{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.7226567019922543,
  "eval_steps": 500,
  "global_step": 50000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.4451772745526446e-05,
      "grad_norm": 182.4078369140625,
      "learning_rate": 2.7225701061802342e-08,
      "loss": 13.9417,
      "step": 1
    },
    {
      "epoch": 0.027225886372763224,
      "grad_norm": 6.484030246734619,
      "learning_rate": 1.3612850530901172e-05,
      "loss": 8.9538,
      "step": 500
    },
    {
      "epoch": 0.05445177274552645,
      "grad_norm": 8.51893138885498,
      "learning_rate": 2.7225701061802345e-05,
      "loss": 6.3731,
      "step": 1000
    },
    {
      "epoch": 0.08167765911828967,
      "grad_norm": 5.699058532714844,
      "learning_rate": 4.083855159270352e-05,
      "loss": 5.4332,
      "step": 1500
    },
    {
      "epoch": 0.1089035454910529,
      "grad_norm": 11.347241401672363,
      "learning_rate": 5.445140212360469e-05,
      "loss": 4.6858,
      "step": 2000
    },
    {
      "epoch": 0.13612943186381612,
      "grad_norm": 2.8045220375061035,
      "learning_rate": 6.806425265450586e-05,
      "loss": 4.1909,
      "step": 2500
    },
    {
      "epoch": 0.16335531823657934,
      "grad_norm": 2.2558796405792236,
      "learning_rate": 8.167710318540703e-05,
      "loss": 3.929,
      "step": 3000
    },
    {
      "epoch": 0.19058120460934255,
      "grad_norm": 1.9897191524505615,
      "learning_rate": 9.52899537163082e-05,
      "loss": 3.7863,
      "step": 3500
    },
    {
      "epoch": 0.2178070909821058,
      "grad_norm": 1.875031590461731,
      "learning_rate": 9.999458213543955e-05,
      "loss": 3.6905,
      "step": 4000
    },
    {
      "epoch": 0.245032977354869,
      "grad_norm": 1.7890335321426392,
      "learning_rate": 9.996535015198725e-05,
      "loss": 3.6033,
      "step": 4500
    },
    {
      "epoch": 0.27225886372763225,
      "grad_norm": 1.7135872840881348,
      "learning_rate": 9.99108024021366e-05,
      "loss": 3.5396,
      "step": 5000
    },
    {
      "epoch": 0.29948475010039544,
      "grad_norm": 1.6878308057785034,
      "learning_rate": 9.983096652340218e-05,
      "loss": 3.4888,
      "step": 5500
    },
    {
      "epoch": 0.3267106364731587,
      "grad_norm": 1.6131703853607178,
      "learning_rate": 9.972588296594496e-05,
      "loss": 3.4457,
      "step": 6000
    },
    {
      "epoch": 0.3539365228459219,
      "grad_norm": 1.5502527952194214,
      "learning_rate": 9.959560497207756e-05,
      "loss": 3.414,
      "step": 6500
    },
    {
      "epoch": 0.3811624092186851,
      "grad_norm": 1.5919506549835205,
      "learning_rate": 9.944019854928815e-05,
      "loss": 3.3923,
      "step": 7000
    },
    {
      "epoch": 0.40838829559144835,
      "grad_norm": 1.558945655822754,
      "learning_rate": 9.925974243679667e-05,
      "loss": 3.3694,
      "step": 7500
    },
    {
      "epoch": 0.4356141819642116,
      "grad_norm": 1.5559173822402954,
      "learning_rate": 9.905432806566039e-05,
      "loss": 3.3422,
      "step": 8000
    },
    {
      "epoch": 0.4628400683369748,
      "grad_norm": 1.5621517896652222,
      "learning_rate": 9.882405951244874e-05,
      "loss": 3.3287,
      "step": 8500
    },
    {
      "epoch": 0.490065954709738,
      "grad_norm": 1.4939193725585938,
      "learning_rate": 9.856905344651121e-05,
      "loss": 3.3063,
      "step": 9000
    },
    {
      "epoch": 0.5172918410825013,
      "grad_norm": 1.4603230953216553,
      "learning_rate": 9.828943907086507e-05,
      "loss": 3.291,
      "step": 9500
    },
    {
      "epoch": 0.5445177274552645,
      "grad_norm": 1.4974786043167114,
      "learning_rate": 9.798535805673234e-05,
      "loss": 3.2838,
      "step": 10000
    },
    {
      "epoch": 0.5717436138280277,
      "grad_norm": 1.5649632215499878,
      "learning_rate": 9.765696447175982e-05,
      "loss": 3.2695,
      "step": 10500
    },
    {
      "epoch": 0.5989695002007909,
      "grad_norm": 1.523169994354248,
      "learning_rate": 9.73044247019582e-05,
      "loss": 3.2554,
      "step": 11000
    },
    {
      "epoch": 0.6261953865735541,
      "grad_norm": 1.4261916875839233,
      "learning_rate": 9.692791736739979e-05,
      "loss": 3.242,
      "step": 11500
    },
    {
      "epoch": 0.6534212729463174,
      "grad_norm": 1.4566397666931152,
      "learning_rate": 9.65276332317178e-05,
      "loss": 3.2229,
      "step": 12000
    },
    {
      "epoch": 0.6806471593190806,
      "grad_norm": 1.4012161493301392,
      "learning_rate": 9.610377510545277e-05,
      "loss": 3.223,
      "step": 12500
    },
    {
      "epoch": 0.7078730456918438,
      "grad_norm": 1.4838244915008545,
      "learning_rate": 9.565655774329525e-05,
      "loss": 3.2162,
      "step": 13000
    },
    {
      "epoch": 0.7350989320646071,
      "grad_norm": 1.4344415664672852,
      "learning_rate": 9.518620773527684e-05,
      "loss": 3.2148,
      "step": 13500
    },
    {
      "epoch": 0.7623248184373702,
      "grad_norm": 1.4415775537490845,
      "learning_rate": 9.469296339196459e-05,
      "loss": 3.206,
      "step": 14000
    },
    {
      "epoch": 0.7895507048101335,
      "grad_norm": 1.469138264656067,
      "learning_rate": 9.417707462371708e-05,
      "loss": 3.1997,
      "step": 14500
    },
    {
      "epoch": 0.8167765911828967,
      "grad_norm": 1.4184919595718384,
      "learning_rate": 9.36388028140632e-05,
      "loss": 3.1877,
      "step": 15000
    },
    {
      "epoch": 0.8440024775556599,
      "grad_norm": 1.4781111478805542,
      "learning_rate": 9.307842068726786e-05,
      "loss": 3.1893,
      "step": 15500
    },
    {
      "epoch": 0.8712283639284232,
      "grad_norm": 1.4547381401062012,
      "learning_rate": 9.249621217015181e-05,
      "loss": 3.1728,
      "step": 16000
    },
    {
      "epoch": 0.8984542503011864,
      "grad_norm": 1.4702293872833252,
      "learning_rate": 9.189247224823542e-05,
      "loss": 3.1727,
      "step": 16500
    },
    {
      "epoch": 0.9256801366739495,
      "grad_norm": 1.5003573894500732,
      "learning_rate": 9.126750681627938e-05,
      "loss": 3.1648,
      "step": 17000
    },
    {
      "epoch": 0.9529060230467128,
      "grad_norm": 1.43069589138031,
      "learning_rate": 9.062163252329815e-05,
      "loss": 3.1589,
      "step": 17500
    },
    {
      "epoch": 0.980131909419476,
      "grad_norm": 1.443028450012207,
      "learning_rate": 8.99551766121245e-05,
      "loss": 3.1447,
      "step": 18000
    },
    {
      "epoch": 1.0073918281502052,
      "grad_norm": 1.4987826347351074,
      "learning_rate": 8.926847675360655e-05,
      "loss": 3.1253,
      "step": 18500
    },
    {
      "epoch": 1.0346177145229685,
      "grad_norm": 1.514776349067688,
      "learning_rate": 8.856188087552123e-05,
      "loss": 3.0292,
      "step": 19000
    },
    {
      "epoch": 1.0618436008957317,
      "grad_norm": 1.464104175567627,
      "learning_rate": 8.783574698629109e-05,
      "loss": 3.0248,
      "step": 19500
    },
    {
      "epoch": 1.089069487268495,
      "grad_norm": 1.4995203018188477,
      "learning_rate": 8.709044299359327e-05,
      "loss": 3.0316,
      "step": 20000
    },
    {
      "epoch": 1.1162953736412582,
      "grad_norm": 1.5172643661499023,
      "learning_rate": 8.632634651795318e-05,
      "loss": 3.0303,
      "step": 20500
    },
    {
      "epoch": 1.1435212600140212,
      "grad_norm": 1.5371851921081543,
      "learning_rate": 8.554384470141671e-05,
      "loss": 3.0278,
      "step": 21000
    },
    {
      "epoch": 1.1707471463867845,
      "grad_norm": 1.4607304334640503,
      "learning_rate": 8.474333401139847e-05,
      "loss": 3.0206,
      "step": 21500
    },
    {
      "epoch": 1.1979730327595477,
      "grad_norm": 1.486690878868103,
      "learning_rate": 8.392522003980496e-05,
      "loss": 3.0251,
      "step": 22000
    },
    {
      "epoch": 1.225198919132311,
      "grad_norm": 1.4945247173309326,
      "learning_rate": 8.308991729753484e-05,
      "loss": 3.03,
      "step": 22500
    },
    {
      "epoch": 1.2524248055050742,
      "grad_norm": 1.4576785564422607,
      "learning_rate": 8.223784900446006e-05,
      "loss": 3.0219,
      "step": 23000
    },
    {
      "epoch": 1.2796506918778374,
      "grad_norm": 1.5146127939224243,
      "learning_rate": 8.136944687499468e-05,
      "loss": 3.015,
      "step": 23500
    },
    {
      "epoch": 1.3068765782506007,
      "grad_norm": 1.5249332189559937,
      "learning_rate": 8.048515089935956e-05,
      "loss": 3.0185,
      "step": 24000
    },
    {
      "epoch": 1.334102464623364,
      "grad_norm": 1.4822596311569214,
      "learning_rate": 7.958540912065415e-05,
      "loss": 3.0271,
      "step": 24500
    },
    {
      "epoch": 1.3613283509961271,
      "grad_norm": 1.5259217023849487,
      "learning_rate": 7.867067740784801e-05,
      "loss": 3.0188,
      "step": 25000
    },
    {
      "epoch": 1.3885542373688904,
      "grad_norm": 1.5163941383361816,
      "learning_rate": 7.774141922480738e-05,
      "loss": 3.0165,
      "step": 25500
    },
    {
      "epoch": 1.4157801237416536,
      "grad_norm": 1.6021302938461304,
      "learning_rate": 7.679810539547351e-05,
      "loss": 3.0178,
      "step": 26000
    },
    {
      "epoch": 1.4430060101144169,
      "grad_norm": 1.4801899194717407,
      "learning_rate": 7.584121386531205e-05,
      "loss": 3.0167,
      "step": 26500
    },
    {
      "epoch": 1.4702318964871801,
      "grad_norm": 1.478881597518921,
      "learning_rate": 7.487122945915404e-05,
      "loss": 3.0134,
      "step": 27000
    },
    {
      "epoch": 1.4974577828599434,
      "grad_norm": 1.5363502502441406,
      "learning_rate": 7.388864363555151e-05,
      "loss": 3.0162,
      "step": 27500
    },
    {
      "epoch": 1.5246836692327066,
      "grad_norm": 1.5000182390213013,
      "learning_rate": 7.289395423777183e-05,
      "loss": 3.0088,
      "step": 28000
    },
    {
      "epoch": 1.5519095556054698,
      "grad_norm": 1.4747732877731323,
      "learning_rate": 7.188766524155734e-05,
      "loss": 3.0004,
      "step": 28500
    },
    {
      "epoch": 1.5791354419782329,
      "grad_norm": 1.5123411417007446,
      "learning_rate": 7.08702864997777e-05,
      "loss": 3.0021,
      "step": 29000
    },
    {
      "epoch": 1.606361328350996,
      "grad_norm": 1.5053893327713013,
      "learning_rate": 6.984233348410447e-05,
      "loss": 3.0063,
      "step": 29500
    },
    {
      "epoch": 1.6335872147237593,
      "grad_norm": 1.4705040454864502,
      "learning_rate": 6.88043270238391e-05,
      "loss": 2.992,
      "step": 30000
    },
    {
      "epoch": 1.6608131010965226,
      "grad_norm": 1.4783118963241577,
      "learning_rate": 6.775679304202598e-05,
      "loss": 2.9957,
      "step": 30500
    },
    {
      "epoch": 1.6880389874692858,
      "grad_norm": 1.4604405164718628,
      "learning_rate": 6.670026228898501e-05,
      "loss": 3.0012,
      "step": 31000
    },
    {
      "epoch": 1.715264873842049,
      "grad_norm": 1.4742037057876587,
      "learning_rate": 6.563527007339827e-05,
      "loss": 2.9946,
      "step": 31500
    },
    {
      "epoch": 1.742490760214812,
      "grad_norm": 1.5092357397079468,
      "learning_rate": 6.456235599108689e-05,
      "loss": 2.9953,
      "step": 32000
    },
    {
      "epoch": 1.7697166465875753,
      "grad_norm": 1.4896336793899536,
      "learning_rate": 6.348206365161614e-05,
      "loss": 2.982,
      "step": 32500
    },
    {
      "epoch": 1.7969425329603386,
      "grad_norm": 1.5187143087387085,
      "learning_rate": 6.239494040286657e-05,
      "loss": 2.9954,
      "step": 33000
    },
    {
      "epoch": 1.8241684193331018,
      "grad_norm": 1.5126278400421143,
      "learning_rate": 6.13015370537113e-05,
      "loss": 2.9847,
      "step": 33500
    },
    {
      "epoch": 1.851394305705865,
      "grad_norm": 1.4799253940582275,
      "learning_rate": 6.0202407594939615e-05,
      "loss": 2.9843,
      "step": 34000
    },
    {
      "epoch": 1.8786201920786283,
      "grad_norm": 1.5327789783477783,
      "learning_rate": 5.9098108918568354e-05,
      "loss": 2.9804,
      "step": 34500
    },
    {
      "epoch": 1.9058460784513915,
      "grad_norm": 1.544638991355896,
      "learning_rate": 5.798920053568351e-05,
      "loss": 2.977,
      "step": 35000
    },
    {
      "epoch": 1.9330719648241548,
      "grad_norm": 1.5565699338912964,
      "learning_rate": 5.687624429295468e-05,
      "loss": 2.9804,
      "step": 35500
    },
    {
      "epoch": 1.960297851196918,
      "grad_norm": 1.4633305072784424,
      "learning_rate": 5.575980408796623e-05,
      "loss": 2.9758,
      "step": 36000
    },
    {
      "epoch": 1.9875237375696813,
      "grad_norm": 1.519126534461975,
      "learning_rate": 5.464044558350929e-05,
      "loss": 2.9697,
      "step": 36500
    },
    {
      "epoch": 2.0147836563004105,
      "grad_norm": 1.5512584447860718,
      "learning_rate": 5.3518735920979434e-05,
      "loss": 2.8931,
      "step": 37000
    },
    {
      "epoch": 2.0420095426731737,
      "grad_norm": 1.5433844327926636,
      "learning_rate": 5.239524343302511e-05,
      "loss": 2.8394,
      "step": 37500
    },
    {
      "epoch": 2.069235429045937,
      "grad_norm": 1.5760645866394043,
      "learning_rate": 5.127053735559263e-05,
      "loss": 2.8356,
      "step": 38000
    },
    {
      "epoch": 2.0964613154187,
      "grad_norm": 1.6160222291946411,
      "learning_rate": 5.0145187539513326e-05,
      "loss": 2.8333,
      "step": 38500
    },
    {
      "epoch": 2.1236872017914634,
      "grad_norm": 1.5705721378326416,
      "learning_rate": 4.901976416177933e-05,
      "loss": 2.8355,
      "step": 39000
    },
    {
      "epoch": 2.1509130881642267,
      "grad_norm": 1.5703595876693726,
      "learning_rate": 4.7894837436654e-05,
      "loss": 2.8399,
      "step": 39500
    },
    {
      "epoch": 2.17813897453699,
      "grad_norm": 1.5459660291671753,
      "learning_rate": 4.677097732676345e-05,
      "loss": 2.8474,
      "step": 40000
    },
    {
      "epoch": 2.205364860909753,
      "grad_norm": 1.6205192804336548,
      "learning_rate": 4.56487532543157e-05,
      "loss": 2.8455,
      "step": 40500
    },
    {
      "epoch": 2.2325907472825164,
      "grad_norm": 1.546263337135315,
      "learning_rate": 4.4528733812593465e-05,
      "loss": 2.8378,
      "step": 41000
    },
    {
      "epoch": 2.2598166336552796,
      "grad_norm": 1.641790509223938,
      "learning_rate": 4.341148647786699e-05,
      "loss": 2.8439,
      "step": 41500
    },
    {
      "epoch": 2.2870425200280424,
      "grad_norm": 1.5860189199447632,
      "learning_rate": 4.2297577321873025e-05,
      "loss": 2.8485,
      "step": 42000
    },
    {
      "epoch": 2.3142684064008057,
      "grad_norm": 1.578778624534607,
      "learning_rate": 4.118757072500509e-05,
      "loss": 2.8428,
      "step": 42500
    },
    {
      "epoch": 2.341494292773569,
      "grad_norm": 1.6200742721557617,
      "learning_rate": 4.008202909036093e-05,
      "loss": 2.8439,
      "step": 43000
    },
    {
      "epoch": 2.368720179146332,
      "grad_norm": 1.5700623989105225,
      "learning_rate": 3.898151255879183e-05,
      "loss": 2.8421,
      "step": 43500
    },
    {
      "epoch": 2.3959460655190954,
      "grad_norm": 1.608221173286438,
      "learning_rate": 3.788657872509786e-05,
      "loss": 2.8472,
      "step": 44000
    },
    {
      "epoch": 2.4231719518918586,
      "grad_norm": 1.5805708169937134,
      "learning_rate": 3.679778235551349e-05,
      "loss": 2.8437,
      "step": 44500
    },
    {
      "epoch": 2.450397838264622,
      "grad_norm": 1.5437690019607544,
      "learning_rate": 3.571567510662597e-05,
      "loss": 2.839,
      "step": 45000
    },
    {
      "epoch": 2.477623724637385,
      "grad_norm": 1.5659681558609009,
      "learning_rate": 3.464080524586959e-05,
      "loss": 2.8338,
      "step": 45500
    },
    {
      "epoch": 2.5048496110101484,
      "grad_norm": 1.5578417778015137,
      "learning_rate": 3.357371737373694e-05,
      "loss": 2.8417,
      "step": 46000
    },
    {
      "epoch": 2.5320754973829116,
      "grad_norm": 1.6232887506484985,
      "learning_rate": 3.2514952147848014e-05,
      "loss": 2.8424,
      "step": 46500
    },
    {
      "epoch": 2.559301383755675,
      "grad_norm": 1.580432415008545,
      "learning_rate": 3.146504600901725e-05,
      "loss": 2.8325,
      "step": 47000
    },
    {
      "epoch": 2.586527270128438,
      "grad_norm": 1.589545488357544,
      "learning_rate": 3.0424530909456973e-05,
      "loss": 2.8342,
      "step": 47500
    },
    {
      "epoch": 2.6137531565012013,
      "grad_norm": 1.5450959205627441,
      "learning_rate": 2.9393934043254943e-05,
      "loss": 2.838,
      "step": 48000
    },
    {
      "epoch": 2.6409790428739646,
      "grad_norm": 1.59015953540802,
      "learning_rate": 2.8373777579262905e-05,
      "loss": 2.8379,
      "step": 48500
    },
    {
      "epoch": 2.668204929246728,
      "grad_norm": 1.586267113685608,
      "learning_rate": 2.7364578396531094e-05,
      "loss": 2.8316,
      "step": 49000
    },
    {
      "epoch": 2.695430815619491,
      "grad_norm": 1.6080377101898193,
      "learning_rate": 2.636684782242303e-05,
      "loss": 2.8333,
      "step": 49500
    },
    {
      "epoch": 2.7226567019922543,
      "grad_norm": 1.6002167463302612,
      "learning_rate": 2.5381091373543046e-05,
      "loss": 2.8373,
      "step": 50000
    }
  ],
  "logging_steps": 500,
  "max_steps": 73456,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 50000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.074111485812618e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}