{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9760,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010245901639344262,
      "grad_norm": 0.055161599069833755,
      "learning_rate": 2.5618035096708087e-07,
      "loss": 2.2315,
      "step": 100
    },
    {
      "epoch": 0.020491803278688523,
      "grad_norm": 0.05540880188345909,
      "learning_rate": 5.123607019341617e-07,
      "loss": 2.2296,
      "step": 200
    },
    {
      "epoch": 0.030737704918032786,
      "grad_norm": 0.0638064593076706,
      "learning_rate": 7.685410529012426e-07,
      "loss": 2.237,
      "step": 300
    },
    {
      "epoch": 0.040983606557377046,
      "grad_norm": 0.0683213621377945,
      "learning_rate": 1.0247214038683235e-06,
      "loss": 2.2204,
      "step": 400
    },
    {
      "epoch": 0.05122950819672131,
      "grad_norm": 0.08443646132946014,
      "learning_rate": 1.2809017548354041e-06,
      "loss": 2.2304,
      "step": 500
    },
    {
      "epoch": 0.06147540983606557,
      "grad_norm": 0.07605498284101486,
      "learning_rate": 1.5370821058024852e-06,
      "loss": 2.2239,
      "step": 600
    },
    {
      "epoch": 0.07172131147540983,
      "grad_norm": 0.08266434073448181,
      "learning_rate": 1.7932624567695659e-06,
      "loss": 2.2219,
      "step": 700
    },
    {
      "epoch": 0.08196721311475409,
      "grad_norm": 0.10109508037567139,
      "learning_rate": 2.049442807736647e-06,
      "loss": 2.2004,
      "step": 800
    },
    {
      "epoch": 0.09221311475409837,
      "grad_norm": 0.11055849492549896,
      "learning_rate": 2.3056231587037276e-06,
      "loss": 2.1945,
      "step": 900
    },
    {
      "epoch": 0.10245901639344263,
      "grad_norm": 0.1108393445611,
      "learning_rate": 2.5618035096708083e-06,
      "loss": 2.1769,
      "step": 1000
    },
    {
      "epoch": 0.11270491803278689,
      "grad_norm": 0.14474108815193176,
      "learning_rate": 2.8179838606378894e-06,
      "loss": 2.1904,
      "step": 1100
    },
    {
      "epoch": 0.12295081967213115,
      "grad_norm": 0.13787466287612915,
      "learning_rate": 3.0741642116049704e-06,
      "loss": 2.1765,
      "step": 1200
    },
    {
      "epoch": 0.13319672131147542,
      "grad_norm": 0.14868797361850739,
      "learning_rate": 3.330344562572051e-06,
      "loss": 2.1609,
      "step": 1300
    },
    {
      "epoch": 0.14344262295081966,
      "grad_norm": 0.15633544325828552,
      "learning_rate": 3.5865249135391318e-06,
      "loss": 2.1708,
      "step": 1400
    },
    {
      "epoch": 0.15368852459016394,
      "grad_norm": 0.17003007233142853,
      "learning_rate": 3.842705264506213e-06,
      "loss": 2.1521,
      "step": 1500
    },
    {
      "epoch": 0.16393442622950818,
      "grad_norm": 0.1722685843706131,
      "learning_rate": 4.098885615473294e-06,
      "loss": 2.1417,
      "step": 1600
    },
    {
      "epoch": 0.17418032786885246,
      "grad_norm": 0.18862290680408478,
      "learning_rate": 4.3550659664403746e-06,
      "loss": 2.1328,
      "step": 1700
    },
    {
      "epoch": 0.18442622950819673,
      "grad_norm": 0.211067795753479,
      "learning_rate": 4.611246317407455e-06,
      "loss": 2.1222,
      "step": 1800
    },
    {
      "epoch": 0.19467213114754098,
      "grad_norm": 0.21424412727355957,
      "learning_rate": 4.867426668374536e-06,
      "loss": 2.1448,
      "step": 1900
    },
    {
      "epoch": 0.20491803278688525,
      "grad_norm": 0.21230654418468475,
      "learning_rate": 5.1236070193416165e-06,
      "loss": 2.1189,
      "step": 2000
    },
    {
      "epoch": 0.2151639344262295,
      "grad_norm": 0.2188699096441269,
      "learning_rate": 5.379787370308698e-06,
      "loss": 2.1208,
      "step": 2100
    },
    {
      "epoch": 0.22540983606557377,
      "grad_norm": 0.23276151716709137,
      "learning_rate": 5.635967721275779e-06,
      "loss": 2.0994,
      "step": 2200
    },
    {
      "epoch": 0.23565573770491804,
      "grad_norm": 0.25028660893440247,
      "learning_rate": 5.892148072242859e-06,
      "loss": 2.0921,
      "step": 2300
    },
    {
      "epoch": 0.2459016393442623,
      "grad_norm": 0.23129107058048248,
      "learning_rate": 6.148328423209941e-06,
      "loss": 2.0986,
      "step": 2400
    },
    {
      "epoch": 0.25614754098360654,
      "grad_norm": 0.30324786901474,
      "learning_rate": 6.404508774177021e-06,
      "loss": 2.0998,
      "step": 2500
    },
    {
      "epoch": 0.26639344262295084,
      "grad_norm": 0.3137739896774292,
      "learning_rate": 6.660689125144102e-06,
      "loss": 2.0831,
      "step": 2600
    },
    {
      "epoch": 0.2766393442622951,
      "grad_norm": 0.29233041405677795,
      "learning_rate": 6.916869476111183e-06,
      "loss": 2.072,
      "step": 2700
    },
    {
      "epoch": 0.28688524590163933,
      "grad_norm": 0.2947608232498169,
      "learning_rate": 7.1730498270782635e-06,
      "loss": 2.0713,
      "step": 2800
    },
    {
      "epoch": 0.29713114754098363,
      "grad_norm": 0.2987135350704193,
      "learning_rate": 7.429230178045345e-06,
      "loss": 2.0745,
      "step": 2900
    },
    {
      "epoch": 0.3073770491803279,
      "grad_norm": 0.3173224627971649,
      "learning_rate": 7.685410529012427e-06,
      "loss": 2.0552,
      "step": 3000
    },
    {
      "epoch": 0.3176229508196721,
      "grad_norm": 0.3080308139324188,
      "learning_rate": 7.941590879979506e-06,
      "loss": 2.058,
      "step": 3100
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 0.3344276249408722,
      "learning_rate": 8.197771230946588e-06,
      "loss": 2.0335,
      "step": 3200
    },
    {
      "epoch": 0.33811475409836067,
      "grad_norm": 0.35821136832237244,
      "learning_rate": 8.453951581913668e-06,
      "loss": 2.0389,
      "step": 3300
    },
    {
      "epoch": 0.3483606557377049,
      "grad_norm": 0.34500202536582947,
      "learning_rate": 8.710131932880749e-06,
      "loss": 2.0473,
      "step": 3400
    },
    {
      "epoch": 0.35860655737704916,
      "grad_norm": 0.3924405872821808,
      "learning_rate": 8.96631228384783e-06,
      "loss": 2.0413,
      "step": 3500
    },
    {
      "epoch": 0.36885245901639346,
      "grad_norm": 0.3596484065055847,
      "learning_rate": 9.22249263481491e-06,
      "loss": 2.0408,
      "step": 3600
    },
    {
      "epoch": 0.3790983606557377,
      "grad_norm": 0.3991853892803192,
      "learning_rate": 9.478672985781992e-06,
      "loss": 2.0143,
      "step": 3700
    },
    {
      "epoch": 0.38934426229508196,
      "grad_norm": 0.36749976873397827,
      "learning_rate": 9.734853336749072e-06,
      "loss": 2.0248,
      "step": 3800
    },
    {
      "epoch": 0.39959016393442626,
      "grad_norm": 0.41294941306114197,
      "learning_rate": 9.991033687716153e-06,
      "loss": 2.0214,
      "step": 3900
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 0.40261194109916687,
      "learning_rate": 1.0247214038683233e-05,
      "loss": 2.0089,
      "step": 4000
    },
    {
      "epoch": 0.42008196721311475,
      "grad_norm": 0.3836148977279663,
      "learning_rate": 1.0503394389650315e-05,
      "loss": 2.0109,
      "step": 4100
    },
    {
      "epoch": 0.430327868852459,
      "grad_norm": 0.38613057136535645,
      "learning_rate": 1.0759574740617396e-05,
      "loss": 1.9881,
      "step": 4200
    },
    {
      "epoch": 0.4405737704918033,
      "grad_norm": 0.39355090260505676,
      "learning_rate": 1.1015755091584478e-05,
      "loss": 1.9919,
      "step": 4300
    },
    {
      "epoch": 0.45081967213114754,
      "grad_norm": 0.5257683992385864,
      "learning_rate": 1.1271935442551557e-05,
      "loss": 1.9929,
      "step": 4400
    },
    {
      "epoch": 0.4610655737704918,
      "grad_norm": 0.43797537684440613,
      "learning_rate": 1.1528115793518637e-05,
      "loss": 2.0092,
      "step": 4500
    },
    {
      "epoch": 0.4713114754098361,
      "grad_norm": 0.5497995018959045,
      "learning_rate": 1.1784296144485719e-05,
      "loss": 1.9967,
      "step": 4600
    },
    {
      "epoch": 0.48155737704918034,
      "grad_norm": 0.4768752157688141,
      "learning_rate": 1.20404764954528e-05,
      "loss": 1.9921,
      "step": 4700
    },
    {
      "epoch": 0.4918032786885246,
      "grad_norm": 0.43218129873275757,
      "learning_rate": 1.2296656846419882e-05,
      "loss": 1.9883,
      "step": 4800
    },
    {
      "epoch": 0.5020491803278688,
      "grad_norm": 0.47497090697288513,
      "learning_rate": 1.2552837197386962e-05,
      "loss": 1.9799,
      "step": 4900
    },
    {
      "epoch": 0.5122950819672131,
      "grad_norm": 0.4906615912914276,
      "learning_rate": 1.2809017548354041e-05,
      "loss": 1.9762,
      "step": 5000
    },
    {
      "epoch": 0.5225409836065574,
      "grad_norm": 0.5236986875534058,
      "learning_rate": 1.3065197899321123e-05,
      "loss": 1.9736,
      "step": 5100
    },
    {
      "epoch": 0.5327868852459017,
      "grad_norm": 0.5295742750167847,
      "learning_rate": 1.3318816446778533e-05,
      "loss": 1.9762,
      "step": 5200
    },
    {
      "epoch": 0.5430327868852459,
      "grad_norm": 0.4924796521663666,
      "learning_rate": 1.3574996797745615e-05,
      "loss": 1.9709,
      "step": 5300
    },
    {
      "epoch": 0.5532786885245902,
      "grad_norm": 0.5359669327735901,
      "learning_rate": 1.3831177148712696e-05,
      "loss": 1.9594,
      "step": 5400
    },
    {
      "epoch": 0.5635245901639344,
      "grad_norm": 0.49581173062324524,
      "learning_rate": 1.4087357499679775e-05,
      "loss": 1.9787,
      "step": 5500
    },
    {
      "epoch": 0.5737704918032787,
      "grad_norm": 0.49295076727867126,
      "learning_rate": 1.4343537850646856e-05,
      "loss": 1.9628,
      "step": 5600
    },
    {
      "epoch": 0.5840163934426229,
      "grad_norm": 0.6119683980941772,
      "learning_rate": 1.4599718201613938e-05,
      "loss": 1.9622,
      "step": 5700
    },
    {
      "epoch": 0.5942622950819673,
      "grad_norm": 0.5446974039077759,
      "learning_rate": 1.4855898552581019e-05,
      "loss": 1.9547,
      "step": 5800
    },
    {
      "epoch": 0.6045081967213115,
      "grad_norm": 0.5105884671211243,
      "learning_rate": 1.5112078903548099e-05,
      "loss": 1.9477,
      "step": 5900
    },
    {
      "epoch": 0.6147540983606558,
      "grad_norm": 0.5642023682594299,
      "learning_rate": 1.536825925451518e-05,
      "loss": 1.9433,
      "step": 6000
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.4802306294441223,
      "learning_rate": 1.562443960548226e-05,
      "loss": 1.984,
      "step": 6100
    },
    {
      "epoch": 0.6352459016393442,
      "grad_norm": 0.5581754446029663,
      "learning_rate": 1.588061995644934e-05,
      "loss": 1.9341,
      "step": 6200
    },
    {
      "epoch": 0.6454918032786885,
      "grad_norm": 0.5883158445358276,
      "learning_rate": 1.6136800307416423e-05,
      "loss": 1.9394,
      "step": 6300
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 0.5291092395782471,
      "learning_rate": 1.6392980658383503e-05,
      "loss": 1.9421,
      "step": 6400
    },
    {
      "epoch": 0.6659836065573771,
      "grad_norm": 0.6235113739967346,
      "learning_rate": 1.6649161009350586e-05,
      "loss": 1.9602,
      "step": 6500
    },
    {
      "epoch": 0.6762295081967213,
      "grad_norm": 0.49288830161094666,
      "learning_rate": 1.6905341360317663e-05,
      "loss": 1.9346,
      "step": 6600
    },
    {
      "epoch": 0.6864754098360656,
      "grad_norm": 0.6099859476089478,
      "learning_rate": 1.7161521711284746e-05,
      "loss": 1.9431,
      "step": 6700
    },
    {
      "epoch": 0.6967213114754098,
      "grad_norm": 0.5682905316352844,
      "learning_rate": 1.7417702062251826e-05,
      "loss": 1.9137,
      "step": 6800
    },
    {
      "epoch": 0.7069672131147541,
      "grad_norm": 0.7166417241096497,
      "learning_rate": 1.767388241321891e-05,
      "loss": 1.9391,
      "step": 6900
    },
    {
      "epoch": 0.7172131147540983,
      "grad_norm": 0.5477021336555481,
      "learning_rate": 1.793006276418599e-05,
      "loss": 1.9429,
      "step": 7000
    },
    {
      "epoch": 0.7274590163934426,
      "grad_norm": 0.54074627161026,
      "learning_rate": 1.818624311515307e-05,
      "loss": 1.9384,
      "step": 7100
    },
    {
      "epoch": 0.7377049180327869,
      "grad_norm": 0.5290340185165405,
      "learning_rate": 1.8442423466120148e-05,
      "loss": 1.9356,
      "step": 7200
    },
    {
      "epoch": 0.7479508196721312,
      "grad_norm": 0.5777207612991333,
      "learning_rate": 1.869860381708723e-05,
      "loss": 1.9184,
      "step": 7300
    },
    {
      "epoch": 0.7581967213114754,
      "grad_norm": 0.6420373320579529,
      "learning_rate": 1.895478416805431e-05,
      "loss": 1.9236,
      "step": 7400
    },
    {
      "epoch": 0.7684426229508197,
      "grad_norm": 0.5742615461349487,
      "learning_rate": 1.9210964519021394e-05,
      "loss": 1.9222,
      "step": 7500
    },
    {
      "epoch": 0.7786885245901639,
      "grad_norm": 0.613657534122467,
      "learning_rate": 1.946714486998847e-05,
      "loss": 1.9362,
      "step": 7600
    },
    {
      "epoch": 0.7889344262295082,
      "grad_norm": 0.5610764622688293,
      "learning_rate": 1.9723325220955554e-05,
      "loss": 1.9035,
      "step": 7700
    },
    {
      "epoch": 0.7991803278688525,
      "grad_norm": 0.6575483083724976,
      "learning_rate": 1.9979505571922634e-05,
      "loss": 1.9312,
      "step": 7800
    },
    {
      "epoch": 0.8094262295081968,
      "grad_norm": 0.5548086166381836,
      "learning_rate": 1.9890692959274893e-05,
      "loss": 1.9221,
      "step": 7900
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 0.6601373553276062,
      "learning_rate": 1.952683486776797e-05,
      "loss": 1.9076,
      "step": 8000
    },
    {
      "epoch": 0.8299180327868853,
      "grad_norm": 0.5061026215553284,
      "learning_rate": 1.891699257502426e-05,
      "loss": 1.9173,
      "step": 8100
    },
    {
      "epoch": 0.8401639344262295,
      "grad_norm": 0.5650314688682556,
      "learning_rate": 1.807691229385665e-05,
      "loss": 1.9175,
      "step": 8200
    },
    {
      "epoch": 0.8504098360655737,
      "grad_norm": 0.7669892311096191,
      "learning_rate": 1.7028285013959614e-05,
      "loss": 1.9074,
      "step": 8300
    },
    {
      "epoch": 0.860655737704918,
      "grad_norm": 0.6210095286369324,
      "learning_rate": 1.5798186437538642e-05,
      "loss": 1.9109,
      "step": 8400
    },
    {
      "epoch": 0.8709016393442623,
      "grad_norm": 0.6777961254119873,
      "learning_rate": 1.4418377880913716e-05,
      "loss": 1.9079,
      "step": 8500
    },
    {
      "epoch": 0.8811475409836066,
      "grad_norm": 0.5773517489433289,
      "learning_rate": 1.2924486192914705e-05,
      "loss": 1.9298,
      "step": 8600
    },
    {
      "epoch": 0.8913934426229508,
      "grad_norm": 0.5826138257980347,
      "learning_rate": 1.1355083864689736e-05,
      "loss": 1.9094,
      "step": 8700
    },
    {
      "epoch": 0.9016393442622951,
      "grad_norm": 0.6318073868751526,
      "learning_rate": 9.750693082619274e-06,
      "loss": 1.9103,
      "step": 8800
    },
    {
      "epoch": 0.9118852459016393,
      "grad_norm": 0.6698086261749268,
      "learning_rate": 8.15273943982811e-06,
      "loss": 1.9159,
      "step": 8900
    },
    {
      "epoch": 0.9221311475409836,
      "grad_norm": 0.612259566783905,
      "learning_rate": 6.602482321609596e-06,
      "loss": 1.9195,
      "step": 9000
    },
    {
      "epoch": 0.9323770491803278,
      "grad_norm": 0.5766026973724365,
      "learning_rate": 5.139949582359195e-06,
      "loss": 1.9105,
      "step": 9100
    },
    {
      "epoch": 0.9426229508196722,
      "grad_norm": 0.6413909196853638,
      "learning_rate": 3.8029040208070833e-06,
      "loss": 1.8875,
      "step": 9200
    },
    {
      "epoch": 0.9528688524590164,
      "grad_norm": 0.651620090007782,
      "learning_rate": 2.6258683393030516e-06,
      "loss": 1.9158,
      "step": 9300
    },
    {
      "epoch": 0.9631147540983607,
      "grad_norm": 0.6111190319061279,
      "learning_rate": 1.6392337628405675e-06,
      "loss": 1.9242,
      "step": 9400
    },
    {
      "epoch": 0.9733606557377049,
      "grad_norm": 0.6417702436447144,
      "learning_rate": 8.68475333400769e-07,
      "loss": 1.891,
      "step": 9500
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 0.5989698767662048,
      "learning_rate": 3.3762624823906577e-07,
      "loss": 1.9093,
      "step": 9600
    },
    {
      "epoch": 0.9938524590163934,
      "grad_norm": 0.7152329683303833,
      "learning_rate": 4.9692246345985905e-08,
      "loss": 1.8982,
      "step": 9700
    },
    {
      "epoch": 1.0,
      "step": 9760,
      "total_flos": 1.421397042852266e+18,
      "train_loss": 2.0159098046724915,
      "train_runtime": 3573.2428,
      "train_samples_per_second": 43.7,
      "train_steps_per_second": 2.731
    }
  ],
  "logging_steps": 100,
  "max_steps": 9760,
  "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": 1.421397042852266e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}