{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998682013882787,
  "eval_steps": 427,
  "global_step": 4266,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030225814954749144,
      "grad_norm": 8.637994766235352,
      "learning_rate": 4.777802058246361e-05,
      "loss": 0.7443,
      "step": 43
    },
    {
      "epoch": 0.06045162990949829,
      "grad_norm": 8.726279258728027,
      "learning_rate": 4.77420943412748e-05,
      "loss": 0.3904,
      "step": 86
    },
    {
      "epoch": 0.09067744486424743,
      "grad_norm": 18.751718521118164,
      "learning_rate": 4.768225729865107e-05,
      "loss": 0.2274,
      "step": 129
    },
    {
      "epoch": 0.12090325981899658,
      "grad_norm": 11.341946601867676,
      "learning_rate": 4.759856945148865e-05,
      "loss": 0.2032,
      "step": 172
    },
    {
      "epoch": 0.1511290747737457,
      "grad_norm": 6.5374932289123535,
      "learning_rate": 4.749111471120532e-05,
      "loss": 0.1546,
      "step": 215
    },
    {
      "epoch": 0.18135488972849487,
      "grad_norm": 7.3821539878845215,
      "learning_rate": 4.7360000819604815e-05,
      "loss": 0.177,
      "step": 258
    },
    {
      "epoch": 0.211580704683244,
      "grad_norm": 4.396503448486328,
      "learning_rate": 4.7205359240847186e-05,
      "loss": 0.1611,
      "step": 301
    },
    {
      "epoch": 0.24180651963799316,
      "grad_norm": 4.967313766479492,
      "learning_rate": 4.702734502963346e-05,
      "loss": 0.1408,
      "step": 344
    },
    {
      "epoch": 0.2720323345927423,
      "grad_norm": 9.22049331665039,
      "learning_rate": 4.6826136675736696e-05,
      "loss": 0.1235,
      "step": 387
    },
    {
      "epoch": 0.3001493717599508,
      "eval_accuracy": 0.9654081834354615,
      "eval_loss": 0.12473645806312561,
      "eval_precision": 0.9140316205533597,
      "eval_runtime": 42.4722,
      "eval_samples_per_second": 119.113,
      "eval_steps_per_second": 29.784,
      "step": 427
    },
    {
      "epoch": 0.3022581495474914,
      "grad_norm": 1.0973700284957886,
      "learning_rate": 4.660193592503536e-05,
      "loss": 0.1789,
      "step": 430
    },
    {
      "epoch": 0.3324839645022406,
      "grad_norm": 12.724760055541992,
      "learning_rate": 4.635496757722852e-05,
      "loss": 0.1526,
      "step": 473
    },
    {
      "epoch": 0.36270977945698973,
      "grad_norm": 0.692093014717102,
      "learning_rate": 4.60854792604356e-05,
      "loss": 0.0998,
      "step": 516
    },
    {
      "epoch": 0.39293559441173886,
      "grad_norm": 8.747737884521484,
      "learning_rate": 4.579374118290669e-05,
      "loss": 0.1417,
      "step": 559
    },
    {
      "epoch": 0.423161409366488,
      "grad_norm": 3.383861541748047,
      "learning_rate": 4.548004586209246e-05,
      "loss": 0.1416,
      "step": 602
    },
    {
      "epoch": 0.4533872243212371,
      "grad_norm": 2.690746545791626,
      "learning_rate": 4.514470783134528e-05,
      "loss": 0.1242,
      "step": 645
    },
    {
      "epoch": 0.4836130392759863,
      "grad_norm": 4.980698585510254,
      "learning_rate": 4.4788063324545544e-05,
      "loss": 0.0938,
      "step": 688
    },
    {
      "epoch": 0.5138388542307354,
      "grad_norm": 3.457531452178955,
      "learning_rate": 4.441046993896959e-05,
      "loss": 0.113,
      "step": 731
    },
    {
      "epoch": 0.5440646691854846,
      "grad_norm": 0.17746348679065704,
      "learning_rate": 4.40123062767371e-05,
      "loss": 0.086,
      "step": 774
    },
    {
      "epoch": 0.5742904841402338,
      "grad_norm": 1.7960216999053955,
      "learning_rate": 4.359397156519751e-05,
      "loss": 0.085,
      "step": 817
    },
    {
      "epoch": 0.6002987435199016,
      "eval_accuracy": 0.9703498715161099,
      "eval_loss": 0.190622478723526,
      "eval_precision": 0.918525703200776,
      "eval_runtime": 51.3537,
      "eval_samples_per_second": 98.513,
      "eval_steps_per_second": 24.633,
      "step": 854
    },
    {
      "epoch": 0.6045162990949828,
      "grad_norm": 0.05534067377448082,
      "learning_rate": 4.315588525663622e-05,
      "loss": 0.1477,
      "step": 860
    },
    {
      "epoch": 0.634742114049732,
      "grad_norm": 5.788769245147705,
      "learning_rate": 4.269848660770166e-05,
      "loss": 0.1854,
      "step": 903
    },
    {
      "epoch": 0.6649679290044812,
      "grad_norm": 4.581574440002441,
      "learning_rate": 4.222223423897522e-05,
      "loss": 0.0987,
      "step": 946
    },
    {
      "epoch": 0.6951937439592303,
      "grad_norm": 0.24250301718711853,
      "learning_rate": 4.1727605675125466e-05,
      "loss": 0.1237,
      "step": 989
    },
    {
      "epoch": 0.7254195589139795,
      "grad_norm": 6.998917579650879,
      "learning_rate": 4.1215096866107764e-05,
      "loss": 0.1304,
      "step": 1032
    },
    {
      "epoch": 0.7556453738687285,
      "grad_norm": 9.230533599853516,
      "learning_rate": 4.068522168988941e-05,
      "loss": 0.1212,
      "step": 1075
    },
    {
      "epoch": 0.7858711888234777,
      "grad_norm": 0.515550971031189,
      "learning_rate": 4.013851143719886e-05,
      "loss": 0.0942,
      "step": 1118
    },
    {
      "epoch": 0.8160970037782269,
      "grad_norm": 14.719630241394043,
      "learning_rate": 3.9575514278815625e-05,
      "loss": 0.1182,
      "step": 1161
    },
    {
      "epoch": 0.846322818732976,
      "grad_norm": 1.9909157752990723,
      "learning_rate": 3.899679471593512e-05,
      "loss": 0.099,
      "step": 1204
    },
    {
      "epoch": 0.8765486336877252,
      "grad_norm": 1.0096828937530518,
      "learning_rate": 3.840293301415938e-05,
      "loss": 0.1158,
      "step": 1247
    },
    {
      "epoch": 0.9004481152798524,
      "eval_accuracy": 0.9756868946432101,
      "eval_loss": 0.10856811702251434,
      "eval_precision": 0.941468253968254,
      "eval_runtime": 41.486,
      "eval_samples_per_second": 121.945,
      "eval_steps_per_second": 30.492,
      "step": 1281
    },
    {
      "epoch": 0.9067744486424743,
      "grad_norm": 0.669986367225647,
      "learning_rate": 3.779452462168126e-05,
      "loss": 0.101,
      "step": 1290
    },
    {
      "epoch": 0.9370002635972234,
      "grad_norm": 0.9615168571472168,
      "learning_rate": 3.7172179572245585e-05,
      "loss": 0.1166,
      "step": 1333
    },
    {
      "epoch": 0.9672260785519726,
      "grad_norm": 13.886622428894043,
      "learning_rate": 3.6536521873485673e-05,
      "loss": 0.1308,
      "step": 1376
    },
    {
      "epoch": 0.9974518935067217,
      "grad_norm": 0.23971609771251678,
      "learning_rate": 3.588818888124863e-05,
      "loss": 0.1266,
      "step": 1419
    },
    {
      "epoch": 1.0276777084614708,
      "grad_norm": 4.089532375335693,
      "learning_rate": 3.5227830660536916e-05,
      "loss": 0.0844,
      "step": 1462
    },
    {
      "epoch": 1.05790352341622,
      "grad_norm": 6.12522554397583,
      "learning_rate": 3.455610933370663e-05,
      "loss": 0.0843,
      "step": 1505
    },
    {
      "epoch": 1.0881293383709691,
      "grad_norm": 0.24417218565940857,
      "learning_rate": 3.3873698416576314e-05,
      "loss": 0.0825,
      "step": 1548
    },
    {
      "epoch": 1.1183551533257183,
      "grad_norm": 5.748846054077148,
      "learning_rate": 3.3181282143111886e-05,
      "loss": 0.0885,
      "step": 1591
    },
    {
      "epoch": 1.1485809682804675,
      "grad_norm": 0.8179301619529724,
      "learning_rate": 3.247955477936471e-05,
      "loss": 0.0882,
      "step": 1634
    },
    {
      "epoch": 1.1788067832352165,
      "grad_norm": 0.9235413670539856,
      "learning_rate": 3.176921992735089e-05,
      "loss": 0.0859,
      "step": 1677
    },
    {
      "epoch": 1.2005974870398033,
      "eval_accuracy": 0.9756868946432101,
      "eval_loss": 0.09278497099876404,
      "eval_precision": 0.9550102249488752,
      "eval_runtime": 41.4152,
      "eval_samples_per_second": 122.153,
      "eval_steps_per_second": 30.544,
      "step": 1708
    },
    {
      "epoch": 1.2090325981899657,
      "grad_norm": 0.0459032878279686,
      "learning_rate": 3.1050989819569484e-05,
      "loss": 0.0603,
      "step": 1720
    },
    {
      "epoch": 1.2392584131447149,
      "grad_norm": 2.4066104888916016,
      "learning_rate": 3.0325584604867403e-05,
      "loss": 0.071,
      "step": 1763
    },
    {
      "epoch": 1.269484228099464,
      "grad_norm": 9.008522987365723,
      "learning_rate": 2.9593731626366644e-05,
      "loss": 0.0678,
      "step": 1806
    },
    {
      "epoch": 1.2997100430542132,
      "grad_norm": 6.748093605041504,
      "learning_rate": 2.8856164692178087e-05,
      "loss": 0.1068,
      "step": 1849
    },
    {
      "epoch": 1.3299358580089624,
      "grad_norm": 0.8588669300079346,
      "learning_rate": 2.8113623339633016e-05,
      "loss": 0.0835,
      "step": 1892
    },
    {
      "epoch": 1.3601616729637114,
      "grad_norm": 2.512568235397339,
      "learning_rate": 2.736685209377016e-05,
      "loss": 0.0368,
      "step": 1935
    },
    {
      "epoch": 1.3903874879184606,
      "grad_norm": 3.670646905899048,
      "learning_rate": 2.661659972082166e-05,
      "loss": 0.0795,
      "step": 1978
    },
    {
      "epoch": 1.4206133028732097,
      "grad_norm": 0.5933941602706909,
      "learning_rate": 2.5863618477446485e-05,
      "loss": 0.0615,
      "step": 2021
    },
    {
      "epoch": 1.450839117827959,
      "grad_norm": 0.8403681516647339,
      "learning_rate": 2.5108663356464178e-05,
      "loss": 0.087,
      "step": 2064
    },
    {
      "epoch": 1.481064932782708,
      "grad_norm": 2.5098040103912354,
      "learning_rate": 2.4352491329845075e-05,
      "loss": 0.0821,
      "step": 2107
    },
    {
      "epoch": 1.500746858799754,
      "eval_accuracy": 0.9764775647361138,
      "eval_loss": 0.09056767076253891,
      "eval_precision": 0.935672514619883,
      "eval_runtime": 41.6907,
      "eval_samples_per_second": 121.346,
      "eval_steps_per_second": 30.342,
      "step": 2135
    },
    {
      "epoch": 1.511290747737457,
      "grad_norm": 8.705094337463379,
      "learning_rate": 2.3595860589716064e-05,
      "loss": 0.0714,
      "step": 2150
    },
    {
      "epoch": 1.5415165626922063,
      "grad_norm": 0.13956770300865173,
      "learning_rate": 2.2839529788143032e-05,
      "loss": 0.0696,
      "step": 2193
    },
    {
      "epoch": 1.5717423776469555,
      "grad_norm": 5.737203121185303,
      "learning_rate": 2.208425727645198e-05,
      "loss": 0.088,
      "step": 2236
    },
    {
      "epoch": 1.6019681926017046,
      "grad_norm": 2.958165407180786,
      "learning_rate": 2.1330800344851852e-05,
      "loss": 0.0797,
      "step": 2279
    },
    {
      "epoch": 1.6321940075564538,
      "grad_norm": 3.428828477859497,
      "learning_rate": 2.0579914463121218e-05,
      "loss": 0.0564,
      "step": 2322
    },
    {
      "epoch": 1.662419822511203,
      "grad_norm": 1.1754554510116577,
      "learning_rate": 1.983235252312024e-05,
      "loss": 0.0611,
      "step": 2365
    },
    {
      "epoch": 1.692645637465952,
      "grad_norm": 6.087173938751221,
      "learning_rate": 1.9088864083887505e-05,
      "loss": 0.0865,
      "step": 2408
    },
    {
      "epoch": 1.7228714524207012,
      "grad_norm": 3.829490900039673,
      "learning_rate": 1.835019462007857e-05,
      "loss": 0.0882,
      "step": 2451
    },
    {
      "epoch": 1.7530972673754504,
      "grad_norm": 7.614952087402344,
      "learning_rate": 1.761708477449973e-05,
      "loss": 0.0445,
      "step": 2494
    },
    {
      "epoch": 1.7833230823301993,
      "grad_norm": 35.23410415649414,
      "learning_rate": 1.689026961548663e-05,
      "loss": 0.0593,
      "step": 2537
    },
    {
      "epoch": 1.800896230559705,
      "eval_accuracy": 0.9794425775845028,
      "eval_loss": 0.07280829548835754,
      "eval_precision": 0.9513406156901688,
      "eval_runtime": 41.7842,
      "eval_samples_per_second": 121.075,
      "eval_steps_per_second": 30.275,
      "step": 2562
    },
    {
      "epoch": 1.8135488972849485,
      "grad_norm": 0.42250823974609375,
      "learning_rate": 1.6170477899872236e-05,
      "loss": 0.0682,
      "step": 2580
    },
    {
      "epoch": 1.8437747122396977,
      "grad_norm": 3.1888535022735596,
      "learning_rate": 1.5458431342283072e-05,
      "loss": 0.1021,
      "step": 2623
    },
    {
      "epoch": 1.8740005271944469,
      "grad_norm": 0.11506301164627075,
      "learning_rate": 1.4754843891496589e-05,
      "loss": 0.0655,
      "step": 2666
    },
    {
      "epoch": 1.904226342149196,
      "grad_norm": 0.1821933388710022,
      "learning_rate": 1.4060421014585115e-05,
      "loss": 0.0718,
      "step": 2709
    },
    {
      "epoch": 1.9344521571039452,
      "grad_norm": 4.954631805419922,
      "learning_rate": 1.3375858989564059e-05,
      "loss": 0.0727,
      "step": 2752
    },
    {
      "epoch": 1.9646779720586944,
      "grad_norm": 3.8650524616241455,
      "learning_rate": 1.270184420725387e-05,
      "loss": 0.0794,
      "step": 2795
    },
    {
      "epoch": 1.9949037870134436,
      "grad_norm": 1.1344525814056396,
      "learning_rate": 1.203905248305552e-05,
      "loss": 0.0832,
      "step": 2838
    },
    {
      "epoch": 2.025129601968193,
      "grad_norm": 0.8010286688804626,
      "learning_rate": 1.1388148379329648e-05,
      "loss": 0.0455,
      "step": 2881
    },
    {
      "epoch": 2.0553554169229415,
      "grad_norm": 0.04408084228634834,
      "learning_rate": 1.074978453905885e-05,
      "loss": 0.0328,
      "step": 2924
    },
    {
      "epoch": 2.0855812318776907,
      "grad_norm": 0.044715262949466705,
      "learning_rate": 1.0124601031461207e-05,
      "loss": 0.0527,
      "step": 2967
    },
    {
      "epoch": 2.1010456023196555,
      "eval_accuracy": 0.9774659023522435,
      "eval_loss": 0.08738358318805695,
      "eval_precision": 0.9481555333998006,
      "eval_runtime": 41.7468,
      "eval_samples_per_second": 121.183,
      "eval_steps_per_second": 30.302,
      "step": 2989
    },
    {
      "epoch": 2.11580704683244,
      "grad_norm": 0.9349909424781799,
      "learning_rate": 9.513224710211058e-06,
      "loss": 0.0401,
      "step": 3010
    },
    {
      "epoch": 2.146032861787189,
      "grad_norm": 9.170902252197266,
      "learning_rate": 8.91626858491075e-06,
      "loss": 0.0347,
      "step": 3053
    },
    {
      "epoch": 2.1762586767419383,
      "grad_norm": 0.14816665649414062,
      "learning_rate": 8.334331206443409e-06,
      "loss": 0.0252,
      "step": 3096
    },
    {
      "epoch": 2.2064844916966875,
      "grad_norm": 0.42291557788848877,
      "learning_rate": 7.767996066822973e-06,
      "loss": 0.054,
      "step": 3139
    },
    {
      "epoch": 2.2367103066514367,
      "grad_norm": 0.06259565055370331,
      "learning_rate": 7.2178310141434304e-06,
      "loss": 0.0534,
      "step": 3182
    },
    {
      "epoch": 2.266936121606186,
      "grad_norm": 0.24870078265666962,
      "learning_rate": 6.684387683213754e-06,
      "loss": 0.0706,
      "step": 3225
    },
    {
      "epoch": 2.297161936560935,
      "grad_norm": 4.754918098449707,
      "learning_rate": 6.168200942449347e-06,
      "loss": 0.0439,
      "step": 3268
    },
    {
      "epoch": 2.3273877515156842,
      "grad_norm": 0.26877379417419434,
      "learning_rate": 5.669788357574799e-06,
      "loss": 0.0586,
      "step": 3311
    },
    {
      "epoch": 2.357613566470433,
      "grad_norm": 0.19906023144721985,
      "learning_rate": 5.189649672675441e-06,
      "loss": 0.0324,
      "step": 3354
    },
    {
      "epoch": 2.387839381425182,
      "grad_norm": 0.01543935015797615,
      "learning_rate": 4.728266309118228e-06,
      "loss": 0.0263,
      "step": 3397
    },
    {
      "epoch": 2.4011949740796066,
      "eval_accuracy": 0.9794425775845028,
      "eval_loss": 0.0934256836771965,
      "eval_precision": 0.954045954045954,
      "eval_runtime": 41.7334,
      "eval_samples_per_second": 121.222,
      "eval_steps_per_second": 30.311,
      "step": 3416
    },
    {
      "epoch": 2.4180651963799313,
      "grad_norm": 23.245332717895508,
      "learning_rate": 4.286100882844285e-06,
      "loss": 0.029,
      "step": 3440
    },
    {
      "epoch": 2.4482910113346805,
      "grad_norm": 0.19245454668998718,
      "learning_rate": 3.863596740517041e-06,
      "loss": 0.0473,
      "step": 3483
    },
    {
      "epoch": 2.4785168262894297,
      "grad_norm": 0.045882448554039,
      "learning_rate": 3.461177514991228e-06,
      "loss": 0.0393,
      "step": 3526
    },
    {
      "epoch": 2.508742641244179,
      "grad_norm": 0.06130468472838402,
      "learning_rate": 3.079246700548261e-06,
      "loss": 0.0425,
      "step": 3569
    },
    {
      "epoch": 2.538968456198928,
      "grad_norm": 0.34582972526550293,
      "learning_rate": 2.718187248324039e-06,
      "loss": 0.0349,
      "step": 3612
    },
    {
      "epoch": 2.5691942711536773,
      "grad_norm": 1.3404638767242432,
      "learning_rate": 2.37836118233478e-06,
      "loss": 0.0362,
      "step": 3655
    },
    {
      "epoch": 2.5994200861084265,
      "grad_norm": 0.06872310489416122,
      "learning_rate": 2.060109236485834e-06,
      "loss": 0.0612,
      "step": 3698
    },
    {
      "epoch": 2.629645901063175,
      "grad_norm": 2.3359293937683105,
      "learning_rate": 1.7637505129275284e-06,
      "loss": 0.048,
      "step": 3741
    },
    {
      "epoch": 2.659871716017925,
      "grad_norm": 5.89841365814209,
      "learning_rate": 1.489582162100559e-06,
      "loss": 0.0357,
      "step": 3784
    },
    {
      "epoch": 2.6900975309726736,
      "grad_norm": 0.1834953874349594,
      "learning_rate": 1.2378790847916738e-06,
      "loss": 0.0566,
      "step": 3827
    },
    {
      "epoch": 2.701344345839557,
      "eval_accuracy": 0.9788495750148251,
      "eval_loss": 0.08470374345779419,
      "eval_precision": 0.9548192771084337,
      "eval_runtime": 41.7346,
      "eval_samples_per_second": 121.218,
      "eval_steps_per_second": 30.311,
      "step": 3843
    },
    {
      "epoch": 2.7203233459274228,
      "grad_norm": 6.001282215118408,
      "learning_rate": 1.008893656498535e-06,
      "loss": 0.0427,
      "step": 3870
    },
    {
      "epoch": 2.750549160882172,
      "grad_norm": 6.217470169067383,
      "learning_rate": 8.028554743799881e-07,
      "loss": 0.0367,
      "step": 3913
    },
    {
      "epoch": 2.780774975836921,
      "grad_norm": 4.40556526184082,
      "learning_rate": 6.199711270455366e-07,
      "loss": 0.0317,
      "step": 3956
    },
    {
      "epoch": 2.8110007907916703,
      "grad_norm": 0.33150240778923035,
      "learning_rate": 4.604239874148275e-07,
      "loss": 0.0429,
      "step": 3999
    },
    {
      "epoch": 2.8412266057464195,
      "grad_norm": 5.160287857055664,
      "learning_rate": 3.2437402885488323e-07,
      "loss": 0.0316,
      "step": 4042
    },
    {
      "epoch": 2.8714524207011687,
      "grad_norm": 0.10348004102706909,
      "learning_rate": 2.1195766477933204e-07,
      "loss": 0.032,
      "step": 4085
    },
    {
      "epoch": 2.901678235655918,
      "grad_norm": 0.10884183645248413,
      "learning_rate": 1.2328761187059095e-07,
      "loss": 0.0151,
      "step": 4128
    },
    {
      "epoch": 2.931904050610667,
      "grad_norm": 5.601053714752197,
      "learning_rate": 5.8452777062061706e-08,
      "loss": 0.0296,
      "step": 4171
    },
    {
      "epoch": 2.962129865565416,
      "grad_norm": 0.09085728973150253,
      "learning_rate": 1.7518168393672898e-08,
      "loss": 0.0384,
      "step": 4214
    },
    {
      "epoch": 2.9923556805201654,
      "grad_norm": 0.4429589807987213,
      "learning_rate": 5.248298301814908e-10,
      "loss": 0.0467,
      "step": 4257
    }
  ],
  "logging_steps": 43,
  "max_steps": 4266,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 4.964461063575552e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}