{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1028,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.019464720194647202,
      "grad_norm": 419.45176209681534,
      "learning_rate": 1.7475728155339808e-06,
      "loss": 7.5347,
      "step": 10
    },
    {
      "epoch": 0.038929440389294405,
      "grad_norm": 440.7543770201198,
      "learning_rate": 3.689320388349515e-06,
      "loss": 5.088,
      "step": 20
    },
    {
      "epoch": 0.058394160583941604,
      "grad_norm": 78.99835155456897,
      "learning_rate": 5.631067961165049e-06,
      "loss": 4.2568,
      "step": 30
    },
    {
      "epoch": 0.07785888077858881,
      "grad_norm": 70.58925751405357,
      "learning_rate": 7.572815533980583e-06,
      "loss": 3.4786,
      "step": 40
    },
    {
      "epoch": 0.09732360097323602,
      "grad_norm": 40.09311796360936,
      "learning_rate": 9.514563106796117e-06,
      "loss": 3.161,
      "step": 50
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 41.72458199935773,
      "learning_rate": 1.145631067961165e-05,
      "loss": 3.104,
      "step": 60
    },
    {
      "epoch": 0.1362530413625304,
      "grad_norm": 149.8142811415225,
      "learning_rate": 1.3398058252427187e-05,
      "loss": 3.5738,
      "step": 70
    },
    {
      "epoch": 0.15571776155717762,
      "grad_norm": 72.86685673927337,
      "learning_rate": 1.533980582524272e-05,
      "loss": 3.0172,
      "step": 80
    },
    {
      "epoch": 0.17518248175182483,
      "grad_norm": 25.03536602970884,
      "learning_rate": 1.7281553398058253e-05,
      "loss": 3.3168,
      "step": 90
    },
    {
      "epoch": 0.19464720194647203,
      "grad_norm": 59.627261461817,
      "learning_rate": 1.922330097087379e-05,
      "loss": 3.4559,
      "step": 100
    },
    {
      "epoch": 0.2141119221411192,
      "grad_norm": 69.6146294827494,
      "learning_rate": 1.999792377815462e-05,
      "loss": 3.0347,
      "step": 110
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 72.85844217141502,
      "learning_rate": 1.9985238877782747e-05,
      "loss": 3.1156,
      "step": 120
    },
    {
      "epoch": 0.25304136253041365,
      "grad_norm": 33.23078781184279,
      "learning_rate": 1.9961037146461106e-05,
      "loss": 3.151,
      "step": 130
    },
    {
      "epoch": 0.2725060827250608,
      "grad_norm": 81.22302004701628,
      "learning_rate": 1.992534649811862e-05,
      "loss": 2.9322,
      "step": 140
    },
    {
      "epoch": 0.291970802919708,
      "grad_norm": 47.33360495853381,
      "learning_rate": 1.9878208097835908e-05,
      "loss": 2.8629,
      "step": 150
    },
    {
      "epoch": 0.31143552311435524,
      "grad_norm": 51.597242465258795,
      "learning_rate": 1.9819676314366084e-05,
      "loss": 2.9709,
      "step": 160
    },
    {
      "epoch": 0.3309002433090024,
      "grad_norm": 82.70654992211028,
      "learning_rate": 1.974981865742661e-05,
      "loss": 2.8736,
      "step": 170
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 129.437045502628,
      "learning_rate": 1.9668715699834553e-05,
      "loss": 3.064,
      "step": 180
    },
    {
      "epoch": 0.36982968369829683,
      "grad_norm": 41.951916677490715,
      "learning_rate": 1.957646098457507e-05,
      "loss": 2.7225,
      "step": 190
    },
    {
      "epoch": 0.38929440389294406,
      "grad_norm": 47.177248106803845,
      "learning_rate": 1.9473160916910268e-05,
      "loss": 2.9907,
      "step": 200
    },
    {
      "epoch": 0.40875912408759124,
      "grad_norm": 28.20808923970595,
      "learning_rate": 1.9358934641652897e-05,
      "loss": 2.8968,
      "step": 210
    },
    {
      "epoch": 0.4282238442822384,
      "grad_norm": 60.66162586721256,
      "learning_rate": 1.923391390574645e-05,
      "loss": 2.9422,
      "step": 220
    },
    {
      "epoch": 0.44768856447688565,
      "grad_norm": 29.677036481098472,
      "learning_rate": 1.909824290631012e-05,
      "loss": 3.0416,
      "step": 230
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 57.8380142267369,
      "learning_rate": 1.8952078124323922e-05,
      "loss": 2.8864,
      "step": 240
    },
    {
      "epoch": 0.48661800486618007,
      "grad_norm": 23.875475400629053,
      "learning_rate": 1.8795588144145784e-05,
      "loss": 2.6229,
      "step": 250
    },
    {
      "epoch": 0.5060827250608273,
      "grad_norm": 25.63414075808507,
      "learning_rate": 1.8628953459068766e-05,
      "loss": 2.764,
      "step": 260
    },
    {
      "epoch": 0.5255474452554745,
      "grad_norm": 50.33174840522443,
      "learning_rate": 1.8452366263142694e-05,
      "loss": 2.711,
      "step": 270
    },
    {
      "epoch": 0.5450121654501217,
      "grad_norm": 40.94875404333496,
      "learning_rate": 1.8266030229500307e-05,
      "loss": 2.8135,
      "step": 280
    },
    {
      "epoch": 0.5644768856447688,
      "grad_norm": 23.116839479892334,
      "learning_rate": 1.807016027544359e-05,
      "loss": 2.782,
      "step": 290
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 18.82174925977683,
      "learning_rate": 1.786498231456125e-05,
      "loss": 2.6532,
      "step": 300
    },
    {
      "epoch": 0.6034063260340633,
      "grad_norm": 22.880354016784672,
      "learning_rate": 1.7650732996163246e-05,
      "loss": 2.8259,
      "step": 310
    },
    {
      "epoch": 0.6228710462287105,
      "grad_norm": 28.332068157479885,
      "learning_rate": 1.7427659432332844e-05,
      "loss": 2.7381,
      "step": 320
    },
    {
      "epoch": 0.6423357664233577,
      "grad_norm": 28.026764464260065,
      "learning_rate": 1.7196018912911126e-05,
      "loss": 2.7222,
      "step": 330
    },
    {
      "epoch": 0.6618004866180048,
      "grad_norm": 21.57301570810096,
      "learning_rate": 1.6956078608742567e-05,
      "loss": 2.722,
      "step": 340
    },
    {
      "epoch": 0.681265206812652,
      "grad_norm": 24.388668227839457,
      "learning_rate": 1.6708115263524047e-05,
      "loss": 2.6107,
      "step": 350
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 18.703736650144812,
      "learning_rate": 1.6452414874612608e-05,
      "loss": 2.593,
      "step": 360
    },
    {
      "epoch": 0.7201946472019465,
      "grad_norm": 31.332283196695716,
      "learning_rate": 1.618927236316026e-05,
      "loss": 2.7911,
      "step": 370
    },
    {
      "epoch": 0.7396593673965937,
      "grad_norm": 23.25616532139509,
      "learning_rate": 1.5918991233956145e-05,
      "loss": 2.7352,
      "step": 380
    },
    {
      "epoch": 0.7591240875912408,
      "grad_norm": 12.936266532050945,
      "learning_rate": 1.5641883225368468e-05,
      "loss": 2.7952,
      "step": 390
    },
    {
      "epoch": 0.7785888077858881,
      "grad_norm": 86.36443076979923,
      "learning_rate": 1.5358267949789968e-05,
      "loss": 3.0855,
      "step": 400
    },
    {
      "epoch": 0.7980535279805353,
      "grad_norm": 46.98416378181883,
      "learning_rate": 1.5068472525001554e-05,
      "loss": 2.8366,
      "step": 410
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 26.85730980671752,
      "learning_rate": 1.4772831196879383e-05,
      "loss": 2.5443,
      "step": 420
    },
    {
      "epoch": 0.8369829683698297,
      "grad_norm": 21.408351349543143,
      "learning_rate": 1.4471684953880458e-05,
      "loss": 2.542,
      "step": 430
    },
    {
      "epoch": 0.8564476885644768,
      "grad_norm": 19.42979950360618,
      "learning_rate": 1.416538113375145e-05,
      "loss": 2.6807,
      "step": 440
    },
    {
      "epoch": 0.8759124087591241,
      "grad_norm": 54.322602669364706,
      "learning_rate": 1.3854273022914333e-05,
      "loss": 2.7384,
      "step": 450
    },
    {
      "epoch": 0.8953771289537713,
      "grad_norm": 29.215560058723682,
      "learning_rate": 1.3538719448990905e-05,
      "loss": 2.6412,
      "step": 460
    },
    {
      "epoch": 0.9148418491484185,
      "grad_norm": 25.527754957600752,
      "learning_rate": 1.3219084366936172e-05,
      "loss": 2.5258,
      "step": 470
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 48.82663237736622,
      "learning_rate": 1.2895736439257933e-05,
      "loss": 2.5599,
      "step": 480
    },
    {
      "epoch": 0.9537712895377128,
      "grad_norm": 27.743882727321346,
      "learning_rate": 1.256904861080674e-05,
      "loss": 2.6751,
      "step": 490
    },
    {
      "epoch": 0.9732360097323601,
      "grad_norm": 27.354218263639577,
      "learning_rate": 1.223939767862668e-05,
      "loss": 2.4962,
      "step": 500
    },
    {
      "epoch": 0.9927007299270073,
      "grad_norm": 25.900686408023237,
      "learning_rate": 1.190716385736307e-05,
      "loss": 2.5036,
      "step": 510
    },
    {
      "epoch": 1.0116788321167882,
      "grad_norm": 19.413128620243246,
      "learning_rate": 1.1572730340728362e-05,
      "loss": 2.0859,
      "step": 520
    },
    {
      "epoch": 1.0311435523114356,
      "grad_norm": 16.59211624117448,
      "learning_rate": 1.1236482859532019e-05,
      "loss": 2.0009,
      "step": 530
    },
    {
      "epoch": 1.0506082725060828,
      "grad_norm": 22.403659201349214,
      "learning_rate": 1.0898809236784152e-05,
      "loss": 2.089,
      "step": 540
    },
    {
      "epoch": 1.07007299270073,
      "grad_norm": 25.224065033796418,
      "learning_rate": 1.0560098940386028e-05,
      "loss": 1.8724,
      "step": 550
    },
    {
      "epoch": 1.0895377128953772,
      "grad_norm": 14.859356677964838,
      "learning_rate": 1.0220742633923393e-05,
      "loss": 1.8878,
      "step": 560
    },
    {
      "epoch": 1.1090024330900243,
      "grad_norm": 17.04142136353786,
      "learning_rate": 9.88113172608072e-06,
      "loss": 1.8977,
      "step": 570
    },
    {
      "epoch": 1.1284671532846715,
      "grad_norm": 13.713070262842116,
      "learning_rate": 9.541657919196049e-06,
      "loss": 1.9069,
      "step": 580
    },
    {
      "epoch": 1.1479318734793187,
      "grad_norm": 17.789458089215415,
      "learning_rate": 9.202712757477145e-06,
      "loss": 1.8461,
      "step": 590
    },
    {
      "epoch": 1.1673965936739659,
      "grad_norm": 19.610482780070903,
      "learning_rate": 8.864687175400045e-06,
      "loss": 1.9158,
      "step": 600
    },
    {
      "epoch": 1.186861313868613,
      "grad_norm": 12.010210734055757,
      "learning_rate": 8.527971046810845e-06,
      "loss": 1.8679,
      "step": 610
    },
    {
      "epoch": 1.2063260340632604,
      "grad_norm": 22.335816412732463,
      "learning_rate": 8.192952735250815e-06,
      "loss": 1.9524,
      "step": 620
    },
    {
      "epoch": 1.2257907542579076,
      "grad_norm": 10.686579448514223,
      "learning_rate": 7.86001864602348e-06,
      "loss": 1.8423,
      "step": 630
    },
    {
      "epoch": 1.2452554744525548,
      "grad_norm": 51.109105653823455,
      "learning_rate": 7.529552780520292e-06,
      "loss": 1.8348,
      "step": 640
    },
    {
      "epoch": 1.264720194647202,
      "grad_norm": 33.402753904185154,
      "learning_rate": 7.201936293318946e-06,
      "loss": 1.8627,
      "step": 650
    },
    {
      "epoch": 1.2841849148418492,
      "grad_norm": 32.98747845535257,
      "learning_rate": 6.877547052565177e-06,
      "loss": 1.9406,
      "step": 660
    },
    {
      "epoch": 1.3036496350364963,
      "grad_norm": 21.46620027744224,
      "learning_rate": 6.556759204145069e-06,
      "loss": 1.9661,
      "step": 670
    },
    {
      "epoch": 1.3231143552311435,
      "grad_norm": 17.85715273280427,
      "learning_rate": 6.239942740150571e-06,
      "loss": 1.941,
      "step": 680
    },
    {
      "epoch": 1.3425790754257907,
      "grad_norm": 16.397413728215973,
      "learning_rate": 5.927463072135936e-06,
      "loss": 1.8711,
      "step": 690
    },
    {
      "epoch": 1.3620437956204379,
      "grad_norm": 36.03813133441638,
      "learning_rate": 5.619680609657294e-06,
      "loss": 1.9032,
      "step": 700
    },
    {
      "epoch": 1.381508515815085,
      "grad_norm": 22.89930004578536,
      "learning_rate": 5.316950344581439e-06,
      "loss": 1.8234,
      "step": 710
    },
    {
      "epoch": 1.4009732360097322,
      "grad_norm": 10.308090714098702,
      "learning_rate": 5.019621441643336e-06,
      "loss": 1.8646,
      "step": 720
    },
    {
      "epoch": 1.4204379562043796,
      "grad_norm": 28.844231344171455,
      "learning_rate": 4.728036835724512e-06,
      "loss": 1.8324,
      "step": 730
    },
    {
      "epoch": 1.4399026763990268,
      "grad_norm": 21.377632574293145,
      "learning_rate": 4.442532836316909e-06,
      "loss": 1.862,
      "step": 740
    },
    {
      "epoch": 1.459367396593674,
      "grad_norm": 17.303686682340725,
      "learning_rate": 4.163438739628359e-06,
      "loss": 1.9171,
      "step": 750
    },
    {
      "epoch": 1.4788321167883212,
      "grad_norm": 37.72983556916591,
      "learning_rate": 3.891076448777046e-06,
      "loss": 1.87,
      "step": 760
    },
    {
      "epoch": 1.4982968369829683,
      "grad_norm": 14.826567789437357,
      "learning_rate": 3.625760102513103e-06,
      "loss": 1.8535,
      "step": 770
    },
    {
      "epoch": 1.5177615571776155,
      "grad_norm": 18.51444546988277,
      "learning_rate": 3.367795712895483e-06,
      "loss": 1.8713,
      "step": 780
    },
    {
      "epoch": 1.537226277372263,
      "grad_norm": 5.383086567424804,
      "learning_rate": 3.117480812342054e-06,
      "loss": 1.7639,
      "step": 790
    },
    {
      "epoch": 1.55669099756691,
      "grad_norm": 26.01516191764228,
      "learning_rate": 2.8751041104599818e-06,
      "loss": 1.757,
      "step": 800
    },
    {
      "epoch": 1.5761557177615573,
      "grad_norm": 25.971694889888745,
      "learning_rate": 2.6409451610522287e-06,
      "loss": 1.8712,
      "step": 810
    },
    {
      "epoch": 1.5956204379562045,
      "grad_norm": 16.826359696774247,
      "learning_rate": 2.4152740396842044e-06,
      "loss": 1.8375,
      "step": 820
    },
    {
      "epoch": 1.6150851581508516,
      "grad_norm": 20.178970467961914,
      "learning_rate": 2.1983510321825053e-06,
      "loss": 1.7858,
      "step": 830
    },
    {
      "epoch": 1.6345498783454988,
      "grad_norm": 17.10427518401472,
      "learning_rate": 1.9904263344249743e-06,
      "loss": 1.7688,
      "step": 840
    },
    {
      "epoch": 1.654014598540146,
      "grad_norm": 26.71032233185063,
      "learning_rate": 1.7917397637683799e-06,
      "loss": 1.7905,
      "step": 850
    },
    {
      "epoch": 1.6734793187347932,
      "grad_norm": 14.410090356585053,
      "learning_rate": 1.602520482446519e-06,
      "loss": 1.8363,
      "step": 860
    },
    {
      "epoch": 1.6929440389294403,
      "grad_norm": 20.353833687299634,
      "learning_rate": 1.4229867332577962e-06,
      "loss": 1.8553,
      "step": 870
    },
    {
      "epoch": 1.7124087591240875,
      "grad_norm": 19.078090551260555,
      "learning_rate": 1.2533455878471158e-06,
      "loss": 1.8013,
      "step": 880
    },
    {
      "epoch": 1.7318734793187347,
      "grad_norm": 15.961285499721651,
      "learning_rate": 1.0937927078724242e-06,
      "loss": 1.7987,
      "step": 890
    },
    {
      "epoch": 1.7513381995133819,
      "grad_norm": 23.255794765895036,
      "learning_rate": 9.445121193313678e-07,
      "loss": 1.7438,
      "step": 900
    },
    {
      "epoch": 1.770802919708029,
      "grad_norm": 15.026924673840433,
      "learning_rate": 8.056760003083519e-07,
      "loss": 1.8305,
      "step": 910
    },
    {
      "epoch": 1.7902676399026762,
      "grad_norm": 33.518276552309146,
      "learning_rate": 6.774444823868153e-07,
      "loss": 1.8236,
      "step": 920
    },
    {
      "epoch": 1.8097323600973236,
      "grad_norm": 19.766136489150675,
      "learning_rate": 5.59965465955763e-07,
      "loss": 1.8234,
      "step": 930
    },
    {
      "epoch": 1.8291970802919708,
      "grad_norm": 33.636630644284125,
      "learning_rate": 4.533744496235859e-07,
      "loss": 1.868,
      "step": 940
    },
    {
      "epoch": 1.848661800486618,
      "grad_norm": 11.42056049503142,
      "learning_rate": 3.577943739359102e-07,
      "loss": 1.7457,
      "step": 950
    },
    {
      "epoch": 1.8681265206812652,
      "grad_norm": 15.746160617813796,
      "learning_rate": 2.7333547957774545e-07,
      "loss": 1.8388,
      "step": 960
    },
    {
      "epoch": 1.8875912408759126,
      "grad_norm": 14.925251495002405,
      "learning_rate": 2.0009518022346075e-07,
      "loss": 1.8665,
      "step": 970
    },
    {
      "epoch": 1.9070559610705597,
      "grad_norm": 16.423759367972707,
      "learning_rate": 1.3815795018125133e-07,
      "loss": 1.8027,
      "step": 980
    },
    {
      "epoch": 1.926520681265207,
      "grad_norm": 7.929471875775863,
      "learning_rate": 8.759522696168865e-08,
      "loss": 1.7531,
      "step": 990
    },
    {
      "epoch": 1.945985401459854,
      "grad_norm": 6.194804501228712,
      "learning_rate": 4.846532888272304e-08,
      "loss": 1.8579,
      "step": 1000
    },
    {
      "epoch": 1.9654501216545013,
      "grad_norm": 14.714124220198908,
      "learning_rate": 2.081338780617337e-08,
      "loss": 1.8148,
      "step": 1010
    },
    {
      "epoch": 1.9849148418491485,
      "grad_norm": 15.437155062866172,
      "learning_rate": 4.671297083285176e-09,
      "loss": 1.8327,
      "step": 1020
    }
  ],
  "logging_steps": 10,
  "max_steps": 1028,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 225241411682304.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}