{
  "best_global_step": 700,
  "best_metric": 2.8837223052978516,
  "best_model_checkpoint": "./qwen-chat-finetuned/checkpoint-700",
  "epoch": 2.094426229508197,
  "eval_steps": 100,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.7374715983867646,
      "epoch": 0.02622950819672131,
      "grad_norm": 36.25,
      "learning_rate": 1.565217391304348e-06,
      "loss": 4.3802,
      "mean_token_accuracy": 0.352320809289813,
      "num_tokens": 53401.0,
      "step": 10
    },
    {
      "entropy": 1.7163084208965302,
      "epoch": 0.05245901639344262,
      "grad_norm": 76.0,
      "learning_rate": 3.3043478260869567e-06,
      "loss": 3.9169,
      "mean_token_accuracy": 0.3523058544844389,
      "num_tokens": 110972.0,
      "step": 20
    },
    {
      "entropy": 1.9017599046230316,
      "epoch": 0.07868852459016394,
      "grad_norm": 20.125,
      "learning_rate": 5.043478260869565e-06,
      "loss": 3.9013,
      "mean_token_accuracy": 0.3457519307732582,
      "num_tokens": 165938.0,
      "step": 30
    },
    {
      "entropy": 2.472129786014557,
      "epoch": 0.10491803278688525,
      "grad_norm": 9.625,
      "learning_rate": 6.782608695652174e-06,
      "loss": 3.3305,
      "mean_token_accuracy": 0.3779371060431004,
      "num_tokens": 222674.0,
      "step": 40
    },
    {
      "entropy": 2.775526815652847,
      "epoch": 0.13114754098360656,
      "grad_norm": 11.5625,
      "learning_rate": 8.521739130434783e-06,
      "loss": 3.2128,
      "mean_token_accuracy": 0.3930466562509537,
      "num_tokens": 278701.0,
      "step": 50
    },
    {
      "entropy": 2.78606663942337,
      "epoch": 0.15737704918032788,
      "grad_norm": 10.6875,
      "learning_rate": 1.0260869565217393e-05,
      "loss": 3.1361,
      "mean_token_accuracy": 0.38859146609902384,
      "num_tokens": 334399.0,
      "step": 60
    },
    {
      "entropy": 2.7956872820854186,
      "epoch": 0.18360655737704917,
      "grad_norm": 12.5625,
      "learning_rate": 1.2e-05,
      "loss": 2.9741,
      "mean_token_accuracy": 0.40764485821127894,
      "num_tokens": 388307.0,
      "step": 70
    },
    {
      "entropy": 2.6394418656826018,
      "epoch": 0.2098360655737705,
      "grad_norm": 8.5625,
      "learning_rate": 1.373913043478261e-05,
      "loss": 3.18,
      "mean_token_accuracy": 0.4003832370042801,
      "num_tokens": 440854.0,
      "step": 80
    },
    {
      "entropy": 2.5192583084106444,
      "epoch": 0.2360655737704918,
      "grad_norm": 7.90625,
      "learning_rate": 1.5478260869565217e-05,
      "loss": 2.8277,
      "mean_token_accuracy": 0.42245824187994,
      "num_tokens": 496198.0,
      "step": 90
    },
    {
      "entropy": 2.4694125235080717,
      "epoch": 0.26229508196721313,
      "grad_norm": 7.40625,
      "learning_rate": 1.721739130434783e-05,
      "loss": 2.9568,
      "mean_token_accuracy": 0.4054948791861534,
      "num_tokens": 549533.0,
      "step": 100
    },
    {
      "epoch": 0.26229508196721313,
      "eval_entropy": 2.3915325473336613,
      "eval_loss": 3.1522939205169678,
      "eval_mean_token_accuracy": 0.41849513439571157,
      "eval_num_tokens": 549533.0,
      "eval_runtime": 5.9294,
      "eval_samples_per_second": 228.521,
      "eval_steps_per_second": 28.67,
      "step": 100
    },
    {
      "entropy": 2.3309587478637694,
      "epoch": 0.28852459016393445,
      "grad_norm": 5.65625,
      "learning_rate": 1.8956521739130434e-05,
      "loss": 3.0738,
      "mean_token_accuracy": 0.4145698197185993,
      "num_tokens": 604712.0,
      "step": 110
    },
    {
      "entropy": 2.3101665914058684,
      "epoch": 0.31475409836065577,
      "grad_norm": 8.9375,
      "learning_rate": 1.9999257208329125e-05,
      "loss": 2.9566,
      "mean_token_accuracy": 0.4204997166991234,
      "num_tokens": 660327.0,
      "step": 120
    },
    {
      "entropy": 2.2639360249042513,
      "epoch": 0.34098360655737703,
      "grad_norm": 12.125,
      "learning_rate": 1.9990902069256562e-05,
      "loss": 3.0436,
      "mean_token_accuracy": 0.4053013317286968,
      "num_tokens": 712709.0,
      "step": 130
    },
    {
      "entropy": 2.2741274297237397,
      "epoch": 0.36721311475409835,
      "grad_norm": 8.625,
      "learning_rate": 1.9973271084541296e-05,
      "loss": 2.8725,
      "mean_token_accuracy": 0.43222680017352105,
      "num_tokens": 762841.0,
      "step": 140
    },
    {
      "entropy": 2.245914125442505,
      "epoch": 0.39344262295081966,
      "grad_norm": 9.6875,
      "learning_rate": 1.9946380623305116e-05,
      "loss": 2.924,
      "mean_token_accuracy": 0.42352988123893737,
      "num_tokens": 813780.0,
      "step": 150
    },
    {
      "entropy": 2.248101031780243,
      "epoch": 0.419672131147541,
      "grad_norm": 7.75,
      "learning_rate": 1.991025565143858e-05,
      "loss": 2.9713,
      "mean_token_accuracy": 0.41958039328455926,
      "num_tokens": 866574.0,
      "step": 160
    },
    {
      "entropy": 2.2917182624340056,
      "epoch": 0.4459016393442623,
      "grad_norm": 10.375,
      "learning_rate": 1.986492970842197e-05,
      "loss": 2.9419,
      "mean_token_accuracy": 0.41459466964006425,
      "num_tokens": 918240.0,
      "step": 170
    },
    {
      "entropy": 2.278603363037109,
      "epoch": 0.4721311475409836,
      "grad_norm": 6.59375,
      "learning_rate": 1.9810444876186235e-05,
      "loss": 2.6958,
      "mean_token_accuracy": 0.4311013951897621,
      "num_tokens": 977542.0,
      "step": 180
    },
    {
      "entropy": 2.246600490808487,
      "epoch": 0.49836065573770494,
      "grad_norm": 8.125,
      "learning_rate": 1.9746851740042906e-05,
      "loss": 2.8167,
      "mean_token_accuracy": 0.4296098329126835,
      "num_tokens": 1033813.0,
      "step": 190
    },
    {
      "entropy": 2.154702216386795,
      "epoch": 0.5245901639344263,
      "grad_norm": 15.125,
      "learning_rate": 1.9674209341719224e-05,
      "loss": 2.7415,
      "mean_token_accuracy": 0.4451119631528854,
      "num_tokens": 1085702.0,
      "step": 200
    },
    {
      "epoch": 0.5245901639344263,
      "eval_entropy": 2.186812935155981,
      "eval_loss": 3.0109927654266357,
      "eval_mean_token_accuracy": 0.43699205885915193,
      "eval_num_tokens": 1085702.0,
      "eval_runtime": 5.0574,
      "eval_samples_per_second": 267.925,
      "eval_steps_per_second": 33.614,
      "step": 200
    },
    {
      "entropy": 2.162086147069931,
      "epoch": 0.5508196721311476,
      "grad_norm": 7.03125,
      "learning_rate": 1.9592585124542055e-05,
      "loss": 2.9784,
      "mean_token_accuracy": 0.4247356228530407,
      "num_tokens": 1137871.0,
      "step": 210
    },
    {
      "entropy": 2.1591184198856355,
      "epoch": 0.5770491803278689,
      "grad_norm": 9.375,
      "learning_rate": 1.950205487082154e-05,
      "loss": 2.7191,
      "mean_token_accuracy": 0.4592340663075447,
      "num_tokens": 1195487.0,
      "step": 220
    },
    {
      "entropy": 2.1033347219228746,
      "epoch": 0.6032786885245902,
      "grad_norm": 6.5625,
      "learning_rate": 1.9402702631492595e-05,
      "loss": 2.8622,
      "mean_token_accuracy": 0.4423126816749573,
      "num_tokens": 1245528.0,
      "step": 230
    },
    {
      "entropy": 2.141283091902733,
      "epoch": 0.6295081967213115,
      "grad_norm": 6.96875,
      "learning_rate": 1.9294620648079543e-05,
      "loss": 2.8314,
      "mean_token_accuracy": 0.4316069222986698,
      "num_tokens": 1300795.0,
      "step": 240
    },
    {
      "entropy": 2.1625925838947295,
      "epoch": 0.6557377049180327,
      "grad_norm": 10.875,
      "learning_rate": 1.9177909267056403e-05,
      "loss": 3.0047,
      "mean_token_accuracy": 0.42482461109757425,
      "num_tokens": 1356910.0,
      "step": 250
    },
    {
      "entropy": 2.1598483502864836,
      "epoch": 0.6819672131147541,
      "grad_norm": 10.8125,
      "learning_rate": 1.9052676846682278e-05,
      "loss": 2.7993,
      "mean_token_accuracy": 0.43536330983042715,
      "num_tokens": 1412641.0,
      "step": 260
    },
    {
      "entropy": 2.128132125735283,
      "epoch": 0.7081967213114754,
      "grad_norm": 6.1875,
      "learning_rate": 1.8919039656398387e-05,
      "loss": 2.8191,
      "mean_token_accuracy": 0.4389350414276123,
      "num_tokens": 1470575.0,
      "step": 270
    },
    {
      "entropy": 2.1286911875009538,
      "epoch": 0.7344262295081967,
      "grad_norm": 11.125,
      "learning_rate": 1.8777121768880108e-05,
      "loss": 2.7836,
      "mean_token_accuracy": 0.4473900467157364,
      "num_tokens": 1529970.0,
      "step": 280
    },
    {
      "entropy": 2.1200373440980913,
      "epoch": 0.760655737704918,
      "grad_norm": 8.9375,
      "learning_rate": 1.862705494484429e-05,
      "loss": 2.9483,
      "mean_token_accuracy": 0.4232171691954136,
      "num_tokens": 1582507.0,
      "step": 290
    },
    {
      "entropy": 2.1294125616550446,
      "epoch": 0.7868852459016393,
      "grad_norm": 8.625,
      "learning_rate": 1.8468978510718734e-05,
      "loss": 2.9155,
      "mean_token_accuracy": 0.4285292446613312,
      "num_tokens": 1633961.0,
      "step": 300
    },
    {
      "epoch": 0.7868852459016393,
      "eval_entropy": 2.1470566707498886,
      "eval_loss": 2.9455931186676025,
      "eval_mean_token_accuracy": 0.4428922993295333,
      "eval_num_tokens": 1633961.0,
      "eval_runtime": 5.0322,
      "eval_samples_per_second": 269.267,
      "eval_steps_per_second": 33.783,
      "step": 300
    },
    {
      "entropy": 2.105194115638733,
      "epoch": 0.8131147540983606,
      "grad_norm": 10.0625,
      "learning_rate": 1.8303039229287476e-05,
      "loss": 2.804,
      "mean_token_accuracy": 0.4374969556927681,
      "num_tokens": 1691177.0,
      "step": 310
    },
    {
      "entropy": 2.056723091006279,
      "epoch": 0.839344262295082,
      "grad_norm": 12.9375,
      "learning_rate": 1.8129391163431912e-05,
      "loss": 2.6799,
      "mean_token_accuracy": 0.459722201526165,
      "num_tokens": 1746899.0,
      "step": 320
    },
    {
      "entropy": 2.0097413182258608,
      "epoch": 0.8655737704918033,
      "grad_norm": 11.0625,
      "learning_rate": 1.7948195533094343e-05,
      "loss": 2.6355,
      "mean_token_accuracy": 0.4685002237558365,
      "num_tokens": 1797548.0,
      "step": 330
    },
    {
      "entropy": 2.033376136422157,
      "epoch": 0.8918032786885246,
      "grad_norm": 13.875,
      "learning_rate": 1.7759620565596628e-05,
      "loss": 2.8326,
      "mean_token_accuracy": 0.4427059732377529,
      "num_tokens": 1850940.0,
      "step": 340
    },
    {
      "entropy": 2.1023750483989714,
      "epoch": 0.9180327868852459,
      "grad_norm": 7.71875,
      "learning_rate": 1.7563841339453053e-05,
      "loss": 2.7896,
      "mean_token_accuracy": 0.4486307971179485,
      "num_tokens": 1906789.0,
      "step": 350
    },
    {
      "entropy": 2.150878208875656,
      "epoch": 0.9442622950819672,
      "grad_norm": 11.3125,
      "learning_rate": 1.7361039621822288e-05,
      "loss": 2.9686,
      "mean_token_accuracy": 0.4197625443339348,
      "num_tokens": 1960496.0,
      "step": 360
    },
    {
      "entropy": 2.115389880537987,
      "epoch": 0.9704918032786886,
      "grad_norm": 12.75,
      "learning_rate": 1.7151403699749467e-05,
      "loss": 2.7354,
      "mean_token_accuracy": 0.4496216416358948,
      "num_tokens": 2018312.0,
      "step": 370
    },
    {
      "entropy": 2.0364568412303923,
      "epoch": 0.9967213114754099,
      "grad_norm": 15.25,
      "learning_rate": 1.693512820535498e-05,
      "loss": 2.7294,
      "mean_token_accuracy": 0.4530358798801899,
      "num_tokens": 2075123.0,
      "step": 380
    },
    {
      "entropy": 2.071015902467676,
      "epoch": 1.020983606557377,
      "grad_norm": 11.6875,
      "learning_rate": 1.6712413935132307e-05,
      "loss": 2.785,
      "mean_token_accuracy": 0.4375705743158186,
      "num_tokens": 2127638.0,
      "step": 390
    },
    {
      "entropy": 1.9601984202861786,
      "epoch": 1.0472131147540984,
      "grad_norm": 8.125,
      "learning_rate": 1.648346766352272e-05,
      "loss": 2.5404,
      "mean_token_accuracy": 0.4696895979344845,
      "num_tokens": 2182937.0,
      "step": 400
    },
    {
      "epoch": 1.0472131147540984,
      "eval_entropy": 1.962728219172534,
      "eval_loss": 2.9205315113067627,
      "eval_mean_token_accuracy": 0.44614641999497134,
      "eval_num_tokens": 2182937.0,
      "eval_runtime": 5.0426,
      "eval_samples_per_second": 268.709,
      "eval_steps_per_second": 33.713,
      "step": 400
    },
    {
      "entropy": 1.9237431466579438,
      "epoch": 1.0734426229508196,
      "grad_norm": 14.0625,
      "learning_rate": 1.6248501950939788e-05,
      "loss": 2.6068,
      "mean_token_accuracy": 0.4642325811088085,
      "num_tokens": 2238757.0,
      "step": 410
    },
    {
      "entropy": 1.9247209161520005,
      "epoch": 1.099672131147541,
      "grad_norm": 23.25,
      "learning_rate": 1.6007734946422125e-05,
      "loss": 2.5779,
      "mean_token_accuracy": 0.48758235424757,
      "num_tokens": 2293678.0,
      "step": 420
    },
    {
      "entropy": 1.9393797785043716,
      "epoch": 1.1259016393442622,
      "grad_norm": 5.9375,
      "learning_rate": 1.576139018509742e-05,
      "loss": 2.4135,
      "mean_token_accuracy": 0.4751130722463131,
      "num_tokens": 2345349.0,
      "step": 430
    },
    {
      "entropy": 1.8991259425878524,
      "epoch": 1.1521311475409837,
      "grad_norm": 11.375,
      "learning_rate": 1.550969638064589e-05,
      "loss": 2.5987,
      "mean_token_accuracy": 0.4642434611916542,
      "num_tokens": 2401578.0,
      "step": 440
    },
    {
      "entropy": 1.9015632569789886,
      "epoch": 1.1783606557377049,
      "grad_norm": 8.1875,
      "learning_rate": 1.525288721295581e-05,
      "loss": 2.4481,
      "mean_token_accuracy": 0.49190146625041964,
      "num_tokens": 2457689.0,
      "step": 450
    },
    {
      "entropy": 1.8667599350214004,
      "epoch": 1.2045901639344263,
      "grad_norm": 7.1875,
      "learning_rate": 1.4991201111168272e-05,
      "loss": 2.458,
      "mean_token_accuracy": 0.48113037571310996,
      "num_tokens": 2512267.0,
      "step": 460
    },
    {
      "entropy": 1.8301506459712982,
      "epoch": 1.2308196721311475,
      "grad_norm": 12.25,
      "learning_rate": 1.4724881032312603e-05,
      "loss": 2.3436,
      "mean_token_accuracy": 0.48804719001054764,
      "num_tokens": 2569873.0,
      "step": 470
    },
    {
      "entropy": 1.8779793590307237,
      "epoch": 1.257049180327869,
      "grad_norm": 6.84375,
      "learning_rate": 1.4454174235737943e-05,
      "loss": 2.5532,
      "mean_token_accuracy": 0.4793685719370842,
      "num_tokens": 2625844.0,
      "step": 480
    },
    {
      "entropy": 1.8586382299661637,
      "epoch": 1.2832786885245901,
      "grad_norm": 8.375,
      "learning_rate": 1.4179332053550452e-05,
      "loss": 2.4023,
      "mean_token_accuracy": 0.4916908532381058,
      "num_tokens": 2684182.0,
      "step": 490
    },
    {
      "entropy": 1.864977639913559,
      "epoch": 1.3095081967213114,
      "grad_norm": 10.8125,
      "learning_rate": 1.3900609657269222e-05,
      "loss": 2.4986,
      "mean_token_accuracy": 0.47516297921538353,
      "num_tokens": 2737200.0,
      "step": 500
    },
    {
      "epoch": 1.3095081967213114,
      "eval_entropy": 1.9044717620400822,
      "eval_loss": 2.9017913341522217,
      "eval_mean_token_accuracy": 0.45116734504699707,
      "eval_num_tokens": 2737200.0,
      "eval_runtime": 5.0643,
      "eval_samples_per_second": 267.557,
      "eval_steps_per_second": 33.568,
      "step": 500
    },
    {
      "entropy": 1.8558655947446823,
      "epoch": 1.3357377049180328,
      "grad_norm": 10.5625,
      "learning_rate": 1.3618265820917565e-05,
      "loss": 2.5359,
      "mean_token_accuracy": 0.46635654121637343,
      "num_tokens": 2788464.0,
      "step": 510
    },
    {
      "entropy": 1.8646740198135376,
      "epoch": 1.3619672131147542,
      "grad_norm": 11.875,
      "learning_rate": 1.333256268076967e-05,
      "loss": 2.5993,
      "mean_token_accuracy": 0.47755324840545654,
      "num_tokens": 2840606.0,
      "step": 520
    },
    {
      "entropy": 1.8370872467756272,
      "epoch": 1.3881967213114754,
      "grad_norm": 9.1875,
      "learning_rate": 1.3043765491975602e-05,
      "loss": 2.5856,
      "mean_token_accuracy": 0.47580934390425683,
      "num_tokens": 2889736.0,
      "step": 530
    },
    {
      "entropy": 1.8797135382890702,
      "epoch": 1.4144262295081966,
      "grad_norm": 11.375,
      "learning_rate": 1.275214238229066e-05,
      "loss": 2.6422,
      "mean_token_accuracy": 0.4679927438497543,
      "num_tokens": 2938632.0,
      "step": 540
    },
    {
      "entropy": 1.8948182821273805,
      "epoch": 1.440655737704918,
      "grad_norm": 8.8125,
      "learning_rate": 1.2457964103137747e-05,
      "loss": 2.5687,
      "mean_token_accuracy": 0.46481930911540986,
      "num_tokens": 2992166.0,
      "step": 550
    },
    {
      "entropy": 1.884895321726799,
      "epoch": 1.4668852459016393,
      "grad_norm": 4.46875,
      "learning_rate": 1.2161503778233813e-05,
      "loss": 2.4362,
      "mean_token_accuracy": 0.49427496194839476,
      "num_tokens": 3047092.0,
      "step": 560
    },
    {
      "entropy": 1.8731955915689469,
      "epoch": 1.4931147540983607,
      "grad_norm": 5.8125,
      "learning_rate": 1.1863036650013817e-05,
      "loss": 2.5261,
      "mean_token_accuracy": 0.48092877194285394,
      "num_tokens": 3103514.0,
      "step": 570
    },
    {
      "entropy": 1.8721953690052033,
      "epoch": 1.519344262295082,
      "grad_norm": 12.8125,
      "learning_rate": 1.1562839824087604e-05,
      "loss": 2.6489,
      "mean_token_accuracy": 0.46040653586387636,
      "num_tokens": 3154223.0,
      "step": 580
    },
    {
      "entropy": 1.8559091806411743,
      "epoch": 1.5455737704918033,
      "grad_norm": 9.0625,
      "learning_rate": 1.1261192011966952e-05,
      "loss": 2.5391,
      "mean_token_accuracy": 0.49064989015460014,
      "num_tokens": 3208440.0,
      "step": 590
    },
    {
      "entropy": 1.8500412583351136,
      "epoch": 1.5718032786885245,
      "grad_norm": 11.375,
      "learning_rate": 1.0958373272301647e-05,
      "loss": 2.667,
      "mean_token_accuracy": 0.45823334604501725,
      "num_tokens": 3261342.0,
      "step": 600
    },
    {
      "epoch": 1.5718032786885245,
      "eval_entropy": 1.8744515650412616,
      "eval_loss": 2.891632556915283,
      "eval_mean_token_accuracy": 0.4510051518678665,
      "eval_num_tokens": 3261342.0,
      "eval_runtime": 5.0441,
      "eval_samples_per_second": 268.629,
      "eval_steps_per_second": 33.703,
      "step": 600
    },
    {
      "entropy": 1.8289855808019637,
      "epoch": 1.5980327868852457,
      "grad_norm": 8.1875,
      "learning_rate": 1.0654664750864851e-05,
      "loss": 2.6261,
      "mean_token_accuracy": 0.4699882663786411,
      "num_tokens": 3312799.0,
      "step": 610
    },
    {
      "entropy": 1.8798219680786132,
      "epoch": 1.6242622950819672,
      "grad_norm": 9.875,
      "learning_rate": 1.0350348419529143e-05,
      "loss": 2.7662,
      "mean_token_accuracy": 0.4601142071187496,
      "num_tokens": 3365971.0,
      "step": 620
    },
    {
      "entropy": 1.8897109836339951,
      "epoch": 1.6504918032786886,
      "grad_norm": 5.15625,
      "learning_rate": 1.004570681447559e-05,
      "loss": 2.4977,
      "mean_token_accuracy": 0.48899385556578634,
      "num_tokens": 3422509.0,
      "step": 630
    },
    {
      "entropy": 1.8842475891113282,
      "epoch": 1.6767213114754098,
      "grad_norm": 5.40625,
      "learning_rate": 9.741022773878885e-06,
      "loss": 2.4265,
      "mean_token_accuracy": 0.48201805353164673,
      "num_tokens": 3482220.0,
      "step": 640
    },
    {
      "entropy": 1.8733120083808898,
      "epoch": 1.702950819672131,
      "grad_norm": 11.4375,
      "learning_rate": 9.436579175312121e-06,
      "loss": 2.5942,
      "mean_token_accuracy": 0.4757945977151394,
      "num_tokens": 3535417.0,
      "step": 650
    },
    {
      "entropy": 1.8299070119857788,
      "epoch": 1.7291803278688525,
      "grad_norm": 8.0625,
      "learning_rate": 9.132658673114977e-06,
      "loss": 2.5514,
      "mean_token_accuracy": 0.4865307964384556,
      "num_tokens": 3591470.0,
      "step": 660
    },
    {
      "entropy": 1.836677971482277,
      "epoch": 1.7554098360655739,
      "grad_norm": 11.5,
      "learning_rate": 8.829543435969176e-06,
      "loss": 2.3891,
      "mean_token_accuracy": 0.5033678129315377,
      "num_tokens": 3645508.0,
      "step": 670
    },
    {
      "entropy": 1.8698895812034606,
      "epoch": 1.781639344262295,
      "grad_norm": 5.59375,
      "learning_rate": 8.527514884924806e-06,
      "loss": 2.2491,
      "mean_token_accuracy": 0.5068600162863731,
      "num_tokens": 3708816.0,
      "step": 680
    },
    {
      "entropy": 1.7854702413082122,
      "epoch": 1.8078688524590163,
      "grad_norm": 8.75,
      "learning_rate": 8.226853432120808e-06,
      "loss": 2.5441,
      "mean_token_accuracy": 0.48928667306900026,
      "num_tokens": 3760527.0,
      "step": 690
    },
    {
      "entropy": 1.8351004660129546,
      "epoch": 1.8340983606557377,
      "grad_norm": 14.0,
      "learning_rate": 7.927838220442138e-06,
      "loss": 2.6377,
      "mean_token_accuracy": 0.46876697689294816,
      "num_tokens": 3815286.0,
      "step": 700
    },
    {
      "epoch": 1.8340983606557377,
      "eval_entropy": 1.8660439694628996,
      "eval_loss": 2.8837223052978516,
      "eval_mean_token_accuracy": 0.4532979618100559,
      "eval_num_tokens": 3815286.0,
      "eval_runtime": 5.0517,
      "eval_samples_per_second": 268.228,
      "eval_steps_per_second": 33.652,
      "step": 700
    },
    {
      "entropy": 1.8734222501516342,
      "epoch": 1.8603278688524592,
      "grad_norm": 11.3125,
      "learning_rate": 7.630746864355353e-06,
      "loss": 2.6619,
      "mean_token_accuracy": 0.46061802059412005,
      "num_tokens": 3867525.0,
      "step": 710
    },
    {
      "entropy": 1.8854580372571945,
      "epoch": 1.8865573770491804,
      "grad_norm": 6.25,
      "learning_rate": 7.335855192163215e-06,
      "loss": 2.5745,
      "mean_token_accuracy": 0.47656749188899994,
      "num_tokens": 3923519.0,
      "step": 720
    },
    {
      "entropy": 1.879018184542656,
      "epoch": 1.9127868852459016,
      "grad_norm": 5.625,
      "learning_rate": 7.043436989917631e-06,
      "loss": 2.528,
      "mean_token_accuracy": 0.4868933133780956,
      "num_tokens": 3981456.0,
      "step": 730
    },
    {
      "entropy": 1.8539914727210998,
      "epoch": 1.939016393442623,
      "grad_norm": 10.375,
      "learning_rate": 6.753763747228687e-06,
      "loss": 2.3218,
      "mean_token_accuracy": 0.5035036608576775,
      "num_tokens": 4036987.0,
      "step": 740
    },
    {
      "entropy": 1.8262196511030198,
      "epoch": 1.9652459016393444,
      "grad_norm": 6.1875,
      "learning_rate": 6.467104405205744e-06,
      "loss": 2.5221,
      "mean_token_accuracy": 0.4888892278075218,
      "num_tokens": 4091343.0,
      "step": 750
    },
    {
      "entropy": 1.8170676976442337,
      "epoch": 1.9914754098360654,
      "grad_norm": 6.4375,
      "learning_rate": 6.183725106764643e-06,
      "loss": 2.3265,
      "mean_token_accuracy": 0.512597742676735,
      "num_tokens": 4146393.0,
      "step": 760
    },
    {
      "entropy": 1.8072978258132935,
      "epoch": 2.015737704918033,
      "grad_norm": 8.9375,
      "learning_rate": 5.903888949532854e-06,
      "loss": 2.4277,
      "mean_token_accuracy": 0.49353572726249695,
      "num_tokens": 4193522.0,
      "step": 770
    },
    {
      "entropy": 1.8114411652088165,
      "epoch": 2.041967213114754,
      "grad_norm": 12.375,
      "learning_rate": 5.627855741581935e-06,
      "loss": 2.4582,
      "mean_token_accuracy": 0.4922221273183823,
      "num_tokens": 4248189.0,
      "step": 780
    },
    {
      "entropy": 1.7977081060409545,
      "epoch": 2.0681967213114754,
      "grad_norm": 4.5625,
      "learning_rate": 5.355881760214121e-06,
      "loss": 2.3459,
      "mean_token_accuracy": 0.49450727328658106,
      "num_tokens": 4300966.0,
      "step": 790
    },
    {
      "entropy": 1.797827911376953,
      "epoch": 2.094426229508197,
      "grad_norm": 7.25,
      "learning_rate": 5.088219514026984e-06,
      "loss": 2.384,
      "mean_token_accuracy": 0.4919115357100964,
      "num_tokens": 4354844.0,
      "step": 800
    },
    {
      "epoch": 2.094426229508197,
      "eval_entropy": 1.8060384897624746,
      "eval_loss": 2.8908982276916504,
      "eval_mean_token_accuracy": 0.4530702908249462,
      "eval_num_tokens": 4354844.0,
      "eval_runtime": 5.0625,
      "eval_samples_per_second": 267.652,
      "eval_steps_per_second": 33.58,
      "step": 800
    }
  ],
  "logging_steps": 10,
  "max_steps": 1146,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.254342993392435e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}