{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 837,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03590664272890485,
      "grad_norm": 2.185314827906911,
      "learning_rate": 1.0714285714285714e-06,
      "loss": 0.5255,
      "step": 10
    },
    {
      "epoch": 0.0718132854578097,
      "grad_norm": 0.9772670854647908,
      "learning_rate": 2.261904761904762e-06,
      "loss": 0.4864,
      "step": 20
    },
    {
      "epoch": 0.10771992818671454,
      "grad_norm": 0.5510959017484935,
      "learning_rate": 3.4523809523809528e-06,
      "loss": 0.4451,
      "step": 30
    },
    {
      "epoch": 0.1436265709156194,
      "grad_norm": 0.36493152532802847,
      "learning_rate": 4.642857142857144e-06,
      "loss": 0.4149,
      "step": 40
    },
    {
      "epoch": 0.17953321364452424,
      "grad_norm": 0.2869233938041407,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.3955,
      "step": 50
    },
    {
      "epoch": 0.21543985637342908,
      "grad_norm": 0.2374087352719959,
      "learning_rate": 7.023809523809524e-06,
      "loss": 0.3807,
      "step": 60
    },
    {
      "epoch": 0.2513464991023339,
      "grad_norm": 0.19085750578643032,
      "learning_rate": 8.214285714285714e-06,
      "loss": 0.3719,
      "step": 70
    },
    {
      "epoch": 0.2872531418312388,
      "grad_norm": 0.18588659208121952,
      "learning_rate": 9.404761904761905e-06,
      "loss": 0.3638,
      "step": 80
    },
    {
      "epoch": 0.3231597845601436,
      "grad_norm": 0.19078133253606472,
      "learning_rate": 9.998912137362367e-06,
      "loss": 0.3602,
      "step": 90
    },
    {
      "epoch": 0.3590664272890485,
      "grad_norm": 0.18585754435457444,
      "learning_rate": 9.990212076323587e-06,
      "loss": 0.357,
      "step": 100
    },
    {
      "epoch": 0.39497307001795334,
      "grad_norm": 0.1879818930694756,
      "learning_rate": 9.97282709575282e-06,
      "loss": 0.3511,
      "step": 110
    },
    {
      "epoch": 0.43087971274685816,
      "grad_norm": 0.17595485058415583,
      "learning_rate": 9.946787452311507e-06,
      "loss": 0.3474,
      "step": 120
    },
    {
      "epoch": 0.466786355475763,
      "grad_norm": 0.18168043823227237,
      "learning_rate": 9.912138465157325e-06,
      "loss": 0.3466,
      "step": 130
    },
    {
      "epoch": 0.5026929982046678,
      "grad_norm": 0.19246537631691138,
      "learning_rate": 9.86894043707114e-06,
      "loss": 0.3439,
      "step": 140
    },
    {
      "epoch": 0.5385996409335727,
      "grad_norm": 0.20913120656523562,
      "learning_rate": 9.81726854950659e-06,
      "loss": 0.3392,
      "step": 150
    },
    {
      "epoch": 0.5745062836624776,
      "grad_norm": 0.1943925752607722,
      "learning_rate": 9.757212731744973e-06,
      "loss": 0.3382,
      "step": 160
    },
    {
      "epoch": 0.6104129263913824,
      "grad_norm": 0.22384055375769804,
      "learning_rate": 9.688877504383158e-06,
      "loss": 0.3386,
      "step": 170
    },
    {
      "epoch": 0.6463195691202872,
      "grad_norm": 0.19509930129044936,
      "learning_rate": 9.612381797426874e-06,
      "loss": 0.3378,
      "step": 180
    },
    {
      "epoch": 0.6822262118491921,
      "grad_norm": 0.23177190223223015,
      "learning_rate": 9.52785874330602e-06,
      "loss": 0.333,
      "step": 190
    },
    {
      "epoch": 0.718132854578097,
      "grad_norm": 0.2006246018340954,
      "learning_rate": 9.435455445172214e-06,
      "loss": 0.3311,
      "step": 200
    },
    {
      "epoch": 0.7540394973070018,
      "grad_norm": 0.2042999512470892,
      "learning_rate": 9.33533272088179e-06,
      "loss": 0.3313,
      "step": 210
    },
    {
      "epoch": 0.7899461400359067,
      "grad_norm": 0.21930626870236752,
      "learning_rate": 9.227664823109884e-06,
      "loss": 0.3325,
      "step": 220
    },
    {
      "epoch": 0.8258527827648114,
      "grad_norm": 0.23685553683202806,
      "learning_rate": 9.11263913608266e-06,
      "loss": 0.3308,
      "step": 230
    },
    {
      "epoch": 0.8617594254937163,
      "grad_norm": 0.21203795699103023,
      "learning_rate": 8.990455849455522e-06,
      "loss": 0.3306,
      "step": 240
    },
    {
      "epoch": 0.8976660682226212,
      "grad_norm": 0.2617908535020091,
      "learning_rate": 8.861327609904859e-06,
      "loss": 0.328,
      "step": 250
    },
    {
      "epoch": 0.933572710951526,
      "grad_norm": 0.20428110286245318,
      "learning_rate": 8.725479151039714e-06,
      "loss": 0.3249,
      "step": 260
    },
    {
      "epoch": 0.9694793536804309,
      "grad_norm": 0.23276915935689596,
      "learning_rate": 8.583146902277464e-06,
      "loss": 0.3287,
      "step": 270
    },
    {
      "epoch": 1.0035906642728905,
      "grad_norm": 0.2445973206934874,
      "learning_rate": 8.434578577364218e-06,
      "loss": 0.3246,
      "step": 280
    },
    {
      "epoch": 1.0394973070017954,
      "grad_norm": 0.21984748841500337,
      "learning_rate": 8.28003274325608e-06,
      "loss": 0.3196,
      "step": 290
    },
    {
      "epoch": 1.0754039497307002,
      "grad_norm": 0.21007463727725653,
      "learning_rate": 8.119778370111566e-06,
      "loss": 0.3129,
      "step": 300
    },
    {
      "epoch": 1.111310592459605,
      "grad_norm": 0.22501623459913725,
      "learning_rate": 7.954094363178421e-06,
      "loss": 0.3145,
      "step": 310
    },
    {
      "epoch": 1.14721723518851,
      "grad_norm": 0.20904438470673392,
      "learning_rate": 7.783269077389447e-06,
      "loss": 0.3132,
      "step": 320
    },
    {
      "epoch": 1.1831238779174147,
      "grad_norm": 0.1929405792634159,
      "learning_rate": 7.607599815512226e-06,
      "loss": 0.3119,
      "step": 330
    },
    {
      "epoch": 1.2190305206463194,
      "grad_norm": 0.20576378490371444,
      "learning_rate": 7.427392310726088e-06,
      "loss": 0.3145,
      "step": 340
    },
    {
      "epoch": 1.2549371633752244,
      "grad_norm": 0.22314438193898609,
      "learning_rate": 7.242960194526893e-06,
      "loss": 0.314,
      "step": 350
    },
    {
      "epoch": 1.2908438061041292,
      "grad_norm": 0.18526094334512155,
      "learning_rate": 7.054624450885621e-06,
      "loss": 0.3147,
      "step": 360
    },
    {
      "epoch": 1.3267504488330342,
      "grad_norm": 0.20944522582021186,
      "learning_rate": 6.862712857610812e-06,
      "loss": 0.3125,
      "step": 370
    },
    {
      "epoch": 1.362657091561939,
      "grad_norm": 0.20970244825805384,
      "learning_rate": 6.667559415887055e-06,
      "loss": 0.3119,
      "step": 380
    },
    {
      "epoch": 1.398563734290844,
      "grad_norm": 0.18929491036558882,
      "learning_rate": 6.469503768982379e-06,
      "loss": 0.3144,
      "step": 390
    },
    {
      "epoch": 1.4344703770197487,
      "grad_norm": 0.18469452603163652,
      "learning_rate": 6.2688906111362115e-06,
      "loss": 0.3131,
      "step": 400
    },
    {
      "epoch": 1.4703770197486534,
      "grad_norm": 0.18021217056496155,
      "learning_rate": 6.066069087656665e-06,
      "loss": 0.3128,
      "step": 410
    },
    {
      "epoch": 1.5062836624775584,
      "grad_norm": 0.31157448102210894,
      "learning_rate": 5.8613921872712446e-06,
      "loss": 0.3115,
      "step": 420
    },
    {
      "epoch": 1.5421903052064632,
      "grad_norm": 0.19847526072216384,
      "learning_rate": 5.655216127788472e-06,
      "loss": 0.3088,
      "step": 430
    },
    {
      "epoch": 1.578096947935368,
      "grad_norm": 0.19215915923161053,
      "learning_rate": 5.447899736139676e-06,
      "loss": 0.3063,
      "step": 440
    },
    {
      "epoch": 1.614003590664273,
      "grad_norm": 0.18221395113654418,
      "learning_rate": 5.239803823879878e-06,
      "loss": 0.3082,
      "step": 450
    },
    {
      "epoch": 1.6499102333931779,
      "grad_norm": 0.1749484556213854,
      "learning_rate": 5.03129055923465e-06,
      "loss": 0.3089,
      "step": 460
    },
    {
      "epoch": 1.6858168761220824,
      "grad_norm": 0.17613434487474727,
      "learning_rate": 4.822722836785842e-06,
      "loss": 0.3068,
      "step": 470
    },
    {
      "epoch": 1.7217235188509874,
      "grad_norm": 0.1751185479216866,
      "learning_rate": 4.614463645893175e-06,
      "loss": 0.3068,
      "step": 480
    },
    {
      "epoch": 1.7576301615798924,
      "grad_norm": 0.1748572053302032,
      "learning_rate": 4.4068754389508616e-06,
      "loss": 0.3093,
      "step": 490
    },
    {
      "epoch": 1.7935368043087971,
      "grad_norm": 0.16507354789633685,
      "learning_rate": 4.2003195005787745e-06,
      "loss": 0.3099,
      "step": 500
    },
    {
      "epoch": 1.829443447037702,
      "grad_norm": 0.170392010993257,
      "learning_rate": 3.995155318845994e-06,
      "loss": 0.3074,
      "step": 510
    },
    {
      "epoch": 1.8653500897666069,
      "grad_norm": 0.18627037895031676,
      "learning_rate": 3.791739959621054e-06,
      "loss": 0.3091,
      "step": 520
    },
    {
      "epoch": 1.9012567324955116,
      "grad_norm": 0.17081506891008494,
      "learning_rate": 3.5904274451377634e-06,
      "loss": 0.3071,
      "step": 530
    },
    {
      "epoch": 1.9371633752244164,
      "grad_norm": 0.1697011202581245,
      "learning_rate": 3.3915681378581416e-06,
      "loss": 0.3093,
      "step": 540
    },
    {
      "epoch": 1.9730700179533214,
      "grad_norm": 0.17582268731336928,
      "learning_rate": 3.195508130704795e-06,
      "loss": 0.3072,
      "step": 550
    },
    {
      "epoch": 2.007181328545781,
      "grad_norm": 0.16800018749204126,
      "learning_rate": 3.0025886447239474e-06,
      "loss": 0.3027,
      "step": 560
    },
    {
      "epoch": 2.043087971274686,
      "grad_norm": 0.17959400944310325,
      "learning_rate": 2.8131454352274525e-06,
      "loss": 0.2975,
      "step": 570
    },
    {
      "epoch": 2.078994614003591,
      "grad_norm": 0.1785065928144257,
      "learning_rate": 2.627508207447308e-06,
      "loss": 0.2986,
      "step": 580
    },
    {
      "epoch": 2.1149012567324954,
      "grad_norm": 0.15727463247343326,
      "learning_rate": 2.4460000427196916e-06,
      "loss": 0.2949,
      "step": 590
    },
    {
      "epoch": 2.1508078994614004,
      "grad_norm": 0.15805698099204835,
      "learning_rate": 2.268936836197144e-06,
      "loss": 0.2952,
      "step": 600
    },
    {
      "epoch": 2.1867145421903054,
      "grad_norm": 0.16920663238784117,
      "learning_rate": 2.0966267470675273e-06,
      "loss": 0.2985,
      "step": 610
    },
    {
      "epoch": 2.22262118491921,
      "grad_norm": 0.16646936573182844,
      "learning_rate": 1.929369662236604e-06,
      "loss": 0.2979,
      "step": 620
    },
    {
      "epoch": 2.258527827648115,
      "grad_norm": 0.3421880613089778,
      "learning_rate": 1.7674566744075882e-06,
      "loss": 0.298,
      "step": 630
    },
    {
      "epoch": 2.29443447037702,
      "grad_norm": 0.15025140288330655,
      "learning_rate": 1.6111695754660667e-06,
      "loss": 0.2942,
      "step": 640
    },
    {
      "epoch": 2.3303411131059244,
      "grad_norm": 0.17473486663498805,
      "learning_rate": 1.4607803660519803e-06,
      "loss": 0.2952,
      "step": 650
    },
    {
      "epoch": 2.3662477558348294,
      "grad_norm": 0.1589150679388623,
      "learning_rate": 1.3165507821721906e-06,
      "loss": 0.2983,
      "step": 660
    },
    {
      "epoch": 2.4021543985637344,
      "grad_norm": 0.1703479004667421,
      "learning_rate": 1.1787318396775188e-06,
      "loss": 0.298,
      "step": 670
    },
    {
      "epoch": 2.438061041292639,
      "grad_norm": 0.1513614749600422,
      "learning_rate": 1.0475633973970573e-06,
      "loss": 0.2959,
      "step": 680
    },
    {
      "epoch": 2.473967684021544,
      "grad_norm": 0.1876551965771974,
      "learning_rate": 9.232737396900543e-07,
      "loss": 0.2996,
      "step": 690
    },
    {
      "epoch": 2.509874326750449,
      "grad_norm": 0.15388873373154505,
      "learning_rate": 8.060791791418887e-07,
      "loss": 0.2966,
      "step": 700
    },
    {
      "epoch": 2.545780969479354,
      "grad_norm": 0.15535759307395802,
      "learning_rate": 6.96183680095639e-07,
      "loss": 0.2991,
      "step": 710
    },
    {
      "epoch": 2.5816876122082584,
      "grad_norm": 0.14993765716059407,
      "learning_rate": 5.937785036743893e-07,
      "loss": 0.298,
      "step": 720
    },
    {
      "epoch": 2.6175942549371634,
      "grad_norm": 0.14735943617696928,
      "learning_rate": 4.990418749121179e-07,
      "loss": 0.2966,
      "step": 730
    },
    {
      "epoch": 2.6535008976660683,
      "grad_norm": 0.1548863032393796,
      "learning_rate": 4.121386725724835e-07,
      "loss": 0.2978,
      "step": 740
    },
    {
      "epoch": 2.6894075403949733,
      "grad_norm": 0.15810495711820732,
      "learning_rate": 3.3322014219532575e-07,
      "loss": 0.2976,
      "step": 750
    },
    {
      "epoch": 2.725314183123878,
      "grad_norm": 0.15072810300321354,
      "learning_rate": 2.6242363287030617e-07,
      "loss": 0.296,
      "step": 760
    },
    {
      "epoch": 2.761220825852783,
      "grad_norm": 0.2424841654880629,
      "learning_rate": 1.9987235819581118e-07,
      "loss": 0.2969,
      "step": 770
    },
    {
      "epoch": 2.797127468581688,
      "grad_norm": 0.22116921613107723,
      "learning_rate": 1.4567518183912887e-07,
      "loss": 0.2964,
      "step": 780
    },
    {
      "epoch": 2.8330341113105924,
      "grad_norm": 0.14618317771038028,
      "learning_rate": 9.992642807111486e-08,
      "loss": 0.2978,
      "step": 790
    },
    {
      "epoch": 2.8689407540394973,
      "grad_norm": 0.1636178141701447,
      "learning_rate": 6.270571760509547e-08,
      "loss": 0.2991,
      "step": 800
    },
    {
      "epoch": 2.9048473967684023,
      "grad_norm": 0.14508080741570623,
      "learning_rate": 3.4077829025703226e-08,
      "loss": 0.2942,
      "step": 810
    },
    {
      "epoch": 2.940754039497307,
      "grad_norm": 0.1398003635276662,
      "learning_rate": 1.4092586048820578e-08,
      "loss": 0.2962,
      "step": 820
    },
    {
      "epoch": 2.976660682226212,
      "grad_norm": 0.1587618807983056,
      "learning_rate": 2.784770808839654e-09,
      "loss": 0.2982,
      "step": 830
    }
  ],
  "logging_steps": 10,
  "max_steps": 837,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 10000000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8093209209077760.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}