{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9924215084806929,
  "eval_steps": 500,
  "global_step": 49500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01002445968162316,
      "grad_norm": 0.1908789426088333,
      "learning_rate": 2.969926620955131e-05,
      "loss": 0.0305,
      "step": 500
    },
    {
      "epoch": 0.02004891936324632,
      "grad_norm": 0.2333366870880127,
      "learning_rate": 2.939853241910261e-05,
      "loss": 0.027,
      "step": 1000
    },
    {
      "epoch": 0.03007337904486948,
      "grad_norm": 0.18207737803459167,
      "learning_rate": 2.9097798628653917e-05,
      "loss": 0.026,
      "step": 1500
    },
    {
      "epoch": 0.04009783872649264,
      "grad_norm": 0.13153837621212006,
      "learning_rate": 2.879706483820522e-05,
      "loss": 0.0253,
      "step": 2000
    },
    {
      "epoch": 0.0501222984081158,
      "grad_norm": 0.20580655336380005,
      "learning_rate": 2.849633104775653e-05,
      "loss": 0.0249,
      "step": 2500
    },
    {
      "epoch": 0.06014675808973896,
      "grad_norm": 0.16466622054576874,
      "learning_rate": 2.8196198724888727e-05,
      "loss": 0.0239,
      "step": 3000
    },
    {
      "epoch": 0.07017121777136212,
      "grad_norm": 0.1590748131275177,
      "learning_rate": 2.7895464934440035e-05,
      "loss": 0.0239,
      "step": 3500
    },
    {
      "epoch": 0.08019567745298528,
      "grad_norm": 0.14150911569595337,
      "learning_rate": 2.759473114399134e-05,
      "loss": 0.0236,
      "step": 4000
    },
    {
      "epoch": 0.09022013713460844,
      "grad_norm": 0.09914161264896393,
      "learning_rate": 2.7293997353542647e-05,
      "loss": 0.0236,
      "step": 4500
    },
    {
      "epoch": 0.1002445968162316,
      "grad_norm": 0.16175945103168488,
      "learning_rate": 2.6993263563093947e-05,
      "loss": 0.0232,
      "step": 5000
    },
    {
      "epoch": 0.11026905649785476,
      "grad_norm": 0.14557033777236938,
      "learning_rate": 2.6692529772645255e-05,
      "loss": 0.0229,
      "step": 5500
    },
    {
      "epoch": 0.12029351617947792,
      "grad_norm": 0.14236459136009216,
      "learning_rate": 2.639179598219656e-05,
      "loss": 0.0228,
      "step": 6000
    },
    {
      "epoch": 0.13031797586110108,
      "grad_norm": 0.1859973669052124,
      "learning_rate": 2.6091062191747867e-05,
      "loss": 0.0226,
      "step": 6500
    },
    {
      "epoch": 0.14034243554272424,
      "grad_norm": 0.21533679962158203,
      "learning_rate": 2.5790929868880065e-05,
      "loss": 0.0226,
      "step": 7000
    },
    {
      "epoch": 0.1503668952243474,
      "grad_norm": 0.13067375123500824,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.0222,
      "step": 7500
    },
    {
      "epoch": 0.16039135490597056,
      "grad_norm": 0.15966835618019104,
      "learning_rate": 2.518946228798268e-05,
      "loss": 0.0223,
      "step": 8000
    },
    {
      "epoch": 0.17041581458759372,
      "grad_norm": 0.16358782351016998,
      "learning_rate": 2.4888728497533985e-05,
      "loss": 0.0218,
      "step": 8500
    },
    {
      "epoch": 0.18044027426921688,
      "grad_norm": 0.2502037286758423,
      "learning_rate": 2.458799470708529e-05,
      "loss": 0.0219,
      "step": 9000
    },
    {
      "epoch": 0.19046473395084004,
      "grad_norm": 0.13426147401332855,
      "learning_rate": 2.4287260916636593e-05,
      "loss": 0.0221,
      "step": 9500
    },
    {
      "epoch": 0.2004891936324632,
      "grad_norm": 0.17509840428829193,
      "learning_rate": 2.39865271261879e-05,
      "loss": 0.0219,
      "step": 10000
    },
    {
      "epoch": 0.21051365331408636,
      "grad_norm": 0.12009692937135696,
      "learning_rate": 2.3685793335739205e-05,
      "loss": 0.0216,
      "step": 10500
    },
    {
      "epoch": 0.22053811299570952,
      "grad_norm": 0.1322207748889923,
      "learning_rate": 2.338505954529051e-05,
      "loss": 0.0211,
      "step": 11000
    },
    {
      "epoch": 0.23056257267733268,
      "grad_norm": 0.14883308112621307,
      "learning_rate": 2.308492722242271e-05,
      "loss": 0.0214,
      "step": 11500
    },
    {
      "epoch": 0.24058703235895584,
      "grad_norm": 0.13109040260314941,
      "learning_rate": 2.278419343197402e-05,
      "loss": 0.0211,
      "step": 12000
    },
    {
      "epoch": 0.25061149204057903,
      "grad_norm": 0.1747232973575592,
      "learning_rate": 2.2483459641525323e-05,
      "loss": 0.0214,
      "step": 12500
    },
    {
      "epoch": 0.26063595172220216,
      "grad_norm": 0.14633044600486755,
      "learning_rate": 2.2182725851076627e-05,
      "loss": 0.0212,
      "step": 13000
    },
    {
      "epoch": 0.27066041140382535,
      "grad_norm": 0.16928279399871826,
      "learning_rate": 2.188199206062793e-05,
      "loss": 0.0209,
      "step": 13500
    },
    {
      "epoch": 0.2806848710854485,
      "grad_norm": 0.1422567069530487,
      "learning_rate": 2.158125827017924e-05,
      "loss": 0.0206,
      "step": 14000
    },
    {
      "epoch": 0.29070933076707167,
      "grad_norm": 0.1617174744606018,
      "learning_rate": 2.1280524479730543e-05,
      "loss": 0.0208,
      "step": 14500
    },
    {
      "epoch": 0.3007337904486948,
      "grad_norm": 0.12710513174533844,
      "learning_rate": 2.0979790689281848e-05,
      "loss": 0.0204,
      "step": 15000
    },
    {
      "epoch": 0.310758250130318,
      "grad_norm": 0.12561199069023132,
      "learning_rate": 2.0679056898833152e-05,
      "loss": 0.0207,
      "step": 15500
    },
    {
      "epoch": 0.3207827098119411,
      "grad_norm": 0.12565076351165771,
      "learning_rate": 2.037832310838446e-05,
      "loss": 0.0204,
      "step": 16000
    },
    {
      "epoch": 0.3308071694935643,
      "grad_norm": 0.16111774742603302,
      "learning_rate": 2.0077589317935764e-05,
      "loss": 0.0204,
      "step": 16500
    },
    {
      "epoch": 0.34083162917518744,
      "grad_norm": 0.08807434886693954,
      "learning_rate": 1.9776855527487068e-05,
      "loss": 0.0204,
      "step": 17000
    },
    {
      "epoch": 0.35085608885681063,
      "grad_norm": 0.12991702556610107,
      "learning_rate": 1.9476121737038372e-05,
      "loss": 0.0203,
      "step": 17500
    },
    {
      "epoch": 0.36088054853843377,
      "grad_norm": 0.17382492125034332,
      "learning_rate": 1.917538794658968e-05,
      "loss": 0.0199,
      "step": 18000
    },
    {
      "epoch": 0.37090500822005695,
      "grad_norm": 0.16158685088157654,
      "learning_rate": 1.8875255623721885e-05,
      "loss": 0.02,
      "step": 18500
    },
    {
      "epoch": 0.3809294679016801,
      "grad_norm": 0.1214708685874939,
      "learning_rate": 1.8575123300854082e-05,
      "loss": 0.0199,
      "step": 19000
    },
    {
      "epoch": 0.3909539275833033,
      "grad_norm": 0.11400719732046127,
      "learning_rate": 1.827438951040539e-05,
      "loss": 0.0197,
      "step": 19500
    },
    {
      "epoch": 0.4009783872649264,
      "grad_norm": 0.09430436044931412,
      "learning_rate": 1.7973655719956694e-05,
      "loss": 0.0197,
      "step": 20000
    },
    {
      "epoch": 0.4110028469465496,
      "grad_norm": 0.13852345943450928,
      "learning_rate": 1.7672921929508002e-05,
      "loss": 0.0199,
      "step": 20500
    },
    {
      "epoch": 0.4210273066281727,
      "grad_norm": 0.13223719596862793,
      "learning_rate": 1.7372188139059303e-05,
      "loss": 0.0198,
      "step": 21000
    },
    {
      "epoch": 0.4310517663097959,
      "grad_norm": 0.08761810511350632,
      "learning_rate": 1.707145434861061e-05,
      "loss": 0.0196,
      "step": 21500
    },
    {
      "epoch": 0.44107622599141905,
      "grad_norm": 0.13589079678058624,
      "learning_rate": 1.6770720558161915e-05,
      "loss": 0.02,
      "step": 22000
    },
    {
      "epoch": 0.45110068567304223,
      "grad_norm": 0.08927265554666519,
      "learning_rate": 1.6469986767713223e-05,
      "loss": 0.0196,
      "step": 22500
    },
    {
      "epoch": 0.46112514535466537,
      "grad_norm": 0.06504929065704346,
      "learning_rate": 1.6169252977264524e-05,
      "loss": 0.0196,
      "step": 23000
    },
    {
      "epoch": 0.47114960503628855,
      "grad_norm": 0.17200474441051483,
      "learning_rate": 1.586851918681583e-05,
      "loss": 0.0192,
      "step": 23500
    },
    {
      "epoch": 0.4811740647179117,
      "grad_norm": 0.15591026842594147,
      "learning_rate": 1.5567785396367136e-05,
      "loss": 0.0195,
      "step": 24000
    },
    {
      "epoch": 0.4911985243995349,
      "grad_norm": 0.07878988236188889,
      "learning_rate": 1.5267051605918443e-05,
      "loss": 0.0197,
      "step": 24500
    },
    {
      "epoch": 0.5012229840811581,
      "grad_norm": 0.14917835593223572,
      "learning_rate": 1.4967520750631541e-05,
      "loss": 0.0194,
      "step": 25000
    },
    {
      "epoch": 0.5112474437627812,
      "grad_norm": 0.13516920804977417,
      "learning_rate": 1.4666786960182846e-05,
      "loss": 0.0191,
      "step": 25500
    },
    {
      "epoch": 0.5212719034444043,
      "grad_norm": 0.13264451920986176,
      "learning_rate": 1.4366053169734152e-05,
      "loss": 0.019,
      "step": 26000
    },
    {
      "epoch": 0.5312963631260275,
      "grad_norm": 0.11801792681217194,
      "learning_rate": 1.4065319379285456e-05,
      "loss": 0.019,
      "step": 26500
    },
    {
      "epoch": 0.5413208228076507,
      "grad_norm": 0.10833830386400223,
      "learning_rate": 1.3764585588836762e-05,
      "loss": 0.0192,
      "step": 27000
    },
    {
      "epoch": 0.5513452824892738,
      "grad_norm": 0.1249992698431015,
      "learning_rate": 1.3463851798388068e-05,
      "loss": 0.0193,
      "step": 27500
    },
    {
      "epoch": 0.561369742170897,
      "grad_norm": 0.1460021287202835,
      "learning_rate": 1.3163118007939374e-05,
      "loss": 0.0189,
      "step": 28000
    },
    {
      "epoch": 0.5713942018525201,
      "grad_norm": 0.10502651333808899,
      "learning_rate": 1.2862384217490678e-05,
      "loss": 0.019,
      "step": 28500
    },
    {
      "epoch": 0.5814186615341433,
      "grad_norm": 0.1526552140712738,
      "learning_rate": 1.2561650427041984e-05,
      "loss": 0.0189,
      "step": 29000
    },
    {
      "epoch": 0.5914431212157665,
      "grad_norm": 0.10992331802845001,
      "learning_rate": 1.2261518104174185e-05,
      "loss": 0.0189,
      "step": 29500
    },
    {
      "epoch": 0.6014675808973896,
      "grad_norm": 0.13150319457054138,
      "learning_rate": 1.1961385781306388e-05,
      "loss": 0.0189,
      "step": 30000
    },
    {
      "epoch": 0.6114920405790127,
      "grad_norm": 0.13109181821346283,
      "learning_rate": 1.1661253458438591e-05,
      "loss": 0.0187,
      "step": 30500
    },
    {
      "epoch": 0.621516500260636,
      "grad_norm": 0.1083877831697464,
      "learning_rate": 1.1360519667989895e-05,
      "loss": 0.0187,
      "step": 31000
    },
    {
      "epoch": 0.6315409599422591,
      "grad_norm": 0.1419740468263626,
      "learning_rate": 1.1059785877541201e-05,
      "loss": 0.0186,
      "step": 31500
    },
    {
      "epoch": 0.6415654196238822,
      "grad_norm": 0.13743899762630463,
      "learning_rate": 1.0759052087092506e-05,
      "loss": 0.0187,
      "step": 32000
    },
    {
      "epoch": 0.6515898793055054,
      "grad_norm": 0.11756709218025208,
      "learning_rate": 1.0458318296643812e-05,
      "loss": 0.0187,
      "step": 32500
    },
    {
      "epoch": 0.6616143389871286,
      "grad_norm": 0.28276416659355164,
      "learning_rate": 1.0157584506195116e-05,
      "loss": 0.0186,
      "step": 33000
    },
    {
      "epoch": 0.6716387986687518,
      "grad_norm": 0.08689709007740021,
      "learning_rate": 9.856850715746422e-06,
      "loss": 0.0185,
      "step": 33500
    },
    {
      "epoch": 0.6816632583503749,
      "grad_norm": 0.16307085752487183,
      "learning_rate": 9.556116925297726e-06,
      "loss": 0.0188,
      "step": 34000
    },
    {
      "epoch": 0.691687718031998,
      "grad_norm": 0.09724652022123337,
      "learning_rate": 9.25598460242993e-06,
      "loss": 0.0186,
      "step": 34500
    },
    {
      "epoch": 0.7017121777136213,
      "grad_norm": 0.1357831209897995,
      "learning_rate": 8.955250811981233e-06,
      "loss": 0.0185,
      "step": 35000
    },
    {
      "epoch": 0.7117366373952444,
      "grad_norm": 0.07240907102823257,
      "learning_rate": 8.65451702153254e-06,
      "loss": 0.0182,
      "step": 35500
    },
    {
      "epoch": 0.7217610970768675,
      "grad_norm": 0.12668687105178833,
      "learning_rate": 8.353783231083844e-06,
      "loss": 0.0184,
      "step": 36000
    },
    {
      "epoch": 0.7317855567584907,
      "grad_norm": 0.10139577835798264,
      "learning_rate": 8.05304944063515e-06,
      "loss": 0.0183,
      "step": 36500
    },
    {
      "epoch": 0.7418100164401139,
      "grad_norm": 0.13369520008563995,
      "learning_rate": 7.752315650186454e-06,
      "loss": 0.0184,
      "step": 37000
    },
    {
      "epoch": 0.751834476121737,
      "grad_norm": 0.10762561112642288,
      "learning_rate": 7.451581859737761e-06,
      "loss": 0.0183,
      "step": 37500
    },
    {
      "epoch": 0.7618589358033602,
      "grad_norm": 0.11992637068033218,
      "learning_rate": 7.150848069289066e-06,
      "loss": 0.0182,
      "step": 38000
    },
    {
      "epoch": 0.7718833954849834,
      "grad_norm": 0.10149220377206802,
      "learning_rate": 6.850114278840371e-06,
      "loss": 0.0184,
      "step": 38500
    },
    {
      "epoch": 0.7819078551666065,
      "grad_norm": 0.11299099028110504,
      "learning_rate": 6.549981955972573e-06,
      "loss": 0.0183,
      "step": 39000
    },
    {
      "epoch": 0.7919323148482297,
      "grad_norm": 0.1773054599761963,
      "learning_rate": 6.249248165523878e-06,
      "loss": 0.0184,
      "step": 39500
    },
    {
      "epoch": 0.8019567745298528,
      "grad_norm": 0.09563998878002167,
      "learning_rate": 5.949115842656081e-06,
      "loss": 0.0184,
      "step": 40000
    },
    {
      "epoch": 0.811981234211476,
      "grad_norm": 0.12298491597175598,
      "learning_rate": 5.648382052207386e-06,
      "loss": 0.0182,
      "step": 40500
    },
    {
      "epoch": 0.8220056938930992,
      "grad_norm": 0.1283864825963974,
      "learning_rate": 5.3476482617586915e-06,
      "loss": 0.0182,
      "step": 41000
    },
    {
      "epoch": 0.8320301535747223,
      "grad_norm": 0.10248496383428574,
      "learning_rate": 5.046914471309997e-06,
      "loss": 0.0182,
      "step": 41500
    },
    {
      "epoch": 0.8420546132563455,
      "grad_norm": 0.09150300920009613,
      "learning_rate": 4.746180680861302e-06,
      "loss": 0.0183,
      "step": 42000
    },
    {
      "epoch": 0.8520790729379687,
      "grad_norm": 0.1290718913078308,
      "learning_rate": 4.445446890412607e-06,
      "loss": 0.0183,
      "step": 42500
    },
    {
      "epoch": 0.8621035326195918,
      "grad_norm": 0.12163352966308594,
      "learning_rate": 4.144713099963912e-06,
      "loss": 0.0179,
      "step": 43000
    },
    {
      "epoch": 0.872127992301215,
      "grad_norm": 0.15341030061244965,
      "learning_rate": 3.843979309515217e-06,
      "loss": 0.0178,
      "step": 43500
    },
    {
      "epoch": 0.8821524519828381,
      "grad_norm": 0.1429668664932251,
      "learning_rate": 3.5438469866474197e-06,
      "loss": 0.0178,
      "step": 44000
    },
    {
      "epoch": 0.8921769116644613,
      "grad_norm": 0.10146286338567734,
      "learning_rate": 3.243113196198725e-06,
      "loss": 0.018,
      "step": 44500
    },
    {
      "epoch": 0.9022013713460845,
      "grad_norm": 0.1440785676240921,
      "learning_rate": 2.9429808733309277e-06,
      "loss": 0.0181,
      "step": 45000
    },
    {
      "epoch": 0.9122258310277076,
      "grad_norm": 0.10165946930646896,
      "learning_rate": 2.642247082882233e-06,
      "loss": 0.0178,
      "step": 45500
    },
    {
      "epoch": 0.9222502907093307,
      "grad_norm": 0.14765135943889618,
      "learning_rate": 2.341513292433538e-06,
      "loss": 0.018,
      "step": 46000
    },
    {
      "epoch": 0.932274750390954,
      "grad_norm": 0.07201279699802399,
      "learning_rate": 2.040779501984843e-06,
      "loss": 0.018,
      "step": 46500
    },
    {
      "epoch": 0.9422992100725771,
      "grad_norm": 0.13335958123207092,
      "learning_rate": 1.7400457115361483e-06,
      "loss": 0.0178,
      "step": 47000
    },
    {
      "epoch": 0.9523236697542002,
      "grad_norm": 0.10147497057914734,
      "learning_rate": 1.4393119210874535e-06,
      "loss": 0.0178,
      "step": 47500
    },
    {
      "epoch": 0.9623481294358234,
      "grad_norm": 0.09820819646120071,
      "learning_rate": 1.1385781306387586e-06,
      "loss": 0.0179,
      "step": 48000
    },
    {
      "epoch": 0.9723725891174466,
      "grad_norm": 0.1263681799173355,
      "learning_rate": 8.378443401900638e-07,
      "loss": 0.018,
      "step": 48500
    },
    {
      "epoch": 0.9823970487990697,
      "grad_norm": 0.08029880374670029,
      "learning_rate": 5.377120173222663e-07,
      "loss": 0.0181,
      "step": 49000
    },
    {
      "epoch": 0.9924215084806929,
      "grad_norm": 0.12001931667327881,
      "learning_rate": 2.369782268735715e-07,
      "loss": 0.018,
      "step": 49500
    }
  ],
  "logging_steps": 500,
  "max_steps": 49878,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}