-05,
      "loss": 0.9535,
      "step": 70
    },
    {
      "epoch": 1.0135135135135136,
      "grad_norm": 1.3775017230764597,
      "learning_rate": 4.967635798282344e-05,
      "loss": 0.9128,
      "step": 75
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 1.1598834626986259,
      "learning_rate": 4.958586518647407e-05,
      "loss": 0.8023,
      "step": 80
    },
    {
      "epoch": 1.1486486486486487,
      "grad_norm": 1.1199196718082898,
      "learning_rate": 4.9484346200630855e-05,
      "loss": 0.7744,
      "step": 85
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 1.1102270321512686,
      "learning_rate": 4.937185170782607e-05,
      "loss": 0.7707,
      "step": 90
    },
    {
      "epoch": 1.2837837837837838,
      "grad_norm": 1.0203324363707762,
      "learning_rate": 4.9248437870025035e-05,
      "loss": 0.7768,
      "step": 95
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 0.9984006736146105,
      "learning_rate": 4.911416630058772e-05,
      "loss": 0.7818,
      "step": 100
    },
    {
      "epoch": 1.4189189189189189,
      "grad_norm": 1.0288042499970187,
      "learning_rate": 4.896910403350873e-05,
      "loss": 0.7864,
      "step": 105
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 1.0840779872937458,
      "learning_rate": 4.88133234899512e-05,
      "loss": 0.7836,
      "step": 110
    },
    {
      "epoch": 1.554054054054054,
      "grad_norm": 1.0695079470515958,
      "learning_rate": 4.864690244209105e-05,
      "loss": 0.7957,
      "step": 115
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.103923214171777,
      "learning_rate": 4.8469923974289874e-05,
      "loss": 0.7943,
      "step": 120
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 0.9711733919756039,
      "learning_rate": 4.828247644161577e-05,
      "loss": 0.7776,
      "step": 125
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 1.064368713968981,
      "learning_rate": 4.808465342573274e-05,
      "loss": 0.7932,
      "step": 130
    },
    {
      "epoch": 1.8243243243243243,
      "grad_norm": 1.0013597653110469,
      "learning_rate": 4.787655368818087e-05,
      "loss": 0.7923,
      "step": 135
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.9961148156869974,
      "learning_rate": 4.765828112107034e-05,
      "loss": 0.7802,
      "step": 140
    },
    {
      "epoch": 1.9594594594594594,
      "grad_norm": 1.037903883462848,
      "learning_rate": 4.742994469521421e-05,
      "loss": 0.7854,
      "step": 145
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 1.2622580047357654,
      "learning_rate": 4.719165840572557e-05,
      "loss": 0.7063,
      "step": 150
    },
    {
      "epoch": 2.0945945945945947,
      "grad_norm": 1.5388907462936379,
      "learning_rate": 4.694354121510644e-05,
      "loss": 0.5487,
      "step": 155
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 1.0947103494672035,
      "learning_rate": 4.668571699385668e-05,
      "loss": 0.5219,
      "step": 160
    },
    {
      "epoch": 2.22972972972973,
      "grad_norm": 1.1651788766706233,
      "learning_rate": 4.641831445863265e-05,
      "loss": 0.5231,
      "step": 165
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 1.1521096427634065,
      "learning_rate": 4.614146710798645e-05,
      "loss": 0.5234,
      "step": 170
    },
    {
      "epoch": 2.364864864864865,
      "grad_norm": 1.1421748306651989,
      "learning_rate": 4.585531315571788e-05,
      "loss": 0.5294,
      "step": 175
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 1.1514841731143928,
      "learning_rate": 4.555999546187229e-05,
      "loss": 0.5381,
      "step": 180
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.0852945108986,
      "learning_rate": 4.5255661461418854e-05,
      "loss": 0.5493,
      "step": 185
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 1.1684382683077765,
      "learning_rate": 4.4942463090644896e-05,
      "loss": 0.5525,
      "step": 190
    },
    {
      "epoch": 2.635135135135135,
      "grad_norm": 1.0788294127649594,
      "learning_rate": 4.462055671130289e-05,
      "loss": 0.5546,
      "step": 195
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 1.0795463247201529,
      "learning_rate": 4.4290103032548094e-05,
      "loss": 0.5619,
      "step": 200
    },
    {
      "epoch": 2.77027027027027,
      "grad_norm": 1.016128366175836,
      "learning_rate": 4.395126703070589e-05,
      "loss": 0.5614,
      "step": 205
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 1.0844580663640953,
      "learning_rate": 4.360421786690862e-05,
      "loss": 0.562,
      "step": 210
    },
    {
      "epoch": 2.9054054054054053,
      "grad_norm": 1.1432810387672725,
      "learning_rate": 4.324912880264326e-05,
      "loss": 0.576,
      "step": 215
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 1.0380987391797107,
      "learning_rate": 4.288617711325207e-05,
      "loss": 0.5697,
      "step": 220
    },
    {
      "epoch": 3.0405405405405403,
      "grad_norm": 1.2503070135517047,
      "learning_rate": 4.251554399942928e-05,
      "loss": 0.4353,
      "step": 225
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 1.6777536935067219,
      "learning_rate": 4.21374144967581e-05,
      "loss": 0.3142,
      "step": 230
    },
    {
      "epoch": 3.175675675675676,
      "grad_norm": 1.1024202684426092,
      "learning_rate": 4.1751977383333224e-05,
      "loss": 0.3107,
      "step": 235
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 1.1910388632740423,
      "learning_rate": 4.1359425085514906e-05,
      "loss": 0.3036,
      "step": 240
    },
    {
      "epoch": 3.310810810810811,
      "grad_norm": 1.231739603690704,
      "learning_rate": 4.095995358186162e-05,
      "loss": 0.3016,
      "step": 245
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 1.1140736912933251,
      "learning_rate": 4.055376230528936e-05,
      "loss": 0.3056,
      "step": 250
    },
    {
      "epoch": 3.445945945945946,
      "grad_norm": 1.1025657791763896,
      "learning_rate": 4.0141054043506406e-05,
      "loss": 0.3185,
      "step": 255
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 1.1791795190101504,
      "learning_rate": 3.972203483777315e-05,
      "loss": 0.3223,
      "step": 260
    },
    {
      "epoch": 3.581081081081081,
      "grad_norm": 1.1814268159199415,
      "learning_rate": 3.929691388003772e-05,
      "loss": 0.3282,
      "step": 265
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 1.2187792717576609,
      "learning_rate": 3.886590340849852e-05,
      "loss": 0.3282,
      "step": 270
    },
    {
      "epoch": 3.7162162162162162,
      "grad_norm": 1.0539496180965828,
      "learning_rate": 3.842921860164607e-05,
      "loss": 0.3317,
      "step": 275
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 1.0934561034346522,
      "learning_rate": 3.798707747083694e-05,
      "loss": 0.3284,
      "step": 280
    },
    {
      "epoch": 3.8513513513513513,
      "grad_norm": 1.1471678068019302,
      "learning_rate": 3.753970075145322e-05,
      "loss": 0.3371,
      "step": 285
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 6.758529885837581,
      "learning_rate": 3.7087311792702265e-05,
      "loss": 0.3967,
      "step": 290
    },
    {
      "epoch": 3.9864864864864864,
      "grad_norm": 1.179358411267412,
      "learning_rate": 3.663013644611139e-05,
      "loss": 0.3428,
      "step": 295
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 1.0255720909850174,
      "learning_rate": 3.616840295277328e-05,
      "loss": 0.2063,
      "step": 300
    },
    {
      "epoch": 4.121621621621622,
      "grad_norm": 1.1681577021267184,
      "learning_rate": 3.5702341829398525e-05,
      "loss": 0.1656,
      "step": 305
    },
    {
      "epoch": 4.1891891891891895,
      "grad_norm": 1.016291058117435,
      "learning_rate": 3.523218575323198e-05,
      "loss": 0.1664,
      "step": 310
    },
    {
      "epoch": 4.256756756756757,
      "grad_norm": 0.9723982730571757,
      "learning_rate": 3.475816944589058e-05,
      "loss": 0.162,
      "step": 315
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 0.990019839692624,
      "learning_rate": 3.4280529556180404e-05,
      "loss": 0.1607,
      "step": 320
    },
    {
      "epoch": 4.391891891891892,
      "grad_norm": 1.067612756627841,
      "learning_rate": 3.379950454195172e-05,
      "loss": 0.163,
      "step": 325
    },
    {
      "epoch": 4.45945945945946,
      "grad_norm": 1.0726343998852998,
      "learning_rate": 3.331533455105084e-05,
      "loss": 0.1709,
      "step": 330
    },
    {
      "epoch": 4.527027027027027,
      "grad_norm": 1.1256065554143042,
      "learning_rate": 3.2828261301428206e-05,
      "loss": 0.1717,
      "step": 335
    },
    {
      "epoch": 4.594594594594595,
      "grad_norm": 1.1551746250455754,
      "learning_rate": 3.23385279604627e-05,
      "loss": 0.1728,
      "step": 340
    },
    {
      "epoch": 4.662162162162162,
      "grad_norm": 1.1309776267452107,
      "learning_rate": 3.18463790235623e-05,
      "loss": 0.1698,
      "step": 345
    },
    {
      "epoch": 4.72972972972973,
      "grad_norm": 1.0126763046166452,
      "learning_rate": 3.135206019210167e-05,
      "loss": 0.1794,
      "step": 350
    },
    {
      "epoch": 4.797297297297297,
      "grad_norm": 1.0529263553265555,
      "learning_rate": 3.085581825075782e-05,
      "loss": 0.1756,
      "step": 355
    },
    {
      "epoch": 4.864864864864865,
      "grad_norm": 1.0824352188484838,
      "learning_rate": 3.0357900944304774e-05,
      "loss": 0.1746,
      "step": 360
    },
    {
      "epoch": 4.9324324324324325,
      "grad_norm": 1.0284704574724253,
      "learning_rate": 2.9858556853929048e-05,
      "loss": 0.1794,
      "step": 365
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0663445785184333,
      "learning_rate": 2.9358035273127483e-05,
      "loss": 0.1758,
      "step": 370
    },
    {
      "epoch": 5.0675675675675675,
      "grad_norm": 0.8110449745410689,
      "learning_rate": 2.8856586083249487e-05,
      "loss": 0.0799,
      "step": 375
    },
    {
      "epoch": 5.135135135135135,
      "grad_norm": 0.8986402275236202,
      "learning_rate": 2.83544596287458e-05,
      "loss": 0.0808,
      "step": 380
    },
    {
      "epoch": 5.202702702702703,
      "grad_norm": 0.9006038868524785,
      "learning_rate": 2.785190659218604e-05,
      "loss": 0.0766,
      "step": 385
    },
    {
      "epoch": 5.27027027027027,
      "grad_norm": 0.8349550048269416,
      "learning_rate": 2.7349177869107462e-05,
      "loss": 0.0771,
      "step": 390
    },
    {
      "epoch": 5.337837837837838,
      "grad_norm": 0.7752306018868771,
      "learning_rate": 2.684652444275741e-05,
      "loss": 0.0782,
      "step": 395
    },
    {
      "epoch": 5.405405405405405,
      "grad_norm": 0.7897117962987327,
      "learning_rate": 2.634419725879193e-05,
      "loss": 0.0775,
      "step": 400
    },
    {
      "epoch": 5.472972972972973,
      "grad_norm": 0.8055715103673516,
      "learning_rate": 2.58424470999932e-05,
      "loss": 0.0783,
      "step": 405
    },
    {
      "epoch": 5.54054054054054,
      "grad_norm": 0.9037399195717022,
      "learning_rate": 2.534152446106825e-05,
      "loss": 0.0762,
      "step": 410
    },
    {
      "epoch": 5.608108108108108,
      "grad_norm": 0.8301708524422988,
      "learning_rate": 2.4841679423591523e-05,
      "loss": 0.0762,
      "step": 415
    },
    {
      "epoch": 5.675675675675675,
      "grad_norm": 0.8143273116216752,
      "learning_rate": 2.4343161531153647e-05,
      "loss": 0.0781,
      "step": 420
    },
    {
      "epoch": 5.743243243243243,
      "grad_norm": 0.8439330014425397,
      "learning_rate": 2.3846219664778824e-05,
      "loss": 0.0793,
      "step": 425
    },
    {
      "epoch": 5.8108108108108105,
      "grad_norm": 0.8142776194662777,
      "learning_rate": 2.3351101918672985e-05,
      "loss": 0.0795,
      "step": 430
    },
    {
      "epoch": 5.878378378378378,
      "grad_norm": 0.8039988882173097,
      "learning_rate": 2.2858055476364822e-05,
      "loss": 0.0784,
      "step": 435
    },
    {
      "epoch": 5.945945945945946,
      "grad_norm": 0.7425653621209469,
      "learning_rate": 2.2367326487301317e-05,
      "loss": 0.0775,
      "step": 440
    },
    {
      "epoch": 6.013513513513513,
      "grad_norm": 0.5266747729292991,
      "learning_rate": 2.1879159943959686e-05,
      "loss": 0.0698,
      "step": 445
    },
    {
      "epoch": 6.081081081081081,
      "grad_norm": 0.5903636908264376,
      "learning_rate": 2.139379955953686e-05,
      "loss": 0.037,
      "step": 450
    },
    {
      "epoch": 6.148648648648648,
      "grad_norm": 0.7057760060859144,
      "learning_rate": 2.0911487646277623e-05,
      "loss": 0.0354,
      "step": 455
    },
    {
      "epoch": 6.216216216216216,
      "grad_norm": 0.6615031551157701,
      "learning_rate": 2.0432464994502203e-05,
      "loss": 0.0338,
      "step": 460
    },
    {
      "epoch": 6.283783783783784,
      "grad_norm": 0.6697026743057446,
      "learning_rate": 1.995697075239365e-05,
      "loss": 0.0333,
      "step": 465
    },
    {
      "epoch": 6.351351351351352,
      "grad_norm": 0.5843277936844908,
      "learning_rate": 1.9485242306605028e-05,
      "loss": 0.0344,
      "step": 470
    },
    {
      "epoch": 6.418918918918919,
      "grad_norm": 0.6767666813757078,
      "learning_rate": 1.9017515163746058e-05,
      "loss": 0.0333,
      "step": 475
    },
    {
      "epoch": 6.486486486486487,
      "grad_norm": 0.612009036541925,
      "learning_rate": 1.855402283280836e-05,
      "loss": 0.0335,
      "step": 480
    },
    {
      "epoch": 6.554054054054054,
      "grad_norm": 0.6242119830720047,
      "learning_rate": 1.8094996708587958e-05,
      "loss": 0.0312,
      "step": 485
    },
    {
      "epoch": 6.621621621621622,
      "grad_norm": 0.6635428778841779,
      "learning_rate": 1.7640665956163306e-05,
      "loss": 0.0343,
      "step": 490
    },
    {
      "epoch": 6.6891891891891895,
      "grad_norm": 0.5640405200429803,
      "learning_rate": 1.719125739648648e-05,
      "loss": 0.0331,
      "step": 495
    },
    {
      "epoch": 6.756756756756757,
      "grad_norm": 0.5504509159363488,
      "learning_rate": 1.6746995393144668e-05,
      "loss": 0.0302,
      "step": 500
    },
    {
      "epoch": 6.824324324324325,
      "grad_norm": 0.6458168054375821,
      "learning_rate": 1.6308101740348433e-05,
      "loss": 0.0328,
      "step": 505
    },
    {
      "epoch": 6.891891891891892,
      "grad_norm": 0.5736755972086186,
      "learning_rate": 1.5874795552202773e-05,
      "loss": 0.0297,
      "step": 510
    },
    {
      "epoch": 6.95945945945946,
      "grad_norm": 0.5847553965309075,
      "learning_rate": 1.5447293153316163e-05,
      "loss": 0.0322,
      "step": 515
    },
    {
      "epoch": 7.027027027027027,
      "grad_norm": 0.3234292566878911,
      "learning_rate": 1.5025807970802252e-05,
      "loss": 0.0237,
      "step": 520
    },
    {
      "epoch": 7.094594594594595,
      "grad_norm": 0.3229946666578016,
      "learning_rate": 1.4610550427728103e-05,
      "loss": 0.0127,
      "step": 525
    },
    {
      "epoch": 7.162162162162162,
      "grad_norm": 0.39666829499344397,
      "learning_rate": 1.4201727838062181e-05,
      "loss": 0.0136,
      "step": 530
    },
    {
      "epoch": 7.22972972972973,
      "grad_norm": 0.4146267850301746,
      "learning_rate": 1.3799544303174514e-05,
      "loss": 0.0129,
      "step": 535
    },
    {
      "epoch": 7.297297297297297,
      "grad_norm": 0.4006318813346767,
      "learning_rate": 1.3404200609940754e-05,
      "loss": 0.0125,
      "step": 540
    },
    {
      "epoch": 7.364864864864865,
      "grad_norm": 0.32483290169015805,
      "learning_rate": 1.3015894130500977e-05,
      "loss": 0.0121,
      "step": 545
    },
    {
      "epoch": 7.4324324324324325,
      "grad_norm": 0.38263371508283517,
      "learning_rate": 1.2634818723723174e-05,
      "loss": 0.0123,
      "step": 550
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.3584723506304905,
      "learning_rate": 1.2261164638420832e-05,
      "loss": 0.0117,
      "step": 555
    },
    {
      "epoch": 7.5675675675675675,
      "grad_norm": 0.32948849737072256,
      "learning_rate": 1.1895118418372734e-05,
      "loss": 0.012,
      "step": 560
    },
    {
      "epoch": 7.635135135135135,
      "grad_norm": 0.3554679755053725,
      "learning_rate": 1.1536862809192518e-05,
      "loss": 0.0123,
      "step": 565
    },
    {
      "epoch": 7.702702702702703,
      "grad_norm": 0.38755695040762,
      "learning_rate": 1.1186576667094342e-05,
      "loss": 0.0126,
      "step": 570
    },
    {
      "epoch": 7.77027027027027,
      "grad_norm": 0.3596675709446449,
      "learning_rate": 1.0844434869600428e-05,
      "loss": 0.0119,
      "step": 575
    },
    {
      "epoch": 7.837837837837838,
      "grad_norm": 0.368664013370026,
      "learning_rate": 1.0510608228234848e-05,
      "loss": 0.0121,
      "step": 580
    },
    {
      "epoch": 7.905405405405405,
      "grad_norm": 0.3689497639593192,
      "learning_rate": 1.0185263403247256e-05,
      "loss": 0.0121,
      "step": 585
    },
    {
      "epoch": 7.972972972972973,
      "grad_norm": 0.37329998652806745,
      "learning_rate": 9.868562820409103e-06,
      "loss": 0.0116,
      "step": 590
    },
    {
      "epoch": 8.04054054054054,
      "grad_norm": 0.17516646949984094,
      "learning_rate": 9.560664589923895e-06,
      "loss": 0.0081,
      "step": 595
    },
    {
      "epoch": 8.108108108108109,
      "grad_norm": 0.19701860916567984,
      "learning_rate": 9.261722427491953e-06,
      "loss": 0.0053,
      "step": 600
    },
    {
      "epoch": 8.175675675675675,
      "grad_norm": 0.16761504423339055,
      "learning_rate": 8.971885577569058e-06,
      "loss": 0.0047,
      "step": 605
    },
    {
      "epoch": 8.243243243243244,
      "grad_norm": 0.23172610604646424,
      "learning_rate": 8.691298738857432e-06,
      "loss": 0.0044,
      "step": 610
    },
    {
      "epoch": 8.31081081081081,
      "grad_norm": 0.1734154562778805,
      "learning_rate": 8.420101992066028e-06,
      "loss": 0.004,
      "step": 615
    },
    {
      "epoch": 8.378378378378379,
      "grad_norm": 0.18641414092933023,
      "learning_rate": 8.158430729976372e-06,
      "loss": 0.0043,
      "step": 620
    },
    {
      "epoch": 8.445945945945946,
      "grad_norm": 0.2824274655739082,
      "learning_rate": 7.906415589848834e-06,
      "loss": 0.0045,
      "step": 625
    },
    {
      "epoch": 8.513513513513514,
      "grad_norm": 0.16976044334904716,
      "learning_rate": 7.664182388203037e-06,
      "loss": 0.0046,
      "step": 630
    },
    {
      "epoch": 8.58108108108108,
      "grad_norm": 0.24818454585022467,
      "learning_rate": 7.4318520580049444e-06,
      "loss": 0.004,
      "step": 635
    },
    {
      "epoch": 8.64864864864865,
      "grad_norm": 0.18073174531439173,
      "learning_rate": 7.209540588292083e-06,
      "loss": 0.0039,
      "step": 640
    },
    {
      "epoch": 8.716216216216216,
      "grad_norm": 0.23113449169489975,
      "learning_rate": 6.9973589662669455e-06,
      "loss": 0.0043,
      "step": 645
    },
    {
      "epoch": 8.783783783783784,
      "grad_norm": 0.2779198237595732,
      "learning_rate": 6.7954131218875404e-06,
      "loss": 0.0046,
      "step": 650
    },
    {
      "epoch": 8.85135135135135,
      "grad_norm": 0.21502404938318828,
      "learning_rate": 6.603803874982687e-06,
      "loss": 0.0039,
      "step": 655
    },
    {
      "epoch": 8.91891891891892,
      "grad_norm": 0.2706988607378338,
      "learning_rate": 6.422626884918559e-06,
      "loss": 0.0043,
      "step": 660
    },
    {
      "epoch": 8.986486486486486,
      "grad_norm": 0.19070682071375314,
      "learning_rate": 6.2519726028415145e-06,
      "loss": 0.0041,
      "step": 665
    },
    {
      "epoch": 9.054054054054054,
      "grad_norm": 0.13080557115625388,
      "learning_rate": 6.091926226521089e-06,
      "loss": 0.0025,
      "step": 670
    },
    {
      "epoch": 9.121621621621621,
      "grad_norm": 0.1076753182673924,
      "learning_rate": 5.942567657815696e-06,
      "loss": 0.0021,
      "step": 675
    },
    {
      "epoch": 9.18918918918919,
      "grad_norm": 0.1002126317760094,
      "learning_rate": 5.8039714627822754e-06,
      "loss": 0.0018,
      "step": 680
    },
    {
      "epoch": 9.256756756756756,
      "grad_norm": 0.12775201749030246,
      "learning_rate": 5.676206834449797e-06,
      "loss": 0.0021,
      "step": 685
    },
    {
      "epoch": 9.324324324324325,
      "grad_norm": 0.07863769364063682,
      "learning_rate": 5.55933755827518e-06,
      "loss": 0.0016,
      "step": 690
    },
    {
      "epoch": 9.391891891891891,
      "grad_norm": 0.12254710433245468,
      "learning_rate": 5.453421980298957e-06,
      "loss": 0.0019,
      "step": 695
    },
    {
      "epoch": 9.45945945945946,
      "grad_norm": 0.13332972976618485,
      "learning_rate": 5.358512978016445e-06,
      "loss": 0.0018,
      "step": 700
    },
    {
      "epoch": 9.527027027027026,
      "grad_norm": 0.12204772827962151,
      "learning_rate": 5.27465793397911e-06,
      "loss": 0.0019,
      "step": 705
    },
    {
      "epoch": 9.594594594594595,
      "grad_norm": 0.25418976259049003,
      "learning_rate": 5.201898712139201e-06,
      "loss": 0.0026,
      "step": 710
    },
    {
      "epoch": 9.662162162162161,
      "grad_norm": 0.12072355515085566,
      "learning_rate": 5.1402716369495194e-06,
      "loss": 0.0018,
      "step": 715
    },
    {
      "epoch": 9.72972972972973,
      "grad_norm": 0.1155679459376018,
      "learning_rate": 5.089807475228711e-06,
      "loss": 0.0021,
      "step": 720
    },
    {
      "epoch": 9.797297297297296,
      "grad_norm": 0.17751733060453512,
      "learning_rate": 5.050531420801205e-06,
      "loss": 0.002,
      "step": 725
    },
    {
      "epoch": 9.864864864864865,
      "grad_norm": 0.08558684167277145,
      "learning_rate": 5.022463081919386e-06,
      "loss": 0.002,
      "step": 730
    },
    {
      "epoch": 9.932432432432432,
      "grad_norm": 0.09909189210286934,
      "learning_rate": 5.005616471474332e-06,
      "loss": 0.0017,
      "step": 735
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.09536830165909363,
      "learning_rate": 5e-06,
      "loss": 0.0018,
      "step": 740
    },
    {
      "epoch": 10.0,
      "step": 740,
      "total_flos": 1440572837462016.0,
      "train_loss": 0.301153065278069,
      "train_runtime": 18722.0332,
      "train_samples_per_second": 5.053,
      "train_steps_per_second": 0.04
    }
  ],
  "logging_steps": 5,
  "max_steps": 740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1440572837462016.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}