{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 8975,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011142061281337047,
      "grad_norm": 0.04801425710320473,
      "learning_rate": 2.7859033291544785e-07,
      "loss": 2.4677,
      "step": 100
    },
    {
      "epoch": 0.022284122562674095,
      "grad_norm": 0.04515479877591133,
      "learning_rate": 5.571806658308957e-07,
      "loss": 2.465,
      "step": 200
    },
    {
      "epoch": 0.033426183844011144,
      "grad_norm": 0.04905751347541809,
      "learning_rate": 8.357709987463436e-07,
      "loss": 2.479,
      "step": 300
    },
    {
      "epoch": 0.04456824512534819,
      "grad_norm": 0.0585104376077652,
      "learning_rate": 1.1143613316617914e-06,
      "loss": 2.4715,
      "step": 400
    },
    {
      "epoch": 0.055710306406685235,
      "grad_norm": 0.06588415056467056,
      "learning_rate": 1.3929516645772392e-06,
      "loss": 2.4912,
      "step": 500
    },
    {
      "epoch": 0.06685236768802229,
      "grad_norm": 0.07128334790468216,
      "learning_rate": 1.6715419974926873e-06,
      "loss": 2.449,
      "step": 600
    },
    {
      "epoch": 0.07799442896935933,
      "grad_norm": 0.08448737859725952,
      "learning_rate": 1.9501323304081347e-06,
      "loss": 2.4613,
      "step": 700
    },
    {
      "epoch": 0.08913649025069638,
      "grad_norm": 0.09213274717330933,
      "learning_rate": 2.2287226633235828e-06,
      "loss": 2.4366,
      "step": 800
    },
    {
      "epoch": 0.10027855153203342,
      "grad_norm": 0.10825659334659576,
      "learning_rate": 2.507312996239031e-06,
      "loss": 2.4342,
      "step": 900
    },
    {
      "epoch": 0.11142061281337047,
      "grad_norm": 0.11989778280258179,
      "learning_rate": 2.7859033291544785e-06,
      "loss": 2.4227,
      "step": 1000
    },
    {
      "epoch": 0.12256267409470752,
      "grad_norm": 0.11828930675983429,
      "learning_rate": 3.064493662069926e-06,
      "loss": 2.4346,
      "step": 1100
    },
    {
      "epoch": 0.13370473537604458,
      "grad_norm": 0.13713237643241882,
      "learning_rate": 3.3430839949853746e-06,
      "loss": 2.3954,
      "step": 1200
    },
    {
      "epoch": 0.14484679665738162,
      "grad_norm": 0.13783149421215057,
      "learning_rate": 3.6216743279008222e-06,
      "loss": 2.4027,
      "step": 1300
    },
    {
      "epoch": 0.15598885793871867,
      "grad_norm": 0.14380280673503876,
      "learning_rate": 3.9002646608162694e-06,
      "loss": 2.3894,
      "step": 1400
    },
    {
      "epoch": 0.1671309192200557,
      "grad_norm": 0.1586257964372635,
      "learning_rate": 4.178854993731718e-06,
      "loss": 2.3923,
      "step": 1500
    },
    {
      "epoch": 0.17827298050139276,
      "grad_norm": 0.15052290260791779,
      "learning_rate": 4.4574453266471655e-06,
      "loss": 2.3792,
      "step": 1600
    },
    {
      "epoch": 0.1894150417827298,
      "grad_norm": 0.16659782826900482,
      "learning_rate": 4.736035659562614e-06,
      "loss": 2.3657,
      "step": 1700
    },
    {
      "epoch": 0.20055710306406685,
      "grad_norm": 0.16844528913497925,
      "learning_rate": 5.014625992478062e-06,
      "loss": 2.3748,
      "step": 1800
    },
    {
      "epoch": 0.2116991643454039,
      "grad_norm": 0.17568887770175934,
      "learning_rate": 5.29321632539351e-06,
      "loss": 2.3609,
      "step": 1900
    },
    {
      "epoch": 0.22284122562674094,
      "grad_norm": 0.19198767840862274,
      "learning_rate": 5.571806658308957e-06,
      "loss": 2.3544,
      "step": 2000
    },
    {
      "epoch": 0.233983286908078,
      "grad_norm": 0.19439826905727386,
      "learning_rate": 5.850396991224405e-06,
      "loss": 2.3589,
      "step": 2100
    },
    {
      "epoch": 0.24512534818941503,
      "grad_norm": 0.2022334635257721,
      "learning_rate": 6.128987324139852e-06,
      "loss": 2.3612,
      "step": 2200
    },
    {
      "epoch": 0.2562674094707521,
      "grad_norm": 0.20067375898361206,
      "learning_rate": 6.4075776570553e-06,
      "loss": 2.3417,
      "step": 2300
    },
    {
      "epoch": 0.26740947075208915,
      "grad_norm": 0.2165047824382782,
      "learning_rate": 6.686167989970749e-06,
      "loss": 2.3481,
      "step": 2400
    },
    {
      "epoch": 0.2785515320334262,
      "grad_norm": 0.21726970374584198,
      "learning_rate": 6.9647583228861955e-06,
      "loss": 2.3331,
      "step": 2500
    },
    {
      "epoch": 0.28969359331476324,
      "grad_norm": 0.21808430552482605,
      "learning_rate": 7.2433486558016444e-06,
      "loss": 2.336,
      "step": 2600
    },
    {
      "epoch": 0.3008356545961003,
      "grad_norm": 0.23211392760276794,
      "learning_rate": 7.5219389887170925e-06,
      "loss": 2.3332,
      "step": 2700
    },
    {
      "epoch": 0.31197771587743733,
      "grad_norm": 0.24134476482868195,
      "learning_rate": 7.800529321632539e-06,
      "loss": 2.3377,
      "step": 2800
    },
    {
      "epoch": 0.3231197771587744,
      "grad_norm": 0.2478109449148178,
      "learning_rate": 8.079119654547987e-06,
      "loss": 2.3209,
      "step": 2900
    },
    {
      "epoch": 0.3342618384401114,
      "grad_norm": 0.25040534138679504,
      "learning_rate": 8.357709987463437e-06,
      "loss": 2.3218,
      "step": 3000
    },
    {
      "epoch": 0.34540389972144847,
      "grad_norm": 0.25530460476875305,
      "learning_rate": 8.636300320378883e-06,
      "loss": 2.3063,
      "step": 3100
    },
    {
      "epoch": 0.3565459610027855,
      "grad_norm": 0.258489727973938,
      "learning_rate": 8.914890653294331e-06,
      "loss": 2.3218,
      "step": 3200
    },
    {
      "epoch": 0.36768802228412256,
      "grad_norm": 0.28399255871772766,
      "learning_rate": 9.193480986209779e-06,
      "loss": 2.3063,
      "step": 3300
    },
    {
      "epoch": 0.3788300835654596,
      "grad_norm": 0.2863540053367615,
      "learning_rate": 9.472071319125227e-06,
      "loss": 2.2997,
      "step": 3400
    },
    {
      "epoch": 0.38997214484679665,
      "grad_norm": 0.2868782877922058,
      "learning_rate": 9.750661652040675e-06,
      "loss": 2.2947,
      "step": 3500
    },
    {
      "epoch": 0.4011142061281337,
      "grad_norm": 0.2859346270561218,
      "learning_rate": 1.0029251984956123e-05,
      "loss": 2.3054,
      "step": 3600
    },
    {
      "epoch": 0.41225626740947074,
      "grad_norm": 0.29307645559310913,
      "learning_rate": 1.030784231787157e-05,
      "loss": 2.2882,
      "step": 3700
    },
    {
      "epoch": 0.4233983286908078,
      "grad_norm": 0.3006058633327484,
      "learning_rate": 1.058643265078702e-05,
      "loss": 2.2893,
      "step": 3800
    },
    {
      "epoch": 0.43454038997214484,
      "grad_norm": 0.31586816906929016,
      "learning_rate": 1.0865022983702467e-05,
      "loss": 2.2967,
      "step": 3900
    },
    {
      "epoch": 0.4456824512534819,
      "grad_norm": 0.3375140428543091,
      "learning_rate": 1.1143613316617914e-05,
      "loss": 2.2694,
      "step": 4000
    },
    {
      "epoch": 0.4568245125348189,
      "grad_norm": 0.30628982186317444,
      "learning_rate": 1.142220364953336e-05,
      "loss": 2.2709,
      "step": 4100
    },
    {
      "epoch": 0.467966573816156,
      "grad_norm": 0.34296101331710815,
      "learning_rate": 1.170079398244881e-05,
      "loss": 2.2783,
      "step": 4200
    },
    {
      "epoch": 0.479108635097493,
      "grad_norm": 0.33544307947158813,
      "learning_rate": 1.1979384315364258e-05,
      "loss": 2.2825,
      "step": 4300
    },
    {
      "epoch": 0.49025069637883006,
      "grad_norm": 0.3508812487125397,
      "learning_rate": 1.2257974648279704e-05,
      "loss": 2.2793,
      "step": 4400
    },
    {
      "epoch": 0.5013927576601671,
      "grad_norm": 0.34851330518722534,
      "learning_rate": 1.2536564981195154e-05,
      "loss": 2.2703,
      "step": 4500
    },
    {
      "epoch": 0.5125348189415042,
      "grad_norm": 0.3589787185192108,
      "learning_rate": 1.28151553141106e-05,
      "loss": 2.2649,
      "step": 4600
    },
    {
      "epoch": 0.5236768802228412,
      "grad_norm": 0.3842375576496124,
      "learning_rate": 1.3093745647026049e-05,
      "loss": 2.264,
      "step": 4700
    },
    {
      "epoch": 0.5348189415041783,
      "grad_norm": 0.3963667154312134,
      "learning_rate": 1.3372335979941498e-05,
      "loss": 2.2567,
      "step": 4800
    },
    {
      "epoch": 0.5459610027855153,
      "grad_norm": 0.40061014890670776,
      "learning_rate": 1.3650926312856945e-05,
      "loss": 2.2567,
      "step": 4900
    },
    {
      "epoch": 0.5571030640668524,
      "grad_norm": 0.3819388747215271,
      "learning_rate": 1.3929516645772391e-05,
      "loss": 2.2545,
      "step": 5000
    },
    {
      "epoch": 0.5682451253481894,
      "grad_norm": 0.40177789330482483,
      "learning_rate": 1.420810697868784e-05,
      "loss": 2.2509,
      "step": 5100
    },
    {
      "epoch": 0.5793871866295265,
      "grad_norm": 0.4258297085762024,
      "learning_rate": 1.4486697311603289e-05,
      "loss": 2.2466,
      "step": 5200
    },
    {
      "epoch": 0.5905292479108635,
      "grad_norm": 0.44000041484832764,
      "learning_rate": 1.4765287644518735e-05,
      "loss": 2.2453,
      "step": 5300
    },
    {
      "epoch": 0.6016713091922006,
      "grad_norm": 0.4116784632205963,
      "learning_rate": 1.5043877977434185e-05,
      "loss": 2.2708,
      "step": 5400
    },
    {
      "epoch": 0.6128133704735376,
      "grad_norm": 0.39489638805389404,
      "learning_rate": 1.5322468310349633e-05,
      "loss": 2.2656,
      "step": 5500
    },
    {
      "epoch": 0.6239554317548747,
      "grad_norm": 0.43062567710876465,
      "learning_rate": 1.5601058643265078e-05,
      "loss": 2.2305,
      "step": 5600
    },
    {
      "epoch": 0.6350974930362117,
      "grad_norm": 0.4567050635814667,
      "learning_rate": 1.587964897618053e-05,
      "loss": 2.2392,
      "step": 5700
    },
    {
      "epoch": 0.6462395543175488,
      "grad_norm": 0.40272602438926697,
      "learning_rate": 1.6158239309095974e-05,
      "loss": 2.2327,
      "step": 5800
    },
    {
      "epoch": 0.6573816155988857,
      "grad_norm": 0.4182547628879547,
      "learning_rate": 1.6436829642011422e-05,
      "loss": 2.227,
      "step": 5900
    },
    {
      "epoch": 0.6685236768802229,
      "grad_norm": 0.4616607129573822,
      "learning_rate": 1.6715419974926873e-05,
      "loss": 2.223,
      "step": 6000
    },
    {
      "epoch": 0.6796657381615598,
      "grad_norm": 0.4605095386505127,
      "learning_rate": 1.6994010307842318e-05,
      "loss": 2.229,
      "step": 6100
    },
    {
      "epoch": 0.6908077994428969,
      "grad_norm": 0.48064395785331726,
      "learning_rate": 1.7272600640757766e-05,
      "loss": 2.2285,
      "step": 6200
    },
    {
      "epoch": 0.7019498607242339,
      "grad_norm": 0.45375484228134155,
      "learning_rate": 1.7551190973673214e-05,
      "loss": 2.231,
      "step": 6300
    },
    {
      "epoch": 0.713091922005571,
      "grad_norm": 0.4386723041534424,
      "learning_rate": 1.7829781306588662e-05,
      "loss": 2.2295,
      "step": 6400
    },
    {
      "epoch": 0.724233983286908,
      "grad_norm": 0.4519370198249817,
      "learning_rate": 1.810837163950411e-05,
      "loss": 2.2185,
      "step": 6500
    },
    {
      "epoch": 0.7353760445682451,
      "grad_norm": 0.45770174264907837,
      "learning_rate": 1.8386961972419558e-05,
      "loss": 2.2146,
      "step": 6600
    },
    {
      "epoch": 0.7465181058495822,
      "grad_norm": 0.46970096230506897,
      "learning_rate": 1.8665552305335006e-05,
      "loss": 2.2268,
      "step": 6700
    },
    {
      "epoch": 0.7576601671309192,
      "grad_norm": 0.4301040768623352,
      "learning_rate": 1.8944142638250454e-05,
      "loss": 2.2186,
      "step": 6800
    },
    {
      "epoch": 0.7688022284122563,
      "grad_norm": 0.49275335669517517,
      "learning_rate": 1.9222732971165902e-05,
      "loss": 2.2143,
      "step": 6900
    },
    {
      "epoch": 0.7799442896935933,
      "grad_norm": 0.4531406760215759,
      "learning_rate": 1.950132330408135e-05,
      "loss": 2.2087,
      "step": 7000
    },
    {
      "epoch": 0.7910863509749304,
      "grad_norm": 0.44058841466903687,
      "learning_rate": 1.97799136369968e-05,
      "loss": 2.2081,
      "step": 7100
    },
    {
      "epoch": 0.8022284122562674,
      "grad_norm": 0.4620118737220764,
      "learning_rate": 1.9993253992196043e-05,
      "loss": 2.2029,
      "step": 7200
    },
    {
      "epoch": 0.8133704735376045,
      "grad_norm": 0.4749448299407959,
      "learning_rate": 1.9776845355951603e-05,
      "loss": 2.1974,
      "step": 7300
    },
    {
      "epoch": 0.8245125348189415,
      "grad_norm": 0.4482609033584595,
      "learning_rate": 1.9262050899616325e-05,
      "loss": 2.2229,
      "step": 7400
    },
    {
      "epoch": 0.8356545961002786,
      "grad_norm": 0.4689159095287323,
      "learning_rate": 1.8464581965697866e-05,
      "loss": 2.2081,
      "step": 7500
    },
    {
      "epoch": 0.8467966573816156,
      "grad_norm": 0.4448583126068115,
      "learning_rate": 1.7408777020560473e-05,
      "loss": 2.2181,
      "step": 7600
    },
    {
      "epoch": 0.8579387186629527,
      "grad_norm": 0.513080894947052,
      "learning_rate": 1.6126858853144854e-05,
      "loss": 2.2045,
      "step": 7700
    },
    {
      "epoch": 0.8690807799442897,
      "grad_norm": 0.4850214421749115,
      "learning_rate": 1.465795114698568e-05,
      "loss": 2.1966,
      "step": 7800
    },
    {
      "epoch": 0.8802228412256268,
      "grad_norm": 0.5757237672805786,
      "learning_rate": 1.3046884439396632e-05,
      "loss": 2.2019,
      "step": 7900
    },
    {
      "epoch": 0.8913649025069638,
      "grad_norm": 0.48911547660827637,
      "learning_rate": 1.1342827909521198e-05,
      "loss": 2.1994,
      "step": 8000
    },
    {
      "epoch": 0.9025069637883009,
      "grad_norm": 0.5143416523933411,
      "learning_rate": 9.597788752588108e-06,
      "loss": 2.207,
      "step": 8100
    },
    {
      "epoch": 0.9136490250696379,
      "grad_norm": 0.483064204454422,
      "learning_rate": 7.882117054290375e-06,
      "loss": 2.1971,
      "step": 8200
    },
    {
      "epoch": 0.924791086350975,
      "grad_norm": 0.4610692262649536,
      "learning_rate": 6.213603769467132e-06,
      "loss": 2.1977,
      "step": 8300
    },
    {
      "epoch": 0.935933147632312,
      "grad_norm": 0.50405353307724,
      "learning_rate": 4.660649942045826e-06,
      "loss": 2.1955,
      "step": 8400
    },
    {
      "epoch": 0.947075208913649,
      "grad_norm": 0.47849225997924805,
      "learning_rate": 3.27065116687201e-06,
      "loss": 2.1906,
      "step": 8500
    },
    {
      "epoch": 0.958217270194986,
      "grad_norm": 0.49040549993515015,
      "learning_rate": 2.0860297089355943e-06,
      "loss": 2.1875,
      "step": 8600
    },
    {
      "epoch": 0.9693593314763231,
      "grad_norm": 0.49369877576828003,
      "learning_rate": 1.1429397910307794e-06,
      "loss": 2.2054,
      "step": 8700
    },
    {
      "epoch": 0.9805013927576601,
      "grad_norm": 0.5104972124099731,
      "learning_rate": 4.701641798198353e-07,
      "loss": 2.2121,
      "step": 8800
    },
    {
      "epoch": 0.9916434540389972,
      "grad_norm": 0.48128950595855713,
      "learning_rate": 8.823574609897134e-08,
      "loss": 2.1928,
      "step": 8900
    },
    {
      "epoch": 1.0,
      "step": 8975,
      "total_flos": 1.3078189256852111e+18,
      "train_loss": 2.2911263094506222,
      "train_runtime": 2843.733,
      "train_samples_per_second": 50.495,
      "train_steps_per_second": 3.156
    }
  ],
  "logging_steps": 100,
  "max_steps": 8975,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1.3078189256852111e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}