{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 198,
  "global_step": 1980,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010108668182966893,
      "grad_norm": 0.4873044788837433,
      "learning_rate": 0.0,
      "loss": 0.9551,
      "num_input_tokens_seen": 1998,
      "step": 1,
      "train_runtime": 4.5289,
      "train_tokens_per_second": 441.167
    },
    {
      "epoch": 0.0020217336365933787,
      "grad_norm": 0.4253540635108948,
      "learning_rate": 4e-05,
      "loss": 0.8258,
      "num_input_tokens_seen": 4262,
      "step": 2,
      "train_runtime": 6.4124,
      "train_tokens_per_second": 664.645
    },
    {
      "epoch": 0.003032600454890068,
      "grad_norm": 0.522351861000061,
      "learning_rate": 8e-05,
      "loss": 0.9482,
      "num_input_tokens_seen": 6200,
      "step": 3,
      "train_runtime": 8.1846,
      "train_tokens_per_second": 757.523
    },
    {
      "epoch": 0.004043467273186757,
      "grad_norm": 0.5153616666793823,
      "learning_rate": 0.00012,
      "loss": 0.8615,
      "num_input_tokens_seen": 8052,
      "step": 4,
      "train_runtime": 9.7901,
      "train_tokens_per_second": 822.461
    },
    {
      "epoch": 0.005054334091483447,
      "grad_norm": 0.7847296595573425,
      "learning_rate": 0.00016,
      "loss": 1.0249,
      "num_input_tokens_seen": 9692,
      "step": 5,
      "train_runtime": 11.4662,
      "train_tokens_per_second": 845.267
    },
    {
      "epoch": 0.006065200909780136,
      "grad_norm": 0.48074302077293396,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "num_input_tokens_seen": 11676,
      "step": 6,
      "train_runtime": 13.3663,
      "train_tokens_per_second": 873.539
    },
    {
      "epoch": 0.0070760677280768255,
      "grad_norm": 0.5613217353820801,
      "learning_rate": 0.0001998987341772152,
      "loss": 0.5166,
      "num_input_tokens_seen": 13238,
      "step": 7,
      "train_runtime": 14.9884,
      "train_tokens_per_second": 883.217
    },
    {
      "epoch": 0.008086934546373515,
      "grad_norm": 0.3947386145591736,
      "learning_rate": 0.0001997974683544304,
      "loss": 0.3962,
      "num_input_tokens_seen": 15272,
      "step": 8,
      "train_runtime": 16.7975,
      "train_tokens_per_second": 909.182
    },
    {
      "epoch": 0.009097801364670205,
      "grad_norm": 0.4128956198692322,
      "learning_rate": 0.00019969620253164558,
      "loss": 0.3897,
      "num_input_tokens_seen": 17090,
      "step": 9,
      "train_runtime": 18.5089,
      "train_tokens_per_second": 923.342
    },
    {
      "epoch": 0.010108668182966895,
      "grad_norm": 0.4657420217990875,
      "learning_rate": 0.00019959493670886075,
      "loss": 0.4569,
      "num_input_tokens_seen": 19018,
      "step": 10,
      "train_runtime": 20.1885,
      "train_tokens_per_second": 942.02
    },
    {
      "epoch": 0.011119535001263583,
      "grad_norm": 0.4372606873512268,
      "learning_rate": 0.00019949367088607596,
      "loss": 0.3667,
      "num_input_tokens_seen": 21072,
      "step": 11,
      "train_runtime": 21.9933,
      "train_tokens_per_second": 958.11
    },
    {
      "epoch": 0.012130401819560273,
      "grad_norm": 0.29161205887794495,
      "learning_rate": 0.00019939240506329115,
      "loss": 0.3203,
      "num_input_tokens_seen": 23612,
      "step": 12,
      "train_runtime": 24.0274,
      "train_tokens_per_second": 982.712
    },
    {
      "epoch": 0.013141268637856963,
      "grad_norm": 0.4845719039440155,
      "learning_rate": 0.00019929113924050632,
      "loss": 0.4122,
      "num_input_tokens_seen": 24886,
      "step": 13,
      "train_runtime": 25.5607,
      "train_tokens_per_second": 973.603
    },
    {
      "epoch": 0.014152135456153651,
      "grad_norm": 0.39806467294692993,
      "learning_rate": 0.00019918987341772154,
      "loss": 0.2891,
      "num_input_tokens_seen": 26718,
      "step": 14,
      "train_runtime": 27.2874,
      "train_tokens_per_second": 979.134
    },
    {
      "epoch": 0.015163002274450341,
      "grad_norm": 0.3543863296508789,
      "learning_rate": 0.00019908860759493673,
      "loss": 0.4272,
      "num_input_tokens_seen": 28968,
      "step": 15,
      "train_runtime": 29.2312,
      "train_tokens_per_second": 990.997
    },
    {
      "epoch": 0.01617386909274703,
      "grad_norm": 0.33626294136047363,
      "learning_rate": 0.0001989873417721519,
      "loss": 0.2618,
      "num_input_tokens_seen": 30542,
      "step": 16,
      "train_runtime": 30.8899,
      "train_tokens_per_second": 988.737
    },
    {
      "epoch": 0.01718473591104372,
      "grad_norm": 0.30460983514785767,
      "learning_rate": 0.0001988860759493671,
      "loss": 0.2975,
      "num_input_tokens_seen": 32276,
      "step": 17,
      "train_runtime": 32.5496,
      "train_tokens_per_second": 991.595
    },
    {
      "epoch": 0.01819560272934041,
      "grad_norm": 0.31722599267959595,
      "learning_rate": 0.00019878481012658227,
      "loss": 0.2809,
      "num_input_tokens_seen": 33716,
      "step": 18,
      "train_runtime": 34.1265,
      "train_tokens_per_second": 987.973
    },
    {
      "epoch": 0.0192064695476371,
      "grad_norm": 0.35153570771217346,
      "learning_rate": 0.00019868354430379746,
      "loss": 0.3387,
      "num_input_tokens_seen": 35506,
      "step": 19,
      "train_runtime": 35.7353,
      "train_tokens_per_second": 993.583
    },
    {
      "epoch": 0.02021733636593379,
      "grad_norm": 0.3597993552684784,
      "learning_rate": 0.00019858227848101268,
      "loss": 0.2891,
      "num_input_tokens_seen": 37408,
      "step": 20,
      "train_runtime": 37.4995,
      "train_tokens_per_second": 997.561
    },
    {
      "epoch": 0.02122820318423048,
      "grad_norm": 0.40770795941352844,
      "learning_rate": 0.00019848101265822785,
      "loss": 0.3535,
      "num_input_tokens_seen": 39344,
      "step": 21,
      "train_runtime": 39.1813,
      "train_tokens_per_second": 1004.154
    },
    {
      "epoch": 0.022239070002527166,
      "grad_norm": 0.29805222153663635,
      "learning_rate": 0.00019837974683544304,
      "loss": 0.3261,
      "num_input_tokens_seen": 41504,
      "step": 22,
      "train_runtime": 41.0211,
      "train_tokens_per_second": 1011.771
    },
    {
      "epoch": 0.023249936820823856,
      "grad_norm": 0.33333706855773926,
      "learning_rate": 0.00019827848101265825,
      "loss": 0.2981,
      "num_input_tokens_seen": 43398,
      "step": 23,
      "train_runtime": 42.7634,
      "train_tokens_per_second": 1014.841
    },
    {
      "epoch": 0.024260803639120546,
      "grad_norm": 0.3034307360649109,
      "learning_rate": 0.00019817721518987342,
      "loss": 0.2549,
      "num_input_tokens_seen": 45312,
      "step": 24,
      "train_runtime": 44.5424,
      "train_tokens_per_second": 1017.277
    },
    {
      "epoch": 0.025271670457417236,
      "grad_norm": 0.33897197246551514,
      "learning_rate": 0.0001980759493670886,
      "loss": 0.268,
      "num_input_tokens_seen": 46780,
      "step": 25,
      "train_runtime": 46.1397,
      "train_tokens_per_second": 1013.878
    },
    {
      "epoch": 0.026282537275713926,
      "grad_norm": 0.3111315667629242,
      "learning_rate": 0.0001979746835443038,
      "loss": 0.2719,
      "num_input_tokens_seen": 48436,
      "step": 26,
      "train_runtime": 47.8076,
      "train_tokens_per_second": 1013.145
    },
    {
      "epoch": 0.027293404094010616,
      "grad_norm": 0.3690702021121979,
      "learning_rate": 0.000197873417721519,
      "loss": 0.3425,
      "num_input_tokens_seen": 50246,
      "step": 27,
      "train_runtime": 49.4904,
      "train_tokens_per_second": 1015.268
    },
    {
      "epoch": 0.028304270912307302,
      "grad_norm": 0.29139432311058044,
      "learning_rate": 0.00019777215189873418,
      "loss": 0.2403,
      "num_input_tokens_seen": 52080,
      "step": 28,
      "train_runtime": 51.2374,
      "train_tokens_per_second": 1016.446
    },
    {
      "epoch": 0.029315137730603992,
      "grad_norm": 0.3654118776321411,
      "learning_rate": 0.00019767088607594937,
      "loss": 0.3419,
      "num_input_tokens_seen": 53724,
      "step": 29,
      "train_runtime": 52.9181,
      "train_tokens_per_second": 1015.23
    },
    {
      "epoch": 0.030326004548900682,
      "grad_norm": 0.3280225098133087,
      "learning_rate": 0.00019756962025316456,
      "loss": 0.2571,
      "num_input_tokens_seen": 55372,
      "step": 30,
      "train_runtime": 54.5194,
      "train_tokens_per_second": 1015.638
    },
    {
      "epoch": 0.03133687136719737,
      "grad_norm": 0.27247095108032227,
      "learning_rate": 0.00019746835443037975,
      "loss": 0.2048,
      "num_input_tokens_seen": 57012,
      "step": 31,
      "train_runtime": 58.5787,
      "train_tokens_per_second": 973.254
    },
    {
      "epoch": 0.03234773818549406,
      "grad_norm": 0.24623878300189972,
      "learning_rate": 0.00019736708860759494,
      "loss": 0.2631,
      "num_input_tokens_seen": 59328,
      "step": 32,
      "train_runtime": 60.5506,
      "train_tokens_per_second": 979.808
    },
    {
      "epoch": 0.03335860500379075,
      "grad_norm": 0.40265941619873047,
      "learning_rate": 0.00019726582278481014,
      "loss": 0.2294,
      "num_input_tokens_seen": 60640,
      "step": 33,
      "train_runtime": 62.0709,
      "train_tokens_per_second": 976.947
    },
    {
      "epoch": 0.03436947182208744,
      "grad_norm": 0.2988276779651642,
      "learning_rate": 0.00019716455696202533,
      "loss": 0.2675,
      "num_input_tokens_seen": 62976,
      "step": 34,
      "train_runtime": 63.956,
      "train_tokens_per_second": 984.678
    },
    {
      "epoch": 0.03538033864038413,
      "grad_norm": 0.37253138422966003,
      "learning_rate": 0.00019706329113924052,
      "loss": 0.2998,
      "num_input_tokens_seen": 64994,
      "step": 35,
      "train_runtime": 65.7974,
      "train_tokens_per_second": 987.79
    },
    {
      "epoch": 0.03639120545868082,
      "grad_norm": 0.37015262246131897,
      "learning_rate": 0.0001969620253164557,
      "loss": 0.2078,
      "num_input_tokens_seen": 66600,
      "step": 36,
      "train_runtime": 67.3916,
      "train_tokens_per_second": 988.254
    },
    {
      "epoch": 0.037402072276977505,
      "grad_norm": 0.34977278113365173,
      "learning_rate": 0.0001968607594936709,
      "loss": 0.3089,
      "num_input_tokens_seen": 68622,
      "step": 37,
      "train_runtime": 69.1144,
      "train_tokens_per_second": 992.875
    },
    {
      "epoch": 0.0384129390952742,
      "grad_norm": 0.3342492878437042,
      "learning_rate": 0.0001967594936708861,
      "loss": 0.3135,
      "num_input_tokens_seen": 70364,
      "step": 38,
      "train_runtime": 70.8181,
      "train_tokens_per_second": 993.587
    },
    {
      "epoch": 0.039423805913570885,
      "grad_norm": 0.34639793634414673,
      "learning_rate": 0.00019665822784810128,
      "loss": 0.2722,
      "num_input_tokens_seen": 72130,
      "step": 39,
      "train_runtime": 72.4387,
      "train_tokens_per_second": 995.739
    },
    {
      "epoch": 0.04043467273186758,
      "grad_norm": 0.30288997292518616,
      "learning_rate": 0.00019655696202531647,
      "loss": 0.1647,
      "num_input_tokens_seen": 74042,
      "step": 40,
      "train_runtime": 74.266,
      "train_tokens_per_second": 996.983
    },
    {
      "epoch": 0.041445539550164265,
      "grad_norm": 0.34732452034950256,
      "learning_rate": 0.00019645569620253166,
      "loss": 0.2737,
      "num_input_tokens_seen": 76178,
      "step": 41,
      "train_runtime": 76.6877,
      "train_tokens_per_second": 993.354
    },
    {
      "epoch": 0.04245640636846096,
      "grad_norm": 0.3373444378376007,
      "learning_rate": 0.00019635443037974683,
      "loss": 0.1763,
      "num_input_tokens_seen": 77916,
      "step": 42,
      "train_runtime": 78.3588,
      "train_tokens_per_second": 994.349
    },
    {
      "epoch": 0.043467273186757645,
      "grad_norm": 0.29695427417755127,
      "learning_rate": 0.00019625316455696204,
      "loss": 0.2235,
      "num_input_tokens_seen": 79524,
      "step": 43,
      "train_runtime": 79.9569,
      "train_tokens_per_second": 994.586
    },
    {
      "epoch": 0.04447814000505433,
      "grad_norm": 0.30779194831848145,
      "learning_rate": 0.00019615189873417723,
      "loss": 0.2355,
      "num_input_tokens_seen": 81330,
      "step": 44,
      "train_runtime": 81.7036,
      "train_tokens_per_second": 995.427
    },
    {
      "epoch": 0.045489006823351025,
      "grad_norm": 0.3187408745288849,
      "learning_rate": 0.0001960506329113924,
      "loss": 0.3051,
      "num_input_tokens_seen": 83266,
      "step": 45,
      "train_runtime": 83.404,
      "train_tokens_per_second": 998.345
    },
    {
      "epoch": 0.04649987364164771,
      "grad_norm": 0.31425026059150696,
      "learning_rate": 0.00019594936708860762,
      "loss": 0.2152,
      "num_input_tokens_seen": 85536,
      "step": 46,
      "train_runtime": 85.3767,
      "train_tokens_per_second": 1001.866
    },
    {
      "epoch": 0.047510740459944405,
      "grad_norm": 0.3499682545661926,
      "learning_rate": 0.0001958481012658228,
      "loss": 0.2051,
      "num_input_tokens_seen": 87530,
      "step": 47,
      "train_runtime": 87.1458,
      "train_tokens_per_second": 1004.409
    },
    {
      "epoch": 0.04852160727824109,
      "grad_norm": 0.29931217432022095,
      "learning_rate": 0.00019574683544303797,
      "loss": 0.2634,
      "num_input_tokens_seen": 89404,
      "step": 48,
      "train_runtime": 88.8415,
      "train_tokens_per_second": 1006.332
    },
    {
      "epoch": 0.04953247409653778,
      "grad_norm": 0.31406113505363464,
      "learning_rate": 0.0001956455696202532,
      "loss": 0.2824,
      "num_input_tokens_seen": 91598,
      "step": 49,
      "train_runtime": 90.6844,
      "train_tokens_per_second": 1010.074
    },
    {
      "epoch": 0.05054334091483447,
      "grad_norm": 0.2785133719444275,
      "learning_rate": 0.00019554430379746835,
      "loss": 0.2575,
      "num_input_tokens_seen": 93776,
      "step": 50,
      "train_runtime": 92.5267,
      "train_tokens_per_second": 1013.502
    },
    {
      "epoch": 0.05155420773313116,
      "grad_norm": 0.31789687275886536,
      "learning_rate": 0.00019544303797468354,
      "loss": 0.1854,
      "num_input_tokens_seen": 95578,
      "step": 51,
      "train_runtime": 94.2097,
      "train_tokens_per_second": 1014.524
    },
    {
      "epoch": 0.05256507455142785,
      "grad_norm": 0.3141697645187378,
      "learning_rate": 0.00019534177215189876,
      "loss": 0.1852,
      "num_input_tokens_seen": 97450,
      "step": 52,
      "train_runtime": 96.012,
      "train_tokens_per_second": 1014.977
    },
    {
      "epoch": 0.05357594136972454,
      "grad_norm": 0.3136524558067322,
      "learning_rate": 0.00019524050632911392,
      "loss": 0.2625,
      "num_input_tokens_seen": 99408,
      "step": 53,
      "train_runtime": 97.7861,
      "train_tokens_per_second": 1016.587
    },
    {
      "epoch": 0.05458680818802123,
      "grad_norm": 0.3242315351963043,
      "learning_rate": 0.00019513924050632912,
      "loss": 0.1751,
      "num_input_tokens_seen": 100788,
      "step": 54,
      "train_runtime": 99.346,
      "train_tokens_per_second": 1014.515
    },
    {
      "epoch": 0.05559767500631792,
      "grad_norm": 0.36038273572921753,
      "learning_rate": 0.00019503797468354433,
      "loss": 0.2772,
      "num_input_tokens_seen": 102596,
      "step": 55,
      "train_runtime": 101.0294,
      "train_tokens_per_second": 1015.506
    },
    {
      "epoch": 0.056608541824614604,
      "grad_norm": 0.35775113105773926,
      "learning_rate": 0.0001949367088607595,
      "loss": 0.3227,
      "num_input_tokens_seen": 104604,
      "step": 56,
      "train_runtime": 102.7993,
      "train_tokens_per_second": 1017.555
    },
    {
      "epoch": 0.0576194086429113,
      "grad_norm": 0.3312723934650421,
      "learning_rate": 0.0001948354430379747,
      "loss": 0.3182,
      "num_input_tokens_seen": 106376,
      "step": 57,
      "train_runtime": 104.4071,
      "train_tokens_per_second": 1018.858
    },
    {
      "epoch": 0.058630275461207984,
      "grad_norm": 0.22781355679035187,
      "learning_rate": 0.00019473417721518988,
      "loss": 0.1965,
      "num_input_tokens_seen": 108598,
      "step": 58,
      "train_runtime": 106.315,
      "train_tokens_per_second": 1021.474
    },
    {
      "epoch": 0.05964114227950468,
      "grad_norm": 0.38329920172691345,
      "learning_rate": 0.00019463291139240507,
      "loss": 0.2272,
      "num_input_tokens_seen": 110334,
      "step": 59,
      "train_runtime": 107.9998,
      "train_tokens_per_second": 1021.613
    },
    {
      "epoch": 0.060652009097801364,
      "grad_norm": 0.42992767691612244,
      "learning_rate": 0.00019453164556962026,
      "loss": 0.1986,
      "num_input_tokens_seen": 111732,
      "step": 60,
      "train_runtime": 109.564,
      "train_tokens_per_second": 1019.788
    },
    {
      "epoch": 0.06166287591609806,
      "grad_norm": 0.32373154163360596,
      "learning_rate": 0.00019443037974683545,
      "loss": 0.2735,
      "num_input_tokens_seen": 114016,
      "step": 61,
      "train_runtime": 113.7251,
      "train_tokens_per_second": 1002.558
    },
    {
      "epoch": 0.06267374273439474,
      "grad_norm": 0.34583255648612976,
      "learning_rate": 0.00019432911392405064,
      "loss": 0.2463,
      "num_input_tokens_seen": 115918,
      "step": 62,
      "train_runtime": 115.4902,
      "train_tokens_per_second": 1003.705
    },
    {
      "epoch": 0.06368460955269144,
      "grad_norm": 0.3125549852848053,
      "learning_rate": 0.00019422784810126583,
      "loss": 0.2226,
      "num_input_tokens_seen": 118102,
      "step": 63,
      "train_runtime": 117.3214,
      "train_tokens_per_second": 1006.653
    },
    {
      "epoch": 0.06469547637098812,
      "grad_norm": 0.29421722888946533,
      "learning_rate": 0.00019412658227848102,
      "loss": 0.1892,
      "num_input_tokens_seen": 120104,
      "step": 64,
      "train_runtime": 119.144,
      "train_tokens_per_second": 1008.057
    },
    {
      "epoch": 0.06570634318928481,
      "grad_norm": 0.41534146666526794,
      "learning_rate": 0.00019402531645569622,
      "loss": 0.2665,
      "num_input_tokens_seen": 121812,
      "step": 65,
      "train_runtime": 120.8223,
      "train_tokens_per_second": 1008.191
    },
    {
      "epoch": 0.0667172100075815,
      "grad_norm": 0.35454341769218445,
      "learning_rate": 0.00019392405063291138,
      "loss": 0.2388,
      "num_input_tokens_seen": 124036,
      "step": 66,
      "train_runtime": 122.7939,
      "train_tokens_per_second": 1010.115
    },
    {
      "epoch": 0.0677280768258782,
      "grad_norm": 0.31525930762290955,
      "learning_rate": 0.0001938227848101266,
      "loss": 0.221,
      "num_input_tokens_seen": 126360,
      "step": 67,
      "train_runtime": 124.7271,
      "train_tokens_per_second": 1013.091
    },
    {
      "epoch": 0.06873894364417488,
      "grad_norm": 0.3576384484767914,
      "learning_rate": 0.0001937215189873418,
      "loss": 0.2088,
      "num_input_tokens_seen": 128576,
      "step": 68,
      "train_runtime": 126.5829,
      "train_tokens_per_second": 1015.746
    },
    {
      "epoch": 0.06974981046247157,
      "grad_norm": 0.3103581368923187,
      "learning_rate": 0.00019362025316455695,
      "loss": 0.3964,
      "num_input_tokens_seen": 130716,
      "step": 69,
      "train_runtime": 128.4967,
      "train_tokens_per_second": 1017.271
    },
    {
      "epoch": 0.07076067728076826,
      "grad_norm": 0.3163643479347229,
      "learning_rate": 0.00019351898734177217,
      "loss": 0.3141,
      "num_input_tokens_seen": 132542,
      "step": 70,
      "train_runtime": 130.1847,
      "train_tokens_per_second": 1018.108
    },
    {
      "epoch": 0.07177154409906494,
      "grad_norm": 0.3137971758842468,
      "learning_rate": 0.00019341772151898736,
      "loss": 0.2044,
      "num_input_tokens_seen": 134086,
      "step": 71,
      "train_runtime": 131.8034,
      "train_tokens_per_second": 1017.319
    },
    {
      "epoch": 0.07278241091736164,
      "grad_norm": 0.29651039838790894,
      "learning_rate": 0.00019331645569620252,
      "loss": 0.157,
      "num_input_tokens_seen": 135710,
      "step": 72,
      "train_runtime": 133.4291,
      "train_tokens_per_second": 1017.095
    },
    {
      "epoch": 0.07379327773565833,
      "grad_norm": 0.29748690128326416,
      "learning_rate": 0.00019321518987341774,
      "loss": 0.2198,
      "num_input_tokens_seen": 137462,
      "step": 73,
      "train_runtime": 135.1885,
      "train_tokens_per_second": 1016.817
    },
    {
      "epoch": 0.07480414455395501,
      "grad_norm": 0.2688922584056854,
      "learning_rate": 0.0001931139240506329,
      "loss": 0.2184,
      "num_input_tokens_seen": 139066,
      "step": 74,
      "train_runtime": 136.793,
      "train_tokens_per_second": 1016.617
    },
    {
      "epoch": 0.0758150113722517,
      "grad_norm": 0.23618099093437195,
      "learning_rate": 0.0001930126582278481,
      "loss": 0.1644,
      "num_input_tokens_seen": 140966,
      "step": 75,
      "train_runtime": 138.4966,
      "train_tokens_per_second": 1017.83
    },
    {
      "epoch": 0.0768258781905484,
      "grad_norm": 0.2990943193435669,
      "learning_rate": 0.00019291139240506331,
      "loss": 0.2145,
      "num_input_tokens_seen": 142822,
      "step": 76,
      "train_runtime": 140.2559,
      "train_tokens_per_second": 1018.296
    },
    {
      "epoch": 0.07783674500884509,
      "grad_norm": 0.26312482357025146,
      "learning_rate": 0.00019281012658227848,
      "loss": 0.3069,
      "num_input_tokens_seen": 145096,
      "step": 77,
      "train_runtime": 142.2699,
      "train_tokens_per_second": 1019.864
    },
    {
      "epoch": 0.07884761182714177,
      "grad_norm": 0.26823654770851135,
      "learning_rate": 0.00019270886075949367,
      "loss": 0.1723,
      "num_input_tokens_seen": 147358,
      "step": 78,
      "train_runtime": 144.1752,
      "train_tokens_per_second": 1022.076
    },
    {
      "epoch": 0.07985847864543846,
      "grad_norm": 0.3408774435520172,
      "learning_rate": 0.0001926075949367089,
      "loss": 0.2358,
      "num_input_tokens_seen": 148978,
      "step": 79,
      "train_runtime": 145.7997,
      "train_tokens_per_second": 1021.799
    },
    {
      "epoch": 0.08086934546373516,
      "grad_norm": 0.4547388553619385,
      "learning_rate": 0.00019250632911392405,
      "loss": 0.2754,
      "num_input_tokens_seen": 150736,
      "step": 80,
      "train_runtime": 147.5363,
      "train_tokens_per_second": 1021.688
    },
    {
      "epoch": 0.08188021228203184,
      "grad_norm": 0.40629148483276367,
      "learning_rate": 0.00019240506329113924,
      "loss": 0.274,
      "num_input_tokens_seen": 152112,
      "step": 81,
      "train_runtime": 149.0717,
      "train_tokens_per_second": 1020.395
    },
    {
      "epoch": 0.08289107910032853,
      "grad_norm": 0.39234459400177,
      "learning_rate": 0.00019230379746835443,
      "loss": 0.2205,
      "num_input_tokens_seen": 154138,
      "step": 82,
      "train_runtime": 150.8524,
      "train_tokens_per_second": 1021.78
    },
    {
      "epoch": 0.08390194591862522,
      "grad_norm": 0.3805030584335327,
      "learning_rate": 0.00019220253164556962,
      "loss": 0.2239,
      "num_input_tokens_seen": 155784,
      "step": 83,
      "train_runtime": 152.511,
      "train_tokens_per_second": 1021.461
    },
    {
      "epoch": 0.08491281273692192,
      "grad_norm": 0.3097169101238251,
      "learning_rate": 0.00019210126582278481,
      "loss": 0.1396,
      "num_input_tokens_seen": 157204,
      "step": 84,
      "train_runtime": 154.0995,
      "train_tokens_per_second": 1020.146
    },
    {
      "epoch": 0.0859236795552186,
      "grad_norm": 0.44679346680641174,
      "learning_rate": 0.000192,
      "loss": 0.2751,
      "num_input_tokens_seen": 158824,
      "step": 85,
      "train_runtime": 155.7905,
      "train_tokens_per_second": 1019.472
    },
    {
      "epoch": 0.08693454637351529,
      "grad_norm": 0.4235474765300751,
      "learning_rate": 0.0001918987341772152,
      "loss": 0.2625,
      "num_input_tokens_seen": 160912,
      "step": 86,
      "train_runtime": 157.6529,
      "train_tokens_per_second": 1020.673
    },
    {
      "epoch": 0.08794541319181198,
      "grad_norm": 0.35191550850868225,
      "learning_rate": 0.0001917974683544304,
      "loss": 0.2039,
      "num_input_tokens_seen": 163074,
      "step": 87,
      "train_runtime": 159.4907,
      "train_tokens_per_second": 1022.467
    },
    {
      "epoch": 0.08895628001010866,
      "grad_norm": 0.4148600399494171,
      "learning_rate": 0.00019169620253164558,
      "loss": 0.1857,
      "num_input_tokens_seen": 164552,
      "step": 88,
      "train_runtime": 161.097,
      "train_tokens_per_second": 1021.447
    },
    {
      "epoch": 0.08996714682840536,
      "grad_norm": 0.4624585211277008,
      "learning_rate": 0.00019159493670886077,
      "loss": 0.242,
      "num_input_tokens_seen": 166358,
      "step": 89,
      "train_runtime": 162.8719,
      "train_tokens_per_second": 1021.404
    },
    {
      "epoch": 0.09097801364670205,
      "grad_norm": 0.4638700783252716,
      "learning_rate": 0.00019149367088607596,
      "loss": 0.3036,
      "num_input_tokens_seen": 168030,
      "step": 90,
      "train_runtime": 164.5661,
      "train_tokens_per_second": 1021.048
    },
    {
      "epoch": 0.09198888046499874,
      "grad_norm": 0.321336030960083,
      "learning_rate": 0.00019139240506329115,
      "loss": 0.2824,
      "num_input_tokens_seen": 170032,
      "step": 91,
      "train_runtime": 168.3683,
      "train_tokens_per_second": 1009.881
    },
    {
      "epoch": 0.09299974728329542,
      "grad_norm": 0.39739641547203064,
      "learning_rate": 0.00019129113924050634,
      "loss": 0.2554,
      "num_input_tokens_seen": 171476,
      "step": 92,
      "train_runtime": 169.9782,
      "train_tokens_per_second": 1008.812
    },
    {
      "epoch": 0.09401061410159212,
      "grad_norm": 0.34846046566963196,
      "learning_rate": 0.00019118987341772153,
      "loss": 0.2037,
      "num_input_tokens_seen": 173370,
      "step": 93,
      "train_runtime": 171.6984,
      "train_tokens_per_second": 1009.736
    },
    {
      "epoch": 0.09502148091988881,
      "grad_norm": 0.3384343683719635,
      "learning_rate": 0.00019108860759493672,
      "loss": 0.1363,
      "num_input_tokens_seen": 175448,
      "step": 94,
      "train_runtime": 173.525,
      "train_tokens_per_second": 1011.082
    },
    {
      "epoch": 0.09603234773818549,
      "grad_norm": 0.3351871073246002,
      "learning_rate": 0.0001909873417721519,
      "loss": 0.1844,
      "num_input_tokens_seen": 177164,
      "step": 95,
      "train_runtime": 175.1545,
      "train_tokens_per_second": 1011.473
    },
    {
      "epoch": 0.09704321455648218,
      "grad_norm": 0.40943101048469543,
      "learning_rate": 0.0001908860759493671,
      "loss": 0.2371,
      "num_input_tokens_seen": 178574,
      "step": 96,
      "train_runtime": 176.7506,
      "train_tokens_per_second": 1010.316
    },
    {
      "epoch": 0.09805408137477888,
      "grad_norm": 0.34281614422798157,
      "learning_rate": 0.0001907848101265823,
      "loss": 0.2115,
      "num_input_tokens_seen": 180732,
      "step": 97,
      "train_runtime": 178.6029,
      "train_tokens_per_second": 1011.921
    },
    {
      "epoch": 0.09906494819307556,
      "grad_norm": 0.3453739285469055,
      "learning_rate": 0.00019068354430379746,
      "loss": 0.2594,
      "num_input_tokens_seen": 183094,
      "step": 98,
      "train_runtime": 180.6157,
      "train_tokens_per_second": 1013.722
    },
    {
      "epoch": 0.10007581501137225,
      "grad_norm": 0.3091121017932892,
      "learning_rate": 0.00019058227848101268,
      "loss": 0.292,
      "num_input_tokens_seen": 185286,
      "step": 99,
      "train_runtime": 182.5425,
      "train_tokens_per_second": 1015.029
    },
    {
      "epoch": 0.10108668182966894,
      "grad_norm": 0.3359338641166687,
      "learning_rate": 0.00019048101265822787,
      "loss": 0.2609,
      "num_input_tokens_seen": 186894,
      "step": 100,
      "train_runtime": 184.2232,
      "train_tokens_per_second": 1014.498
    },
    {
      "epoch": 0.10209754864796564,
      "grad_norm": 0.39591389894485474,
      "learning_rate": 0.00019037974683544303,
      "loss": 0.3126,
      "num_input_tokens_seen": 188862,
      "step": 101,
      "train_runtime": 186.0581,
      "train_tokens_per_second": 1015.07
    },
    {
      "epoch": 0.10310841546626232,
      "grad_norm": 0.2912481129169464,
      "learning_rate": 0.00019027848101265825,
      "loss": 0.3256,
      "num_input_tokens_seen": 190642,
      "step": 102,
      "train_runtime": 187.7639,
      "train_tokens_per_second": 1015.328
    },
    {
      "epoch": 0.10411928228455901,
      "grad_norm": 0.30756765604019165,
      "learning_rate": 0.00019017721518987344,
      "loss": 0.2803,
      "num_input_tokens_seen": 192360,
      "step": 103,
      "train_runtime": 189.5118,
      "train_tokens_per_second": 1015.029
    },
    {
      "epoch": 0.1051301491028557,
      "grad_norm": 0.2889096736907959,
      "learning_rate": 0.0001900759493670886,
      "loss": 0.2099,
      "num_input_tokens_seen": 194740,
      "step": 104,
      "train_runtime": 191.5138,
      "train_tokens_per_second": 1016.846
    },
    {
      "epoch": 0.10614101592115238,
      "grad_norm": 0.33699703216552734,
      "learning_rate": 0.00018997468354430382,
      "loss": 0.2379,
      "num_input_tokens_seen": 197156,
      "step": 105,
      "train_runtime": 193.5292,
      "train_tokens_per_second": 1018.74
    },
    {
      "epoch": 0.10715188273944908,
      "grad_norm": 0.42451784014701843,
      "learning_rate": 0.00018987341772151899,
      "loss": 0.2275,
      "num_input_tokens_seen": 198724,
      "step": 106,
      "train_runtime": 195.1297,
      "train_tokens_per_second": 1018.42
    },
    {
      "epoch": 0.10816274955774577,
      "grad_norm": 0.35566970705986023,
      "learning_rate": 0.00018977215189873418,
      "loss": 0.3391,
      "num_input_tokens_seen": 200392,
      "step": 107,
      "train_runtime": 196.7609,
      "train_tokens_per_second": 1018.454
    },
    {
      "epoch": 0.10917361637604246,
      "grad_norm": 0.30473417043685913,
      "learning_rate": 0.0001896708860759494,
      "loss": 0.332,
      "num_input_tokens_seen": 202462,
      "step": 108,
      "train_runtime": 198.6138,
      "train_tokens_per_second": 1019.376
    },
    {
      "epoch": 0.11018448319433914,
      "grad_norm": 0.451630175113678,
      "learning_rate": 0.00018956962025316456,
      "loss": 0.1844,
      "num_input_tokens_seen": 204072,
      "step": 109,
      "train_runtime": 200.2296,
      "train_tokens_per_second": 1019.19
    },
    {
      "epoch": 0.11119535001263584,
      "grad_norm": 0.43709272146224976,
      "learning_rate": 0.00018946835443037975,
      "loss": 0.3005,
      "num_input_tokens_seen": 205860,
      "step": 110,
      "train_runtime": 201.8833,
      "train_tokens_per_second": 1019.698
    },
    {
      "epoch": 0.11220621683093253,
      "grad_norm": 0.28327277302742004,
      "learning_rate": 0.00018936708860759497,
      "loss": 0.1795,
      "num_input_tokens_seen": 207538,
      "step": 111,
      "train_runtime": 203.5769,
      "train_tokens_per_second": 1019.458
    },
    {
      "epoch": 0.11321708364922921,
      "grad_norm": 0.5263530611991882,
      "learning_rate": 0.00018926582278481013,
      "loss": 0.2676,
      "num_input_tokens_seen": 209106,
      "step": 112,
      "train_runtime": 205.2913,
      "train_tokens_per_second": 1018.582
    },
    {
      "epoch": 0.1142279504675259,
      "grad_norm": 0.37919676303863525,
      "learning_rate": 0.00018916455696202532,
      "loss": 0.2701,
      "num_input_tokens_seen": 210810,
      "step": 113,
      "train_runtime": 206.9779,
      "train_tokens_per_second": 1018.514
    },
    {
      "epoch": 0.1152388172858226,
      "grad_norm": 0.45175403356552124,
      "learning_rate": 0.0001890632911392405,
      "loss": 0.2468,
      "num_input_tokens_seen": 212672,
      "step": 114,
      "train_runtime": 208.6858,
      "train_tokens_per_second": 1019.101
    },
    {
      "epoch": 0.11624968410411929,
      "grad_norm": 0.431393027305603,
      "learning_rate": 0.0001889620253164557,
      "loss": 0.2553,
      "num_input_tokens_seen": 214800,
      "step": 115,
      "train_runtime": 210.6283,
      "train_tokens_per_second": 1019.806
    },
    {
      "epoch": 0.11726055092241597,
      "grad_norm": 0.43366989493370056,
      "learning_rate": 0.0001888607594936709,
      "loss": 0.2309,
      "num_input_tokens_seen": 216784,
      "step": 116,
      "train_runtime": 212.4119,
      "train_tokens_per_second": 1020.583
    },
    {
      "epoch": 0.11827141774071266,
      "grad_norm": 0.48804113268852234,
      "learning_rate": 0.00018875949367088608,
      "loss": 0.2464,
      "num_input_tokens_seen": 218508,
      "step": 117,
      "train_runtime": 214.1126,
      "train_tokens_per_second": 1020.528
    },
    {
      "epoch": 0.11928228455900936,
      "grad_norm": 0.49241119623184204,
      "learning_rate": 0.00018865822784810128,
      "loss": 0.246,
      "num_input_tokens_seen": 220070,
      "step": 118,
      "train_runtime": 215.7771,
      "train_tokens_per_second": 1019.895
    },
    {
      "epoch": 0.12029315137730603,
      "grad_norm": 0.3341336250305176,
      "learning_rate": 0.00018855696202531647,
      "loss": 0.1921,
      "num_input_tokens_seen": 221972,
      "step": 119,
      "train_runtime": 217.5522,
      "train_tokens_per_second": 1020.316
    },
    {
      "epoch": 0.12130401819560273,
      "grad_norm": 0.37389177083969116,
      "learning_rate": 0.00018845569620253166,
      "loss": 0.1806,
      "num_input_tokens_seen": 223738,
      "step": 120,
      "train_runtime": 219.291,
      "train_tokens_per_second": 1020.279
    },
    {
      "epoch": 0.12231488501389942,
      "grad_norm": 0.3373318016529083,
      "learning_rate": 0.00018835443037974685,
      "loss": 0.146,
      "num_input_tokens_seen": 225128,
      "step": 121,
      "train_runtime": 223.1066,
      "train_tokens_per_second": 1009.06
    },
    {
      "epoch": 0.12332575183219611,
      "grad_norm": 0.3741705119609833,
      "learning_rate": 0.000188253164556962,
      "loss": 0.2245,
      "num_input_tokens_seen": 226914,
      "step": 122,
      "train_runtime": 224.8128,
      "train_tokens_per_second": 1009.346
    },
    {
      "epoch": 0.1243366186504928,
      "grad_norm": 0.35638299584388733,
      "learning_rate": 0.00018815189873417723,
      "loss": 0.1852,
      "num_input_tokens_seen": 228850,
      "step": 123,
      "train_runtime": 226.5813,
      "train_tokens_per_second": 1010.013
    },
    {
      "epoch": 0.1253474854687895,
      "grad_norm": 0.3549748957157135,
      "learning_rate": 0.00018805063291139242,
      "loss": 0.2209,
      "num_input_tokens_seen": 230448,
      "step": 124,
      "train_runtime": 228.1882,
      "train_tokens_per_second": 1009.903
    },
    {
      "epoch": 0.12635835228708617,
      "grad_norm": 0.425045371055603,
      "learning_rate": 0.00018794936708860758,
      "loss": 0.2184,
      "num_input_tokens_seen": 232126,
      "step": 125,
      "train_runtime": 229.8679,
      "train_tokens_per_second": 1009.824
    },
    {
      "epoch": 0.12736921910538287,
      "grad_norm": 0.3064917027950287,
      "learning_rate": 0.0001878481012658228,
      "loss": 0.2005,
      "num_input_tokens_seen": 234350,
      "step": 126,
      "train_runtime": 231.7397,
      "train_tokens_per_second": 1011.264
    },
    {
      "epoch": 0.12838008592367955,
      "grad_norm": 0.3639183044433594,
      "learning_rate": 0.000187746835443038,
      "loss": 0.3228,
      "num_input_tokens_seen": 236724,
      "step": 127,
      "train_runtime": 233.7157,
      "train_tokens_per_second": 1012.872
    },
    {
      "epoch": 0.12939095274197623,
      "grad_norm": 0.4401392638683319,
      "learning_rate": 0.00018764556962025316,
      "loss": 0.2403,
      "num_input_tokens_seen": 238836,
      "step": 128,
      "train_runtime": 235.6082,
      "train_tokens_per_second": 1013.7
    },
    {
      "epoch": 0.13040181956027294,
      "grad_norm": 0.40510451793670654,
      "learning_rate": 0.00018754430379746837,
      "loss": 0.2402,
      "num_input_tokens_seen": 240594,
      "step": 129,
      "train_runtime": 237.2283,
      "train_tokens_per_second": 1014.187
    },
    {
      "epoch": 0.13141268637856962,
      "grad_norm": 0.4582486152648926,
      "learning_rate": 0.00018744303797468354,
      "loss": 0.27,
      "num_input_tokens_seen": 242398,
      "step": 130,
      "train_runtime": 238.8715,
      "train_tokens_per_second": 1014.763
    },
    {
      "epoch": 0.1324235531968663,
      "grad_norm": 0.35521116852760315,
      "learning_rate": 0.00018734177215189873,
      "loss": 0.2177,
      "num_input_tokens_seen": 244080,
      "step": 131,
      "train_runtime": 240.4803,
      "train_tokens_per_second": 1014.969
    },
    {
      "epoch": 0.133434420015163,
      "grad_norm": 0.3867698013782501,
      "learning_rate": 0.00018724050632911395,
      "loss": 0.3857,
      "num_input_tokens_seen": 245858,
      "step": 132,
      "train_runtime": 242.2765,
      "train_tokens_per_second": 1014.782
    },
    {
      "epoch": 0.1344452868334597,
      "grad_norm": 0.29385513067245483,
      "learning_rate": 0.0001871392405063291,
      "loss": 0.2192,
      "num_input_tokens_seen": 247804,
      "step": 133,
      "train_runtime": 244.066,
      "train_tokens_per_second": 1015.315
    },
    {
      "epoch": 0.1354561536517564,
      "grad_norm": 0.2909926176071167,
      "learning_rate": 0.0001870379746835443,
      "loss": 0.2115,
      "num_input_tokens_seen": 249694,
      "step": 134,
      "train_runtime": 245.8824,
      "train_tokens_per_second": 1015.502
    },
    {
      "epoch": 0.13646702047005307,
      "grad_norm": 0.3936549127101898,
      "learning_rate": 0.00018693670886075952,
      "loss": 0.222,
      "num_input_tokens_seen": 251606,
      "step": 135,
      "train_runtime": 247.7136,
      "train_tokens_per_second": 1015.713
    },
    {
      "epoch": 0.13747788728834975,
      "grad_norm": 0.3142557740211487,
      "learning_rate": 0.00018683544303797468,
      "loss": 0.2112,
      "num_input_tokens_seen": 253332,
      "step": 136,
      "train_runtime": 249.3943,
      "train_tokens_per_second": 1015.789
    },
    {
      "epoch": 0.13848875410664646,
      "grad_norm": 0.3500168025493622,
      "learning_rate": 0.00018673417721518987,
      "loss": 0.1939,
      "num_input_tokens_seen": 254974,
      "step": 137,
      "train_runtime": 251.1025,
      "train_tokens_per_second": 1015.418
    },
    {
      "epoch": 0.13949962092494314,
      "grad_norm": 0.3195367157459259,
      "learning_rate": 0.00018663291139240507,
      "loss": 0.3256,
      "num_input_tokens_seen": 256608,
      "step": 138,
      "train_runtime": 252.7901,
      "train_tokens_per_second": 1015.103
    },
    {
      "epoch": 0.14051048774323982,
      "grad_norm": 0.2949298620223999,
      "learning_rate": 0.00018653164556962026,
      "loss": 0.198,
      "num_input_tokens_seen": 258288,
      "step": 139,
      "train_runtime": 254.4778,
      "train_tokens_per_second": 1014.973
    },
    {
      "epoch": 0.14152135456153653,
      "grad_norm": 0.313806414604187,
      "learning_rate": 0.00018643037974683545,
      "loss": 0.1996,
      "num_input_tokens_seen": 260148,
      "step": 140,
      "train_runtime": 256.3055,
      "train_tokens_per_second": 1014.992
    },
    {
      "epoch": 0.1425322213798332,
      "grad_norm": 0.4024338126182556,
      "learning_rate": 0.00018632911392405064,
      "loss": 0.1987,
      "num_input_tokens_seen": 261654,
      "step": 141,
      "train_runtime": 257.8953,
      "train_tokens_per_second": 1014.574
    },
    {
      "epoch": 0.1435430881981299,
      "grad_norm": 0.45130419731140137,
      "learning_rate": 0.00018622784810126583,
      "loss": 0.1909,
      "num_input_tokens_seen": 263218,
      "step": 142,
      "train_runtime": 259.4938,
      "train_tokens_per_second": 1014.352
    },
    {
      "epoch": 0.1445539550164266,
      "grad_norm": 0.541994571685791,
      "learning_rate": 0.00018612658227848102,
      "loss": 0.2816,
      "num_input_tokens_seen": 264632,
      "step": 143,
      "train_runtime": 261.0919,
      "train_tokens_per_second": 1013.559
    },
    {
      "epoch": 0.14556482183472327,
      "grad_norm": 0.3926849663257599,
      "learning_rate": 0.0001860253164556962,
      "loss": 0.1541,
      "num_input_tokens_seen": 266864,
      "step": 144,
      "train_runtime": 262.9617,
      "train_tokens_per_second": 1014.84
    },
    {
      "epoch": 0.14657568865301995,
      "grad_norm": 0.39028400182724,
      "learning_rate": 0.0001859240506329114,
      "loss": 0.2019,
      "num_input_tokens_seen": 268848,
      "step": 145,
      "train_runtime": 264.8027,
      "train_tokens_per_second": 1015.277
    },
    {
      "epoch": 0.14758655547131666,
      "grad_norm": 0.417856901884079,
      "learning_rate": 0.0001858227848101266,
      "loss": 0.1816,
      "num_input_tokens_seen": 270722,
      "step": 146,
      "train_runtime": 266.5571,
      "train_tokens_per_second": 1015.625
    },
    {
      "epoch": 0.14859742228961334,
      "grad_norm": 0.41069379448890686,
      "learning_rate": 0.00018572151898734178,
      "loss": 0.1997,
      "num_input_tokens_seen": 272600,
      "step": 147,
      "train_runtime": 268.3283,
      "train_tokens_per_second": 1015.92
    },
    {
      "epoch": 0.14960828910791002,
      "grad_norm": 0.3944964110851288,
      "learning_rate": 0.00018562025316455697,
      "loss": 0.2941,
      "num_input_tokens_seen": 274420,
      "step": 148,
      "train_runtime": 269.984,
      "train_tokens_per_second": 1016.431
    },
    {
      "epoch": 0.15061915592620673,
      "grad_norm": 0.27445194125175476,
      "learning_rate": 0.00018551898734177216,
      "loss": 0.2761,
      "num_input_tokens_seen": 276326,
      "step": 149,
      "train_runtime": 271.7695,
      "train_tokens_per_second": 1016.766
    },
    {
      "epoch": 0.1516300227445034,
      "grad_norm": 0.5121089220046997,
      "learning_rate": 0.00018541772151898736,
      "loss": 0.2627,
      "num_input_tokens_seen": 277910,
      "step": 150,
      "train_runtime": 273.3918,
      "train_tokens_per_second": 1016.526
    },
    {
      "epoch": 0.15264088956280011,
      "grad_norm": 0.45720142126083374,
      "learning_rate": 0.00018531645569620255,
      "loss": 0.1706,
      "num_input_tokens_seen": 279798,
      "step": 151,
      "train_runtime": 276.9048,
      "train_tokens_per_second": 1010.448
    },
    {
      "epoch": 0.1536517563810968,
      "grad_norm": 0.31884294748306274,
      "learning_rate": 0.00018521518987341774,
      "loss": 0.1553,
      "num_input_tokens_seen": 281512,
      "step": 152,
      "train_runtime": 278.6516,
      "train_tokens_per_second": 1010.265
    },
    {
      "epoch": 0.15466262319939347,
      "grad_norm": 0.37187591195106506,
      "learning_rate": 0.00018511392405063293,
      "loss": 0.2538,
      "num_input_tokens_seen": 283374,
      "step": 153,
      "train_runtime": 280.3843,
      "train_tokens_per_second": 1010.663
    },
    {
      "epoch": 0.15567349001769018,
      "grad_norm": 0.4017643630504608,
      "learning_rate": 0.0001850126582278481,
      "loss": 0.2104,
      "num_input_tokens_seen": 285236,
      "step": 154,
      "train_runtime": 282.035,
      "train_tokens_per_second": 1011.35
    },
    {
      "epoch": 0.15668435683598686,
      "grad_norm": 0.3263136148452759,
      "learning_rate": 0.0001849113924050633,
      "loss": 0.2665,
      "num_input_tokens_seen": 287594,
      "step": 155,
      "train_runtime": 284.034,
      "train_tokens_per_second": 1012.534
    },
    {
      "epoch": 0.15769522365428354,
      "grad_norm": 0.39439326524734497,
      "learning_rate": 0.0001848101265822785,
      "loss": 0.216,
      "num_input_tokens_seen": 289730,
      "step": 156,
      "train_runtime": 285.8955,
      "train_tokens_per_second": 1013.412
    },
    {
      "epoch": 0.15870609047258025,
      "grad_norm": 0.39812779426574707,
      "learning_rate": 0.00018470886075949366,
      "loss": 0.2137,
      "num_input_tokens_seen": 291880,
      "step": 157,
      "train_runtime": 287.7431,
      "train_tokens_per_second": 1014.377
    },
    {
      "epoch": 0.15971695729087693,
      "grad_norm": 0.43934911489486694,
      "learning_rate": 0.00018460759493670888,
      "loss": 0.2193,
      "num_input_tokens_seen": 293984,
      "step": 158,
      "train_runtime": 289.5963,
      "train_tokens_per_second": 1015.151
    },
    {
      "epoch": 0.1607278241091736,
      "grad_norm": 0.4106093943119049,
      "learning_rate": 0.00018450632911392407,
      "loss": 0.2226,
      "num_input_tokens_seen": 296006,
      "step": 159,
      "train_runtime": 291.4338,
      "train_tokens_per_second": 1015.689
    },
    {
      "epoch": 0.1617386909274703,
      "grad_norm": 0.45763731002807617,
      "learning_rate": 0.00018440506329113924,
      "loss": 0.2325,
      "num_input_tokens_seen": 297474,
      "step": 160,
      "train_runtime": 293.06,
      "train_tokens_per_second": 1015.062
    },
    {
      "epoch": 0.162749557745767,
      "grad_norm": 0.4370255768299103,
      "learning_rate": 0.00018430379746835445,
      "loss": 0.2749,
      "num_input_tokens_seen": 299154,
      "step": 161,
      "train_runtime": 294.7502,
      "train_tokens_per_second": 1014.941
    },
    {
      "epoch": 0.16376042456406367,
      "grad_norm": 0.4296151101589203,
      "learning_rate": 0.00018420253164556962,
      "loss": 0.2165,
      "num_input_tokens_seen": 300562,
      "step": 162,
      "train_runtime": 296.3494,
      "train_tokens_per_second": 1014.215
    },
    {
      "epoch": 0.16477129138236038,
      "grad_norm": 0.3881031274795532,
      "learning_rate": 0.0001841012658227848,
      "loss": 0.2646,
      "num_input_tokens_seen": 302242,
      "step": 163,
      "train_runtime": 298.027,
      "train_tokens_per_second": 1014.143
    },
    {
      "epoch": 0.16578215820065706,
      "grad_norm": 0.39300647377967834,
      "learning_rate": 0.00018400000000000003,
      "loss": 0.1683,
      "num_input_tokens_seen": 303996,
      "step": 164,
      "train_runtime": 299.7329,
      "train_tokens_per_second": 1014.223
    },
    {
      "epoch": 0.16679302501895377,
      "grad_norm": 0.3293854892253876,
      "learning_rate": 0.0001838987341772152,
      "loss": 0.2492,
      "num_input_tokens_seen": 305954,
      "step": 165,
      "train_runtime": 301.5051,
      "train_tokens_per_second": 1014.756
    },
    {
      "epoch": 0.16780389183725045,
      "grad_norm": 0.3706817328929901,
      "learning_rate": 0.00018379746835443038,
      "loss": 0.242,
      "num_input_tokens_seen": 307566,
      "step": 166,
      "train_runtime": 303.1229,
      "train_tokens_per_second": 1014.658
    },
    {
      "epoch": 0.16881475865554713,
      "grad_norm": 0.3900180160999298,
      "learning_rate": 0.0001836962025316456,
      "loss": 0.2467,
      "num_input_tokens_seen": 309354,
      "step": 167,
      "train_runtime": 304.841,
      "train_tokens_per_second": 1014.804
    },
    {
      "epoch": 0.16982562547384383,
      "grad_norm": 0.41680508852005005,
      "learning_rate": 0.00018359493670886076,
      "loss": 0.2333,
      "num_input_tokens_seen": 310996,
      "step": 168,
      "train_runtime": 306.5227,
      "train_tokens_per_second": 1014.594
    },
    {
      "epoch": 0.1708364922921405,
      "grad_norm": 0.29431745409965515,
      "learning_rate": 0.00018349367088607595,
      "loss": 0.157,
      "num_input_tokens_seen": 313404,
      "step": 169,
      "train_runtime": 308.5847,
      "train_tokens_per_second": 1015.618
    },
    {
      "epoch": 0.1718473591104372,
      "grad_norm": 0.3451148569583893,
      "learning_rate": 0.00018339240506329115,
      "loss": 0.2472,
      "num_input_tokens_seen": 315656,
      "step": 170,
      "train_runtime": 310.5227,
      "train_tokens_per_second": 1016.531
    },
    {
      "epoch": 0.1728582259287339,
      "grad_norm": 0.5661376118659973,
      "learning_rate": 0.00018329113924050634,
      "loss": 0.343,
      "num_input_tokens_seen": 317108,
      "step": 171,
      "train_runtime": 312.097,
      "train_tokens_per_second": 1016.056
    },
    {
      "epoch": 0.17386909274703058,
      "grad_norm": 0.3814907968044281,
      "learning_rate": 0.00018318987341772153,
      "loss": 0.2181,
      "num_input_tokens_seen": 318558,
      "step": 172,
      "train_runtime": 313.6739,
      "train_tokens_per_second": 1015.571
    },
    {
      "epoch": 0.17487995956532726,
      "grad_norm": 0.3652172088623047,
      "learning_rate": 0.00018308860759493672,
      "loss": 0.2136,
      "num_input_tokens_seen": 320342,
      "step": 173,
      "train_runtime": 315.3775,
      "train_tokens_per_second": 1015.741
    },
    {
      "epoch": 0.17589082638362397,
      "grad_norm": 0.3934604823589325,
      "learning_rate": 0.0001829873417721519,
      "loss": 0.2771,
      "num_input_tokens_seen": 322282,
      "step": 174,
      "train_runtime": 317.2426,
      "train_tokens_per_second": 1015.885
    },
    {
      "epoch": 0.17690169320192065,
      "grad_norm": 0.2914677560329437,
      "learning_rate": 0.0001828860759493671,
      "loss": 0.2556,
      "num_input_tokens_seen": 324562,
      "step": 175,
      "train_runtime": 319.2256,
      "train_tokens_per_second": 1016.717
    },
    {
      "epoch": 0.17791256002021733,
      "grad_norm": 0.359348326921463,
      "learning_rate": 0.0001827848101265823,
      "loss": 0.2204,
      "num_input_tokens_seen": 326460,
      "step": 176,
      "train_runtime": 320.9467,
      "train_tokens_per_second": 1017.178
    },
    {
      "epoch": 0.17892342683851403,
      "grad_norm": 0.36025404930114746,
      "learning_rate": 0.00018268354430379748,
      "loss": 0.2518,
      "num_input_tokens_seen": 328256,
      "step": 177,
      "train_runtime": 322.6783,
      "train_tokens_per_second": 1017.285
    },
    {
      "epoch": 0.1799342936568107,
      "grad_norm": 0.42412030696868896,
      "learning_rate": 0.00018258227848101267,
      "loss": 0.2573,
      "num_input_tokens_seen": 330068,
      "step": 178,
      "train_runtime": 324.4329,
      "train_tokens_per_second": 1017.369
    },
    {
      "epoch": 0.1809451604751074,
      "grad_norm": 0.3933882713317871,
      "learning_rate": 0.00018248101265822786,
      "loss": 0.2604,
      "num_input_tokens_seen": 331562,
      "step": 179,
      "train_runtime": 326.0377,
      "train_tokens_per_second": 1016.944
    },
    {
      "epoch": 0.1819560272934041,
      "grad_norm": 0.34556353092193604,
      "learning_rate": 0.00018237974683544305,
      "loss": 0.1406,
      "num_input_tokens_seen": 332972,
      "step": 180,
      "train_runtime": 327.6457,
      "train_tokens_per_second": 1016.256
    },
    {
      "epoch": 0.18296689411170078,
      "grad_norm": 0.3858505189418793,
      "learning_rate": 0.00018227848101265824,
      "loss": 0.4204,
      "num_input_tokens_seen": 334670,
      "step": 181,
      "train_runtime": 331.3835,
      "train_tokens_per_second": 1009.917
    },
    {
      "epoch": 0.18397776092999749,
      "grad_norm": 0.3960411250591278,
      "learning_rate": 0.00018217721518987344,
      "loss": 0.1473,
      "num_input_tokens_seen": 336074,
      "step": 182,
      "train_runtime": 332.9713,
      "train_tokens_per_second": 1009.318
    },
    {
      "epoch": 0.18498862774829417,
      "grad_norm": 0.2916383743286133,
      "learning_rate": 0.00018207594936708863,
      "loss": 0.1958,
      "num_input_tokens_seen": 338036,
      "step": 183,
      "train_runtime": 334.7676,
      "train_tokens_per_second": 1009.763
    },
    {
      "epoch": 0.18599949456659085,
      "grad_norm": 0.3205072581768036,
      "learning_rate": 0.0001819746835443038,
      "loss": 0.2232,
      "num_input_tokens_seen": 339804,
      "step": 184,
      "train_runtime": 336.5239,
      "train_tokens_per_second": 1009.747
    },
    {
      "epoch": 0.18701036138488755,
      "grad_norm": 0.3922429382801056,
      "learning_rate": 0.000181873417721519,
      "loss": 0.1972,
      "num_input_tokens_seen": 341750,
      "step": 185,
      "train_runtime": 338.2584,
      "train_tokens_per_second": 1010.322
    },
    {
      "epoch": 0.18802122820318423,
      "grad_norm": 0.39398542046546936,
      "learning_rate": 0.00018177215189873417,
      "loss": 0.3266,
      "num_input_tokens_seen": 343670,
      "step": 186,
      "train_runtime": 340.0841,
      "train_tokens_per_second": 1010.544
    },
    {
      "epoch": 0.1890320950214809,
      "grad_norm": 0.3316863775253296,
      "learning_rate": 0.00018167088607594936,
      "loss": 0.1562,
      "num_input_tokens_seen": 345674,
      "step": 187,
      "train_runtime": 341.8852,
      "train_tokens_per_second": 1011.082
    },
    {
      "epoch": 0.19004296183977762,
      "grad_norm": 0.4948942959308624,
      "learning_rate": 0.00018156962025316458,
      "loss": 0.2739,
      "num_input_tokens_seen": 347344,
      "step": 188,
      "train_runtime": 343.5796,
      "train_tokens_per_second": 1010.956
    },
    {
      "epoch": 0.1910538286580743,
      "grad_norm": 0.3622047007083893,
      "learning_rate": 0.00018146835443037974,
      "loss": 0.1862,
      "num_input_tokens_seen": 349468,
      "step": 189,
      "train_runtime": 345.3919,
      "train_tokens_per_second": 1011.801
    },
    {
      "epoch": 0.19206469547637098,
      "grad_norm": 0.37027546763420105,
      "learning_rate": 0.00018136708860759493,
      "loss": 0.2266,
      "num_input_tokens_seen": 351544,
      "step": 190,
      "train_runtime": 347.2422,
      "train_tokens_per_second": 1012.389
    },
    {
      "epoch": 0.19307556229466769,
      "grad_norm": 0.3605174422264099,
      "learning_rate": 0.00018126582278481013,
      "loss": 0.3304,
      "num_input_tokens_seen": 353346,
      "step": 191,
      "train_runtime": 348.9445,
      "train_tokens_per_second": 1012.614
    },
    {
      "epoch": 0.19408642911296436,
      "grad_norm": 0.4563489258289337,
      "learning_rate": 0.00018116455696202532,
      "loss": 0.3384,
      "num_input_tokens_seen": 355628,
      "step": 192,
      "train_runtime": 350.9503,
      "train_tokens_per_second": 1013.329
    },
    {
      "epoch": 0.19509729593126104,
      "grad_norm": 0.5172436237335205,
      "learning_rate": 0.0001810632911392405,
      "loss": 0.278,
      "num_input_tokens_seen": 357234,
      "step": 193,
      "train_runtime": 352.5674,
      "train_tokens_per_second": 1013.236
    },
    {
      "epoch": 0.19610816274955775,
      "grad_norm": 0.41404810547828674,
      "learning_rate": 0.0001809620253164557,
      "loss": 0.2162,
      "num_input_tokens_seen": 359288,
      "step": 194,
      "train_runtime": 354.414,
      "train_tokens_per_second": 1013.752
    },
    {
      "epoch": 0.19711902956785443,
      "grad_norm": 0.3381623327732086,
      "learning_rate": 0.0001808607594936709,
      "loss": 0.1617,
      "num_input_tokens_seen": 360876,
      "step": 195,
      "train_runtime": 356.0269,
      "train_tokens_per_second": 1013.62
    },
    {
      "epoch": 0.1981298963861511,
      "grad_norm": 0.2769710421562195,
      "learning_rate": 0.00018075949367088608,
      "loss": 0.2259,
      "num_input_tokens_seen": 362952,
      "step": 196,
      "train_runtime": 357.8201,
      "train_tokens_per_second": 1014.342
    },
    {
      "epoch": 0.19914076320444782,
      "grad_norm": 0.3348582684993744,
      "learning_rate": 0.00018065822784810127,
      "loss": 0.2149,
      "num_input_tokens_seen": 365048,
      "step": 197,
      "train_runtime": 359.7149,
      "train_tokens_per_second": 1014.826
    },
    {
      "epoch": 0.2001516300227445,
      "grad_norm": 0.3476252555847168,
      "learning_rate": 0.00018055696202531646,
      "loss": 0.2743,
      "num_input_tokens_seen": 366994,
      "step": 198,
      "train_runtime": 361.4935,
      "train_tokens_per_second": 1015.216
    },
    {
      "epoch": 0.2001516300227445,
      "eval_loss": 0.2194075733423233,
      "eval_runtime": 61.429,
      "eval_samples_per_second": 14.309,
      "eval_steps_per_second": 7.163,
      "num_input_tokens_seen": 366994,
      "step": 198
    },
    {
      "epoch": 0.2011624968410412,
      "grad_norm": 0.42954811453819275,
      "learning_rate": 0.00018045569620253165,
      "loss": 0.2171,
      "num_input_tokens_seen": 368404,
      "step": 199,
      "train_runtime": 424.4999,
      "train_tokens_per_second": 867.854
    },
    {
      "epoch": 0.20217336365933788,
      "grad_norm": 0.39108145236968994,
      "learning_rate": 0.00018035443037974684,
      "loss": 0.2095,
      "num_input_tokens_seen": 370232,
      "step": 200,
      "train_runtime": 426.1921,
      "train_tokens_per_second": 868.697
    },
    {
      "epoch": 0.20318423047763456,
      "grad_norm": 0.3512146472930908,
      "learning_rate": 0.00018025316455696203,
      "loss": 0.2122,
      "num_input_tokens_seen": 372250,
      "step": 201,
      "train_runtime": 428.0407,
      "train_tokens_per_second": 869.66
    },
    {
      "epoch": 0.20419509729593127,
      "grad_norm": 0.3715672194957733,
      "learning_rate": 0.00018015189873417723,
      "loss": 0.189,
      "num_input_tokens_seen": 374026,
      "step": 202,
      "train_runtime": 429.7707,
      "train_tokens_per_second": 870.292
    },
    {
      "epoch": 0.20520596411422795,
      "grad_norm": 0.45648568868637085,
      "learning_rate": 0.00018005063291139242,
      "loss": 0.2295,
      "num_input_tokens_seen": 375588,
      "step": 203,
      "train_runtime": 431.4083,
      "train_tokens_per_second": 870.609
    },
    {
      "epoch": 0.20621683093252463,
      "grad_norm": 0.4618614912033081,
      "learning_rate": 0.0001799493670886076,
      "loss": 0.2562,
      "num_input_tokens_seen": 377394,
      "step": 204,
      "train_runtime": 433.1983,
      "train_tokens_per_second": 871.181
    },
    {
      "epoch": 0.20722769775082134,
      "grad_norm": 0.4303542673587799,
      "learning_rate": 0.0001798481012658228,
      "loss": 0.1736,
      "num_input_tokens_seen": 379322,
      "step": 205,
      "train_runtime": 434.9163,
      "train_tokens_per_second": 872.172
    },
    {
      "epoch": 0.20823856456911802,
      "grad_norm": 0.48945361375808716,
      "learning_rate": 0.000179746835443038,
      "loss": 0.2474,
      "num_input_tokens_seen": 380978,
      "step": 206,
      "train_runtime": 436.5303,
      "train_tokens_per_second": 872.741
    },
    {
      "epoch": 0.2092494313874147,
      "grad_norm": 0.3990488350391388,
      "learning_rate": 0.00017964556962025315,
      "loss": 0.1922,
      "num_input_tokens_seen": 382954,
      "step": 207,
      "train_runtime": 438.3619,
      "train_tokens_per_second": 873.602
    },
    {
      "epoch": 0.2102602982057114,
      "grad_norm": 0.3929692208766937,
      "learning_rate": 0.00017954430379746837,
      "loss": 0.2035,
      "num_input_tokens_seen": 384998,
      "step": 208,
      "train_runtime": 440.2896,
      "train_tokens_per_second": 874.42
    },
    {
      "epoch": 0.21127116502400808,
      "grad_norm": 0.41196030378341675,
      "learning_rate": 0.00017944303797468356,
      "loss": 0.2038,
      "num_input_tokens_seen": 386652,
      "step": 209,
      "train_runtime": 441.9145,
      "train_tokens_per_second": 874.948
    },
    {
      "epoch": 0.21228203184230476,
      "grad_norm": 0.5543971061706543,
      "learning_rate": 0.00017934177215189872,
      "loss": 0.1938,
      "num_input_tokens_seen": 388026,
      "step": 210,
      "train_runtime": 443.4899,
      "train_tokens_per_second": 874.938
    },
    {
      "epoch": 0.21329289866060147,
      "grad_norm": 0.3725636303424835,
      "learning_rate": 0.00017924050632911394,
      "loss": 0.2182,
      "num_input_tokens_seen": 389782,
      "step": 211,
      "train_runtime": 447.2462,
      "train_tokens_per_second": 871.515
    },
    {
      "epoch": 0.21430376547889815,
      "grad_norm": 0.3938060998916626,
      "learning_rate": 0.00017913924050632913,
      "loss": 0.1772,
      "num_input_tokens_seen": 391886,
      "step": 212,
      "train_runtime": 449.0725,
      "train_tokens_per_second": 872.656
    },
    {
      "epoch": 0.21531463229719486,
      "grad_norm": 0.3155343234539032,
      "learning_rate": 0.0001790379746835443,
      "loss": 0.1668,
      "num_input_tokens_seen": 393538,
      "step": 213,
      "train_runtime": 450.7652,
      "train_tokens_per_second": 873.044
    },
    {
      "epoch": 0.21632549911549154,
      "grad_norm": 0.34781721234321594,
      "learning_rate": 0.00017893670886075952,
      "loss": 0.1614,
      "num_input_tokens_seen": 395470,
      "step": 214,
      "train_runtime": 452.6355,
      "train_tokens_per_second": 873.705
    },
    {
      "epoch": 0.21733636593378822,
      "grad_norm": 0.38830557465553284,
      "learning_rate": 0.00017883544303797468,
      "loss": 0.1959,
      "num_input_tokens_seen": 397572,
      "step": 215,
      "train_runtime": 454.4439,
      "train_tokens_per_second": 874.854
    },
    {
      "epoch": 0.21834723275208492,
      "grad_norm": 0.434148907661438,
      "learning_rate": 0.00017873417721518987,
      "loss": 0.2942,
      "num_input_tokens_seen": 399248,
      "step": 216,
      "train_runtime": 456.0767,
      "train_tokens_per_second": 875.397
    },
    {
      "epoch": 0.2193580995703816,
      "grad_norm": 0.3435661196708679,
      "learning_rate": 0.0001786329113924051,
      "loss": 0.1627,
      "num_input_tokens_seen": 401300,
      "step": 217,
      "train_runtime": 457.9212,
      "train_tokens_per_second": 876.352
    },
    {
      "epoch": 0.22036896638867828,
      "grad_norm": 0.31093019247055054,
      "learning_rate": 0.00017853164556962025,
      "loss": 0.2145,
      "num_input_tokens_seen": 403792,
      "step": 218,
      "train_runtime": 460.0322,
      "train_tokens_per_second": 877.747
    },
    {
      "epoch": 0.221379833206975,
      "grad_norm": 0.4084714651107788,
      "learning_rate": 0.00017843037974683544,
      "loss": 0.3023,
      "num_input_tokens_seen": 405542,
      "step": 219,
      "train_runtime": 461.669,
      "train_tokens_per_second": 878.426
    },
    {
      "epoch": 0.22239070002527167,
      "grad_norm": 0.3440285921096802,
      "learning_rate": 0.00017832911392405066,
      "loss": 0.1894,
      "num_input_tokens_seen": 407284,
      "step": 220,
      "train_runtime": 463.3644,
      "train_tokens_per_second": 878.971
    },
    {
      "epoch": 0.22340156684356835,
      "grad_norm": 0.38901078701019287,
      "learning_rate": 0.00017822784810126582,
      "loss": 0.2994,
      "num_input_tokens_seen": 408822,
      "step": 221,
      "train_runtime": 464.9742,
      "train_tokens_per_second": 879.236
    },
    {
      "epoch": 0.22441243366186506,
      "grad_norm": 0.3624526858329773,
      "learning_rate": 0.00017812658227848101,
      "loss": 0.2181,
      "num_input_tokens_seen": 410890,
      "step": 222,
      "train_runtime": 466.8144,
      "train_tokens_per_second": 880.2
    },
    {
      "epoch": 0.22542330048016174,
      "grad_norm": 0.3111014664173126,
      "learning_rate": 0.0001780253164556962,
      "loss": 0.1725,
      "num_input_tokens_seen": 413154,
      "step": 223,
      "train_runtime": 468.769,
      "train_tokens_per_second": 881.36
    },
    {
      "epoch": 0.22643416729845842,
      "grad_norm": 0.41633716225624084,
      "learning_rate": 0.0001779240506329114,
      "loss": 0.1863,
      "num_input_tokens_seen": 414952,
      "step": 224,
      "train_runtime": 470.4852,
      "train_tokens_per_second": 881.966
    },
    {
      "epoch": 0.22744503411675512,
      "grad_norm": 0.326198011636734,
      "learning_rate": 0.0001778227848101266,
      "loss": 0.2084,
      "num_input_tokens_seen": 417118,
      "step": 225,
      "train_runtime": 472.4207,
      "train_tokens_per_second": 882.938
    },
    {
      "epoch": 0.2284559009350518,
      "grad_norm": 0.4119671881198883,
      "learning_rate": 0.00017772151898734178,
      "loss": 0.1664,
      "num_input_tokens_seen": 418874,
      "step": 226,
      "train_runtime": 474.1206,
      "train_tokens_per_second": 883.476
    },
    {
      "epoch": 0.22946676775334848,
      "grad_norm": 0.4336074888706207,
      "learning_rate": 0.00017762025316455697,
      "loss": 0.2245,
      "num_input_tokens_seen": 420976,
      "step": 227,
      "train_runtime": 475.9922,
      "train_tokens_per_second": 884.418
    },
    {
      "epoch": 0.2304776345716452,
      "grad_norm": 0.3529372215270996,
      "learning_rate": 0.00017751898734177216,
      "loss": 0.3078,
      "num_input_tokens_seen": 422772,
      "step": 228,
      "train_runtime": 477.7538,
      "train_tokens_per_second": 884.916
    },
    {
      "epoch": 0.23148850138994187,
      "grad_norm": 0.36478540301322937,
      "learning_rate": 0.00017741772151898735,
      "loss": 0.236,
      "num_input_tokens_seen": 424808,
      "step": 229,
      "train_runtime": 479.5427,
      "train_tokens_per_second": 885.861
    },
    {
      "epoch": 0.23249936820823858,
      "grad_norm": 0.36925044655799866,
      "learning_rate": 0.00017731645569620254,
      "loss": 0.3449,
      "num_input_tokens_seen": 426316,
      "step": 230,
      "train_runtime": 481.1515,
      "train_tokens_per_second": 886.033
    },
    {
      "epoch": 0.23351023502653526,
      "grad_norm": 0.35536423325538635,
      "learning_rate": 0.00017721518987341773,
      "loss": 0.2149,
      "num_input_tokens_seen": 428032,
      "step": 231,
      "train_runtime": 482.8356,
      "train_tokens_per_second": 886.496
    },
    {
      "epoch": 0.23452110184483194,
      "grad_norm": 0.3737644553184509,
      "learning_rate": 0.00017711392405063292,
      "loss": 0.2188,
      "num_input_tokens_seen": 430370,
      "step": 232,
      "train_runtime": 484.8382,
      "train_tokens_per_second": 887.657
    },
    {
      "epoch": 0.23553196866312864,
      "grad_norm": 0.30169618129730225,
      "learning_rate": 0.00017701265822784811,
      "loss": 0.1643,
      "num_input_tokens_seen": 432208,
      "step": 233,
      "train_runtime": 486.5744,
      "train_tokens_per_second": 888.267
    },
    {
      "epoch": 0.23654283548142532,
      "grad_norm": 0.42626598477363586,
      "learning_rate": 0.0001769113924050633,
      "loss": 0.2891,
      "num_input_tokens_seen": 434056,
      "step": 234,
      "train_runtime": 488.3877,
      "train_tokens_per_second": 888.753
    },
    {
      "epoch": 0.237553702299722,
      "grad_norm": 0.4076259136199951,
      "learning_rate": 0.0001768101265822785,
      "loss": 0.2513,
      "num_input_tokens_seen": 436298,
      "step": 235,
      "train_runtime": 490.3713,
      "train_tokens_per_second": 889.73
    },
    {
      "epoch": 0.2385645691180187,
      "grad_norm": 0.39652830362319946,
      "learning_rate": 0.0001767088607594937,
      "loss": 0.3282,
      "num_input_tokens_seen": 438264,
      "step": 236,
      "train_runtime": 492.1979,
      "train_tokens_per_second": 890.422
    },
    {
      "epoch": 0.2395754359363154,
      "grad_norm": 0.43892091512680054,
      "learning_rate": 0.00017660759493670888,
      "loss": 0.2807,
      "num_input_tokens_seen": 440096,
      "step": 237,
      "train_runtime": 493.8945,
      "train_tokens_per_second": 891.073
    },
    {
      "epoch": 0.24058630275461207,
      "grad_norm": 0.45497018098831177,
      "learning_rate": 0.00017650632911392407,
      "loss": 0.1972,
      "num_input_tokens_seen": 441640,
      "step": 238,
      "train_runtime": 495.5074,
      "train_tokens_per_second": 891.288
    },
    {
      "epoch": 0.24159716957290878,
      "grad_norm": 0.441567599773407,
      "learning_rate": 0.00017640506329113923,
      "loss": 0.2892,
      "num_input_tokens_seen": 443558,
      "step": 239,
      "train_runtime": 497.226,
      "train_tokens_per_second": 892.065
    },
    {
      "epoch": 0.24260803639120546,
      "grad_norm": 0.5874553918838501,
      "learning_rate": 0.00017630379746835445,
      "loss": 0.1985,
      "num_input_tokens_seen": 444900,
      "step": 240,
      "train_runtime": 498.8049,
      "train_tokens_per_second": 891.932
    },
    {
      "epoch": 0.24361890320950214,
      "grad_norm": 0.46553996205329895,
      "learning_rate": 0.00017620253164556964,
      "loss": 0.2307,
      "num_input_tokens_seen": 446356,
      "step": 241,
      "train_runtime": 502.4771,
      "train_tokens_per_second": 888.311
    },
    {
      "epoch": 0.24462977002779884,
      "grad_norm": 0.3993649482727051,
      "learning_rate": 0.0001761012658227848,
      "loss": 0.1905,
      "num_input_tokens_seen": 447858,
      "step": 242,
      "train_runtime": 504.1376,
      "train_tokens_per_second": 888.365
    },
    {
      "epoch": 0.24564063684609552,
      "grad_norm": 0.3534109592437744,
      "learning_rate": 0.00017600000000000002,
      "loss": 0.1715,
      "num_input_tokens_seen": 449478,
      "step": 243,
      "train_runtime": 505.7563,
      "train_tokens_per_second": 888.725
    },
    {
      "epoch": 0.24665150366439223,
      "grad_norm": 0.40606966614723206,
      "learning_rate": 0.0001758987341772152,
      "loss": 0.2179,
      "num_input_tokens_seen": 451442,
      "step": 244,
      "train_runtime": 507.6019,
      "train_tokens_per_second": 889.362
    },
    {
      "epoch": 0.2476623704826889,
      "grad_norm": 0.3612362742424011,
      "learning_rate": 0.00017579746835443038,
      "loss": 0.1328,
      "num_input_tokens_seen": 452948,
      "step": 245,
      "train_runtime": 509.2145,
      "train_tokens_per_second": 889.503
    },
    {
      "epoch": 0.2486732373009856,
      "grad_norm": 0.38469746708869934,
      "learning_rate": 0.0001756962025316456,
      "loss": 0.187,
      "num_input_tokens_seen": 454934,
      "step": 246,
      "train_runtime": 511.0563,
      "train_tokens_per_second": 890.184
    },
    {
      "epoch": 0.2496841041192823,
      "grad_norm": 0.4440872073173523,
      "learning_rate": 0.00017559493670886076,
      "loss": 0.3736,
      "num_input_tokens_seen": 456716,
      "step": 247,
      "train_runtime": 512.7952,
      "train_tokens_per_second": 890.64
    },
    {
      "epoch": 0.250694970937579,
      "grad_norm": 0.43680766224861145,
      "learning_rate": 0.00017549367088607595,
      "loss": 0.161,
      "num_input_tokens_seen": 458266,
      "step": 248,
      "train_runtime": 514.4249,
      "train_tokens_per_second": 890.832
    },
    {
      "epoch": 0.2517058377558757,
      "grad_norm": 0.3658226430416107,
      "learning_rate": 0.00017539240506329117,
      "loss": 0.2598,
      "num_input_tokens_seen": 459798,
      "step": 249,
      "train_runtime": 516.1364,
      "train_tokens_per_second": 890.846
    },
    {
      "epoch": 0.25271670457417234,
      "grad_norm": 0.31666117906570435,
      "learning_rate": 0.00017529113924050633,
      "loss": 0.1772,
      "num_input_tokens_seen": 461594,
      "step": 250,
      "train_runtime": 517.9112,
      "train_tokens_per_second": 891.261
    },
    {
      "epoch": 0.25372757139246904,
      "grad_norm": 0.31696584820747375,
      "learning_rate": 0.00017518987341772152,
      "loss": 0.3119,
      "num_input_tokens_seen": 463962,
      "step": 251,
      "train_runtime": 519.9303,
      "train_tokens_per_second": 892.354
    },
    {
      "epoch": 0.25473843821076575,
      "grad_norm": 0.36062178015708923,
      "learning_rate": 0.00017508860759493674,
      "loss": 0.1571,
      "num_input_tokens_seen": 465908,
      "step": 252,
      "train_runtime": 521.7069,
      "train_tokens_per_second": 893.046
    },
    {
      "epoch": 0.2557493050290624,
      "grad_norm": 0.41598281264305115,
      "learning_rate": 0.0001749873417721519,
      "loss": 0.2414,
      "num_input_tokens_seen": 467770,
      "step": 253,
      "train_runtime": 523.471,
      "train_tokens_per_second": 893.593
    },
    {
      "epoch": 0.2567601718473591,
      "grad_norm": 0.361608624458313,
      "learning_rate": 0.0001748860759493671,
      "loss": 0.201,
      "num_input_tokens_seen": 469856,
      "step": 254,
      "train_runtime": 525.3382,
      "train_tokens_per_second": 894.388
    },
    {
      "epoch": 0.2577710386656558,
      "grad_norm": 0.28574812412261963,
      "learning_rate": 0.00017478481012658229,
      "loss": 0.2026,
      "num_input_tokens_seen": 472246,
      "step": 255,
      "train_runtime": 527.4042,
      "train_tokens_per_second": 895.416
    },
    {
      "epoch": 0.25878190548395247,
      "grad_norm": 0.39967218041419983,
      "learning_rate": 0.00017468354430379748,
      "loss": 0.1907,
      "num_input_tokens_seen": 473990,
      "step": 256,
      "train_runtime": 529.0961,
      "train_tokens_per_second": 895.849
    },
    {
      "epoch": 0.2597927723022492,
      "grad_norm": 0.3359277546405792,
      "learning_rate": 0.00017458227848101267,
      "loss": 0.279,
      "num_input_tokens_seen": 476312,
      "step": 257,
      "train_runtime": 531.0411,
      "train_tokens_per_second": 896.94
    },
    {
      "epoch": 0.2608036391205459,
      "grad_norm": 0.3367970287799835,
      "learning_rate": 0.00017448101265822786,
      "loss": 0.2644,
      "num_input_tokens_seen": 478152,
      "step": 258,
      "train_runtime": 532.6894,
      "train_tokens_per_second": 897.619
    },
    {
      "epoch": 0.26181450593884253,
      "grad_norm": 0.39386895298957825,
      "learning_rate": 0.00017437974683544305,
      "loss": 0.1985,
      "num_input_tokens_seen": 480216,
      "step": 259,
      "train_runtime": 534.5423,
      "train_tokens_per_second": 898.369
    },
    {
      "epoch": 0.26282537275713924,
      "grad_norm": 0.33743321895599365,
      "learning_rate": 0.00017427848101265824,
      "loss": 0.2987,
      "num_input_tokens_seen": 482288,
      "step": 260,
      "train_runtime": 536.3894,
      "train_tokens_per_second": 899.138
    },
    {
      "epoch": 0.26383623957543595,
      "grad_norm": 0.49012792110443115,
      "learning_rate": 0.00017417721518987343,
      "loss": 0.2798,
      "num_input_tokens_seen": 483850,
      "step": 261,
      "train_runtime": 538.0804,
      "train_tokens_per_second": 899.215
    },
    {
      "epoch": 0.2648471063937326,
      "grad_norm": 0.41628557443618774,
      "learning_rate": 0.00017407594936708862,
      "loss": 0.2674,
      "num_input_tokens_seen": 485646,
      "step": 262,
      "train_runtime": 539.8519,
      "train_tokens_per_second": 899.591
    },
    {
      "epoch": 0.2658579732120293,
      "grad_norm": 0.49774986505508423,
      "learning_rate": 0.00017397468354430379,
      "loss": 0.27,
      "num_input_tokens_seen": 487634,
      "step": 263,
      "train_runtime": 541.6853,
      "train_tokens_per_second": 900.216
    },
    {
      "epoch": 0.266868840030326,
      "grad_norm": 0.35518673062324524,
      "learning_rate": 0.000173873417721519,
      "loss": 0.1975,
      "num_input_tokens_seen": 489650,
      "step": 264,
      "train_runtime": 543.461,
      "train_tokens_per_second": 900.985
    },
    {
      "epoch": 0.26787970684862267,
      "grad_norm": 0.43562984466552734,
      "learning_rate": 0.0001737721518987342,
      "loss": 0.2114,
      "num_input_tokens_seen": 491834,
      "step": 265,
      "train_runtime": 545.341,
      "train_tokens_per_second": 901.883
    },
    {
      "epoch": 0.2688905736669194,
      "grad_norm": 0.3897487223148346,
      "learning_rate": 0.00017367088607594936,
      "loss": 0.2255,
      "num_input_tokens_seen": 494118,
      "step": 266,
      "train_runtime": 547.3362,
      "train_tokens_per_second": 902.769
    },
    {
      "epoch": 0.2699014404852161,
      "grad_norm": 0.41832661628723145,
      "learning_rate": 0.00017356962025316458,
      "loss": 0.235,
      "num_input_tokens_seen": 495884,
      "step": 267,
      "train_runtime": 549.1008,
      "train_tokens_per_second": 903.084
    },
    {
      "epoch": 0.2709123073035128,
      "grad_norm": 0.4256960153579712,
      "learning_rate": 0.00017346835443037977,
      "loss": 0.2551,
      "num_input_tokens_seen": 497916,
      "step": 268,
      "train_runtime": 550.8381,
      "train_tokens_per_second": 903.924
    },
    {
      "epoch": 0.27192317412180944,
      "grad_norm": 0.43746358156204224,
      "learning_rate": 0.00017336708860759493,
      "loss": 0.1557,
      "num_input_tokens_seen": 499520,
      "step": 269,
      "train_runtime": 552.5099,
      "train_tokens_per_second": 904.092
    },
    {
      "epoch": 0.27293404094010615,
      "grad_norm": 0.43507421016693115,
      "learning_rate": 0.00017326582278481015,
      "loss": 0.2928,
      "num_input_tokens_seen": 501704,
      "step": 270,
      "train_runtime": 554.4475,
      "train_tokens_per_second": 904.872
    },
    {
      "epoch": 0.27394490775840286,
      "grad_norm": 0.43564385175704956,
      "learning_rate": 0.0001731645569620253,
      "loss": 0.2278,
      "num_input_tokens_seen": 503400,
      "step": 271,
      "train_runtime": 557.8461,
      "train_tokens_per_second": 902.399
    },
    {
      "epoch": 0.2749557745766995,
      "grad_norm": 0.47292283177375793,
      "learning_rate": 0.0001730632911392405,
      "loss": 0.2152,
      "num_input_tokens_seen": 505150,
      "step": 272,
      "train_runtime": 559.6529,
      "train_tokens_per_second": 902.613
    },
    {
      "epoch": 0.2759666413949962,
      "grad_norm": 0.420528769493103,
      "learning_rate": 0.00017296202531645572,
      "loss": 0.1803,
      "num_input_tokens_seen": 506914,
      "step": 273,
      "train_runtime": 561.4075,
      "train_tokens_per_second": 902.934
    },
    {
      "epoch": 0.2769775082132929,
      "grad_norm": 0.3233083188533783,
      "learning_rate": 0.00017286075949367088,
      "loss": 0.2905,
      "num_input_tokens_seen": 509386,
      "step": 274,
      "train_runtime": 563.4828,
      "train_tokens_per_second": 903.996
    },
    {
      "epoch": 0.2779883750315896,
      "grad_norm": 0.3598622977733612,
      "learning_rate": 0.00017275949367088608,
      "loss": 0.2707,
      "num_input_tokens_seen": 510774,
      "step": 275,
      "train_runtime": 565.0835,
      "train_tokens_per_second": 903.891
    },
    {
      "epoch": 0.2789992418498863,
      "grad_norm": 0.4031928479671478,
      "learning_rate": 0.0001726582278481013,
      "loss": 0.1487,
      "num_input_tokens_seen": 512408,
      "step": 276,
      "train_runtime": 566.7789,
      "train_tokens_per_second": 904.07
    },
    {
      "epoch": 0.280010108668183,
      "grad_norm": 0.35309258103370667,
      "learning_rate": 0.00017255696202531646,
      "loss": 0.2066,
      "num_input_tokens_seen": 514444,
      "step": 277,
      "train_runtime": 568.6336,
      "train_tokens_per_second": 904.702
    },
    {
      "epoch": 0.28102097548647964,
      "grad_norm": 0.3811109960079193,
      "learning_rate": 0.00017245569620253165,
      "loss": 0.2566,
      "num_input_tokens_seen": 516154,
      "step": 278,
      "train_runtime": 570.3865,
      "train_tokens_per_second": 904.92
    },
    {
      "epoch": 0.28203184230477635,
      "grad_norm": 0.39953625202178955,
      "learning_rate": 0.00017235443037974684,
      "loss": 0.2417,
      "num_input_tokens_seen": 518388,
      "step": 279,
      "train_runtime": 572.267,
      "train_tokens_per_second": 905.85
    },
    {
      "epoch": 0.28304270912307306,
      "grad_norm": 0.4009864926338196,
      "learning_rate": 0.00017225316455696203,
      "loss": 0.2438,
      "num_input_tokens_seen": 520276,
      "step": 280,
      "train_runtime": 573.9768,
      "train_tokens_per_second": 906.441
    },
    {
      "epoch": 0.2840535759413697,
      "grad_norm": 0.45036402344703674,
      "learning_rate": 0.00017215189873417722,
      "loss": 0.1924,
      "num_input_tokens_seen": 522032,
      "step": 281,
      "train_runtime": 575.7105,
      "train_tokens_per_second": 906.761
    },
    {
      "epoch": 0.2850644427596664,
      "grad_norm": 0.48415297269821167,
      "learning_rate": 0.0001720506329113924,
      "loss": 0.1766,
      "num_input_tokens_seen": 523640,
      "step": 282,
      "train_runtime": 577.335,
      "train_tokens_per_second": 906.995
    },
    {
      "epoch": 0.2860753095779631,
      "grad_norm": 0.4514244794845581,
      "learning_rate": 0.0001719493670886076,
      "loss": 0.1805,
      "num_input_tokens_seen": 525432,
      "step": 283,
      "train_runtime": 578.9738,
      "train_tokens_per_second": 907.523
    },
    {
      "epoch": 0.2870861763962598,
      "grad_norm": 0.4642978310585022,
      "learning_rate": 0.0001718481012658228,
      "loss": 0.1902,
      "num_input_tokens_seen": 527470,
      "step": 284,
      "train_runtime": 580.7784,
      "train_tokens_per_second": 908.212
    },
    {
      "epoch": 0.2880970432145565,
      "grad_norm": 0.4242786169052124,
      "learning_rate": 0.00017174683544303798,
      "loss": 0.1567,
      "num_input_tokens_seen": 529352,
      "step": 285,
      "train_runtime": 582.5114,
      "train_tokens_per_second": 908.741
    },
    {
      "epoch": 0.2891079100328532,
      "grad_norm": 0.46856752038002014,
      "learning_rate": 0.00017164556962025317,
      "loss": 0.249,
      "num_input_tokens_seen": 531024,
      "step": 286,
      "train_runtime": 584.1487,
      "train_tokens_per_second": 909.056
    },
    {
      "epoch": 0.29011877685114984,
      "grad_norm": 0.4047328531742096,
      "learning_rate": 0.00017154430379746837,
      "loss": 0.228,
      "num_input_tokens_seen": 533078,
      "step": 287,
      "train_runtime": 585.9552,
      "train_tokens_per_second": 909.759
    },
    {
      "epoch": 0.29112964366944655,
      "grad_norm": 0.3408379852771759,
      "learning_rate": 0.00017144303797468356,
      "loss": 0.1971,
      "num_input_tokens_seen": 535470,
      "step": 288,
      "train_runtime": 587.9542,
      "train_tokens_per_second": 910.734
    },
    {
      "epoch": 0.29214051048774325,
      "grad_norm": 0.489050954580307,
      "learning_rate": 0.00017134177215189875,
      "loss": 0.1943,
      "num_input_tokens_seen": 537330,
      "step": 289,
      "train_runtime": 589.7218,
      "train_tokens_per_second": 911.158
    },
    {
      "epoch": 0.2931513773060399,
      "grad_norm": 0.400348037481308,
      "learning_rate": 0.00017124050632911394,
      "loss": 0.2157,
      "num_input_tokens_seen": 539286,
      "step": 290,
      "train_runtime": 591.4963,
      "train_tokens_per_second": 911.732
    },
    {
      "epoch": 0.2941622441243366,
      "grad_norm": 0.3712882399559021,
      "learning_rate": 0.00017113924050632913,
      "loss": 0.168,
      "num_input_tokens_seen": 541432,
      "step": 291,
      "train_runtime": 593.4921,
      "train_tokens_per_second": 912.282
    },
    {
      "epoch": 0.2951731109426333,
      "grad_norm": 0.32833173871040344,
      "learning_rate": 0.00017103797468354432,
      "loss": 0.1658,
      "num_input_tokens_seen": 543382,
      "step": 292,
      "train_runtime": 595.3108,
      "train_tokens_per_second": 912.77
    },
    {
      "epoch": 0.29618397776093,
      "grad_norm": 0.36027055978775024,
      "learning_rate": 0.0001709367088607595,
      "loss": 0.1858,
      "num_input_tokens_seen": 545264,
      "step": 293,
      "train_runtime": 597.0765,
      "train_tokens_per_second": 913.223
    },
    {
      "epoch": 0.2971948445792267,
      "grad_norm": 0.27264341711997986,
      "learning_rate": 0.0001708354430379747,
      "loss": 0.195,
      "num_input_tokens_seen": 547576,
      "step": 294,
      "train_runtime": 599.0353,
      "train_tokens_per_second": 914.096
    },
    {
      "epoch": 0.2982057113975234,
      "grad_norm": 0.46067139506340027,
      "learning_rate": 0.00017073417721518987,
      "loss": 0.2759,
      "num_input_tokens_seen": 549512,
      "step": 295,
      "train_runtime": 600.8085,
      "train_tokens_per_second": 914.621
    },
    {
      "epoch": 0.29921657821582004,
      "grad_norm": 0.3848015367984772,
      "learning_rate": 0.00017063291139240508,
      "loss": 0.1802,
      "num_input_tokens_seen": 550774,
      "step": 296,
      "train_runtime": 602.3542,
      "train_tokens_per_second": 914.369
    },
    {
      "epoch": 0.30022744503411675,
      "grad_norm": 0.36973848938941956,
      "learning_rate": 0.00017053164556962027,
      "loss": 0.181,
      "num_input_tokens_seen": 552806,
      "step": 297,
      "train_runtime": 604.2172,
      "train_tokens_per_second": 914.913
    },
    {
      "epoch": 0.30123831185241345,
      "grad_norm": 0.2873566150665283,
      "learning_rate": 0.00017043037974683544,
      "loss": 0.1467,
      "num_input_tokens_seen": 555204,
      "step": 298,
      "train_runtime": 606.2635,
      "train_tokens_per_second": 915.78
    },
    {
      "epoch": 0.30224917867071016,
      "grad_norm": 0.37974682450294495,
      "learning_rate": 0.00017032911392405066,
      "loss": 0.1634,
      "num_input_tokens_seen": 556868,
      "step": 299,
      "train_runtime": 607.9181,
      "train_tokens_per_second": 916.025
    },
    {
      "epoch": 0.3032600454890068,
      "grad_norm": 0.3785713016986847,
      "learning_rate": 0.00017022784810126585,
      "loss": 0.1746,
      "num_input_tokens_seen": 558456,
      "step": 300,
      "train_runtime": 609.5133,
      "train_tokens_per_second": 916.233
    },
    {
      "epoch": 0.3042709123073035,
      "grad_norm": 0.4307490289211273,
      "learning_rate": 0.000170126582278481,
      "loss": 0.2887,
      "num_input_tokens_seen": 560334,
      "step": 301,
      "train_runtime": 612.9674,
      "train_tokens_per_second": 914.133
    },
    {
      "epoch": 0.30528177912560023,
      "grad_norm": 0.36402326822280884,
      "learning_rate": 0.00017002531645569623,
      "loss": 0.1913,
      "num_input_tokens_seen": 562316,
      "step": 302,
      "train_runtime": 614.8226,
      "train_tokens_per_second": 914.599
    },
    {
      "epoch": 0.3062926459438969,
      "grad_norm": 0.33439892530441284,
      "learning_rate": 0.0001699240506329114,
      "loss": 0.1823,
      "num_input_tokens_seen": 564448,
      "step": 303,
      "train_runtime": 616.6247,
      "train_tokens_per_second": 915.383
    },
    {
      "epoch": 0.3073035127621936,
      "grad_norm": 0.5432685017585754,
      "learning_rate": 0.00016982278481012658,
      "loss": 0.1848,
      "num_input_tokens_seen": 565648,
      "step": 304,
      "train_runtime": 618.1611,
      "train_tokens_per_second": 915.049
    },
    {
      "epoch": 0.3083143795804903,
      "grad_norm": 0.5558014512062073,
      "learning_rate": 0.0001697215189873418,
      "loss": 0.2158,
      "num_input_tokens_seen": 567142,
      "step": 305,
      "train_runtime": 619.8092,
      "train_tokens_per_second": 915.027
    },
    {
      "epoch": 0.30932524639878695,
      "grad_norm": 0.43292495608329773,
      "learning_rate": 0.00016962025316455696,
      "loss": 0.3157,
      "num_input_tokens_seen": 569422,
      "step": 306,
      "train_runtime": 621.7428,
      "train_tokens_per_second": 915.848
    },
    {
      "epoch": 0.31033611321708365,
      "grad_norm": 0.4133246839046478,
      "learning_rate": 0.00016951898734177216,
      "loss": 0.1982,
      "num_input_tokens_seen": 571246,
      "step": 307,
      "train_runtime": 623.4886,
      "train_tokens_per_second": 916.209
    },
    {
      "epoch": 0.31134698003538036,
      "grad_norm": 0.36801382899284363,
      "learning_rate": 0.00016941772151898737,
      "loss": 0.272,
      "num_input_tokens_seen": 573048,
      "step": 308,
      "train_runtime": 625.1869,
      "train_tokens_per_second": 916.603
    },
    {
      "epoch": 0.312357846853677,
      "grad_norm": 0.41676491498947144,
      "learning_rate": 0.00016931645569620254,
      "loss": 0.2582,
      "num_input_tokens_seen": 575496,
      "step": 309,
      "train_runtime": 627.2635,
      "train_tokens_per_second": 917.471
    },
    {
      "epoch": 0.3133687136719737,
      "grad_norm": 0.44287529587745667,
      "learning_rate": 0.00016921518987341773,
      "loss": 0.2554,
      "num_input_tokens_seen": 577466,
      "step": 310,
      "train_runtime": 629.0362,
      "train_tokens_per_second": 918.017
    },
    {
      "epoch": 0.3143795804902704,
      "grad_norm": 0.5288241505622864,
      "learning_rate": 0.00016911392405063292,
      "loss": 0.2612,
      "num_input_tokens_seen": 579274,
      "step": 311,
      "train_runtime": 630.6707,
      "train_tokens_per_second": 918.505
    },
    {
      "epoch": 0.3153904473085671,
      "grad_norm": 0.36141201853752136,
      "learning_rate": 0.0001690126582278481,
      "loss": 0.2379,
      "num_input_tokens_seen": 580918,
      "step": 312,
      "train_runtime": 632.2878,
      "train_tokens_per_second": 918.756
    },
    {
      "epoch": 0.3164013141268638,
      "grad_norm": 0.34109199047088623,
      "learning_rate": 0.0001689113924050633,
      "loss": 0.2135,
      "num_input_tokens_seen": 583062,
      "step": 313,
      "train_runtime": 634.2596,
      "train_tokens_per_second": 919.28
    },
    {
      "epoch": 0.3174121809451605,
      "grad_norm": 0.40509626269340515,
      "learning_rate": 0.0001688101265822785,
      "loss": 0.1945,
      "num_input_tokens_seen": 585200,
      "step": 314,
      "train_runtime": 636.1207,
      "train_tokens_per_second": 919.951
    },
    {
      "epoch": 0.31842304776345715,
      "grad_norm": 0.38416141271591187,
      "learning_rate": 0.00016870886075949368,
      "loss": 0.1894,
      "num_input_tokens_seen": 586980,
      "step": 315,
      "train_runtime": 637.755,
      "train_tokens_per_second": 920.385
    },
    {
      "epoch": 0.31943391458175385,
      "grad_norm": 0.4075499176979065,
      "learning_rate": 0.00016860759493670887,
      "loss": 0.2793,
      "num_input_tokens_seen": 589100,
      "step": 316,
      "train_runtime": 639.6146,
      "train_tokens_per_second": 921.023
    },
    {
      "epoch": 0.32044478140005056,
      "grad_norm": 0.3548603355884552,
      "learning_rate": 0.00016850632911392406,
      "loss": 0.2055,
      "num_input_tokens_seen": 591336,
      "step": 317,
      "train_runtime": 641.5444,
      "train_tokens_per_second": 921.738
    },
    {
      "epoch": 0.3214556482183472,
      "grad_norm": 0.4027518332004547,
      "learning_rate": 0.00016840506329113925,
      "loss": 0.2338,
      "num_input_tokens_seen": 593528,
      "step": 318,
      "train_runtime": 643.4589,
      "train_tokens_per_second": 922.402
    },
    {
      "epoch": 0.3224665150366439,
      "grad_norm": 0.4063768684864044,
      "learning_rate": 0.00016830379746835442,
      "loss": 0.1914,
      "num_input_tokens_seen": 595364,
      "step": 319,
      "train_runtime": 645.3109,
      "train_tokens_per_second": 922.6
    },
    {
      "epoch": 0.3234773818549406,
      "grad_norm": 0.41555434465408325,
      "learning_rate": 0.00016820253164556964,
      "loss": 0.2672,
      "num_input_tokens_seen": 597190,
      "step": 320,
      "train_runtime": 647.035,
      "train_tokens_per_second": 922.964
    },
    {
      "epoch": 0.3244882486732373,
      "grad_norm": 0.4873950183391571,
      "learning_rate": 0.00016810126582278483,
      "loss": 0.1835,
      "num_input_tokens_seen": 598648,
      "step": 321,
      "train_runtime": 648.6144,
      "train_tokens_per_second": 922.964
    },
    {
      "epoch": 0.325499115491534,
      "grad_norm": 0.3299708664417267,
      "learning_rate": 0.000168,
      "loss": 0.218,
      "num_input_tokens_seen": 600346,
      "step": 322,
      "train_runtime": 650.307,
      "train_tokens_per_second": 923.173
    },
    {
      "epoch": 0.3265099823098307,
      "grad_norm": 0.4436740577220917,
      "learning_rate": 0.0001678987341772152,
      "loss": 0.2922,
      "num_input_tokens_seen": 601936,
      "step": 323,
      "train_runtime": 651.9296,
      "train_tokens_per_second": 923.314
    },
    {
      "epoch": 0.32752084912812734,
      "grad_norm": 0.39118847250938416,
      "learning_rate": 0.0001677974683544304,
      "loss": 0.2583,
      "num_input_tokens_seen": 604046,
      "step": 324,
      "train_runtime": 653.7406,
      "train_tokens_per_second": 923.984
    },
    {
      "epoch": 0.32853171594642405,
      "grad_norm": 0.34720638394355774,
      "learning_rate": 0.00016769620253164556,
      "loss": 0.1586,
      "num_input_tokens_seen": 605926,
      "step": 325,
      "train_runtime": 655.518,
      "train_tokens_per_second": 924.347
    },
    {
      "epoch": 0.32954258276472076,
      "grad_norm": 0.45301494002342224,
      "learning_rate": 0.00016759493670886078,
      "loss": 0.2335,
      "num_input_tokens_seen": 607350,
      "step": 326,
      "train_runtime": 657.1179,
      "train_tokens_per_second": 924.263
    },
    {
      "epoch": 0.3305534495830174,
      "grad_norm": 0.5036170482635498,
      "learning_rate": 0.00016749367088607594,
      "loss": 0.2673,
      "num_input_tokens_seen": 608644,
      "step": 327,
      "train_runtime": 658.6868,
      "train_tokens_per_second": 924.026
    },
    {
      "epoch": 0.3315643164013141,
      "grad_norm": 0.5227355360984802,
      "learning_rate": 0.00016739240506329114,
      "loss": 0.1654,
      "num_input_tokens_seen": 609744,
      "step": 328,
      "train_runtime": 660.2289,
      "train_tokens_per_second": 923.534
    },
    {
      "epoch": 0.3325751832196108,
      "grad_norm": 0.3550918400287628,
      "learning_rate": 0.00016729113924050635,
      "loss": 0.1822,
      "num_input_tokens_seen": 611858,
      "step": 329,
      "train_runtime": 662.079,
      "train_tokens_per_second": 924.146
    },
    {
      "epoch": 0.33358605003790753,
      "grad_norm": 0.46741363406181335,
      "learning_rate": 0.00016718987341772152,
      "loss": 0.2668,
      "num_input_tokens_seen": 613746,
      "step": 330,
      "train_runtime": 663.838,
      "train_tokens_per_second": 924.542
    },
    {
      "epoch": 0.3345969168562042,
      "grad_norm": 0.42439013719558716,
      "learning_rate": 0.0001670886075949367,
      "loss": 0.2303,
      "num_input_tokens_seen": 615626,
      "step": 331,
      "train_runtime": 667.3259,
      "train_tokens_per_second": 922.527
    },
    {
      "epoch": 0.3356077836745009,
      "grad_norm": 0.3509574830532074,
      "learning_rate": 0.00016698734177215193,
      "loss": 0.2095,
      "num_input_tokens_seen": 618124,
      "step": 332,
      "train_runtime": 669.3305,
      "train_tokens_per_second": 923.496
    },
    {
      "epoch": 0.3366186504927976,
      "grad_norm": 0.39720645546913147,
      "learning_rate": 0.0001668860759493671,
      "loss": 0.1346,
      "num_input_tokens_seen": 619796,
      "step": 333,
      "train_runtime": 671.0706,
      "train_tokens_per_second": 923.593
    },
    {
      "epoch": 0.33762951731109425,
      "grad_norm": 0.36094117164611816,
      "learning_rate": 0.00016678481012658228,
      "loss": 0.1887,
      "num_input_tokens_seen": 621398,
      "step": 334,
      "train_runtime": 672.701,
      "train_tokens_per_second": 923.736
    },
    {
      "epoch": 0.33864038412939096,
      "grad_norm": 0.29542970657348633,
      "learning_rate": 0.00016668354430379747,
      "loss": 0.2556,
      "num_input_tokens_seen": 623336,
      "step": 335,
      "train_runtime": 674.4235,
      "train_tokens_per_second": 924.25
    },
    {
      "epoch": 0.33965125094768767,
      "grad_norm": 0.4342578649520874,
      "learning_rate": 0.00016658227848101266,
      "loss": 0.184,
      "num_input_tokens_seen": 625088,
      "step": 336,
      "train_runtime": 676.1745,
      "train_tokens_per_second": 924.448
    },
    {
      "epoch": 0.3406621177659843,
      "grad_norm": 0.3464232087135315,
      "learning_rate": 0.00016648101265822785,
      "loss": 0.2506,
      "num_input_tokens_seen": 627178,
      "step": 337,
      "train_runtime": 678.1014,
      "train_tokens_per_second": 924.903
    },
    {
      "epoch": 0.341672984584281,
      "grad_norm": 0.4655736982822418,
      "learning_rate": 0.00016637974683544304,
      "loss": 0.2721,
      "num_input_tokens_seen": 629028,
      "step": 338,
      "train_runtime": 679.8682,
      "train_tokens_per_second": 925.221
    },
    {
      "epoch": 0.34268385140257773,
      "grad_norm": 0.3931097090244293,
      "learning_rate": 0.00016627848101265824,
      "loss": 0.2022,
      "num_input_tokens_seen": 630700,
      "step": 339,
      "train_runtime": 681.5671,
      "train_tokens_per_second": 925.367
    },
    {
      "epoch": 0.3436947182208744,
      "grad_norm": 0.3931916654109955,
      "learning_rate": 0.00016617721518987343,
      "loss": 0.2413,
      "num_input_tokens_seen": 632710,
      "step": 340,
      "train_runtime": 683.408,
      "train_tokens_per_second": 925.816
    },
    {
      "epoch": 0.3447055850391711,
      "grad_norm": 0.5007351636886597,
      "learning_rate": 0.00016607594936708862,
      "loss": 0.2408,
      "num_input_tokens_seen": 634168,
      "step": 341,
      "train_runtime": 685.0014,
      "train_tokens_per_second": 925.791
    },
    {
      "epoch": 0.3457164518574678,
      "grad_norm": 0.36407172679901123,
      "learning_rate": 0.0001659746835443038,
      "loss": 0.2262,
      "num_input_tokens_seen": 636054,
      "step": 342,
      "train_runtime": 686.7729,
      "train_tokens_per_second": 926.149
    },
    {
      "epoch": 0.34672731867576445,
      "grad_norm": 0.39440491795539856,
      "learning_rate": 0.000165873417721519,
      "loss": 0.194,
      "num_input_tokens_seen": 638242,
      "step": 343,
      "train_runtime": 688.6932,
      "train_tokens_per_second": 926.744
    },
    {
      "epoch": 0.34773818549406116,
      "grad_norm": 0.45356428623199463,
      "learning_rate": 0.0001657721518987342,
      "loss": 0.1776,
      "num_input_tokens_seen": 640084,
      "step": 344,
      "train_runtime": 690.4041,
      "train_tokens_per_second": 927.115
    },
    {
      "epoch": 0.34874905231235787,
      "grad_norm": 0.47393128275871277,
      "learning_rate": 0.00016567088607594938,
      "loss": 0.1357,
      "num_input_tokens_seen": 641340,
      "step": 345,
      "train_runtime": 691.9391,
      "train_tokens_per_second": 926.873
    },
    {
      "epoch": 0.3497599191306545,
      "grad_norm": 0.5046817064285278,
      "learning_rate": 0.00016556962025316457,
      "loss": 0.241,
      "num_input_tokens_seen": 643024,
      "step": 346,
      "train_runtime": 693.5728,
      "train_tokens_per_second": 927.118
    },
    {
      "epoch": 0.3507707859489512,
      "grad_norm": 0.3809065520763397,
      "learning_rate": 0.00016546835443037976,
      "loss": 0.2097,
      "num_input_tokens_seen": 645680,
      "step": 347,
      "train_runtime": 695.7296,
      "train_tokens_per_second": 928.062
    },
    {
      "epoch": 0.35178165276724793,
      "grad_norm": 0.5797727108001709,
      "learning_rate": 0.00016536708860759495,
      "loss": 0.1894,
      "num_input_tokens_seen": 646868,
      "step": 348,
      "train_runtime": 697.2946,
      "train_tokens_per_second": 927.683
    },
    {
      "epoch": 0.3527925195855446,
      "grad_norm": 0.3871445953845978,
      "learning_rate": 0.00016526582278481014,
      "loss": 0.1193,
      "num_input_tokens_seen": 648398,
      "step": 349,
      "train_runtime": 698.9061,
      "train_tokens_per_second": 927.733
    },
    {
      "epoch": 0.3538033864038413,
      "grad_norm": 0.3448233902454376,
      "learning_rate": 0.00016516455696202533,
      "loss": 0.1705,
      "num_input_tokens_seen": 650210,
      "step": 350,
      "train_runtime": 700.6107,
      "train_tokens_per_second": 928.062
    },
    {
      "epoch": 0.354814253222138,
      "grad_norm": 0.5008251667022705,
      "learning_rate": 0.0001650632911392405,
      "loss": 0.1996,
      "num_input_tokens_seen": 652020,
      "step": 351,
      "train_runtime": 702.3609,
      "train_tokens_per_second": 928.326
    },
    {
      "epoch": 0.35582512004043465,
      "grad_norm": 0.453442245721817,
      "learning_rate": 0.00016496202531645572,
      "loss": 0.2222,
      "num_input_tokens_seen": 653632,
      "step": 352,
      "train_runtime": 703.9744,
      "train_tokens_per_second": 928.488
    },
    {
      "epoch": 0.35683598685873136,
      "grad_norm": 0.46873143315315247,
      "learning_rate": 0.0001648607594936709,
      "loss": 0.224,
      "num_input_tokens_seen": 655572,
      "step": 353,
      "train_runtime": 705.6982,
      "train_tokens_per_second": 928.969
    },
    {
      "epoch": 0.35784685367702807,
      "grad_norm": 0.44376879930496216,
      "learning_rate": 0.00016475949367088607,
      "loss": 0.1884,
      "num_input_tokens_seen": 657532,
      "step": 354,
      "train_runtime": 707.5266,
      "train_tokens_per_second": 929.339
    },
    {
      "epoch": 0.3588577204953247,
      "grad_norm": 0.26693281531333923,
      "learning_rate": 0.0001646582278481013,
      "loss": 0.1841,
      "num_input_tokens_seen": 660468,
      "step": 355,
      "train_runtime": 709.8611,
      "train_tokens_per_second": 930.419
    },
    {
      "epoch": 0.3598685873136214,
      "grad_norm": 0.41272827982902527,
      "learning_rate": 0.00016455696202531648,
      "loss": 0.1723,
      "num_input_tokens_seen": 661950,
      "step": 356,
      "train_runtime": 711.4576,
      "train_tokens_per_second": 930.414
    },
    {
      "epoch": 0.36087945413191813,
      "grad_norm": 0.45240455865859985,
      "learning_rate": 0.00016445569620253164,
      "loss": 0.2,
      "num_input_tokens_seen": 664054,
      "step": 357,
      "train_runtime": 713.2613,
      "train_tokens_per_second": 931.011
    },
    {
      "epoch": 0.3618903209502148,
      "grad_norm": 0.47014230489730835,
      "learning_rate": 0.00016435443037974686,
      "loss": 0.147,
      "num_input_tokens_seen": 665740,
      "step": 358,
      "train_runtime": 714.8885,
      "train_tokens_per_second": 931.25
    },
    {
      "epoch": 0.3629011877685115,
      "grad_norm": 0.44238078594207764,
      "learning_rate": 0.00016425316455696202,
      "loss": 0.2292,
      "num_input_tokens_seen": 667738,
      "step": 359,
      "train_runtime": 716.669,
      "train_tokens_per_second": 931.724
    },
    {
      "epoch": 0.3639120545868082,
      "grad_norm": 0.4482760429382324,
      "learning_rate": 0.00016415189873417722,
      "loss": 0.2046,
      "num_input_tokens_seen": 669706,
      "step": 360,
      "train_runtime": 718.5261,
      "train_tokens_per_second": 932.055
    },
    {
      "epoch": 0.36492292140510485,
      "grad_norm": 0.4080590307712555,
      "learning_rate": 0.00016405063291139243,
      "loss": 0.1379,
      "num_input_tokens_seen": 671454,
      "step": 361,
      "train_runtime": 722.292,
      "train_tokens_per_second": 929.616
    },
    {
      "epoch": 0.36593378822340156,
      "grad_norm": 0.5153780579566956,
      "learning_rate": 0.0001639493670886076,
      "loss": 0.302,
      "num_input_tokens_seen": 673490,
      "step": 362,
      "train_runtime": 724.0677,
      "train_tokens_per_second": 930.148
    },
    {
      "epoch": 0.36694465504169826,
      "grad_norm": 0.492509126663208,
      "learning_rate": 0.0001638481012658228,
      "loss": 0.2379,
      "num_input_tokens_seen": 675348,
      "step": 363,
      "train_runtime": 725.8596,
      "train_tokens_per_second": 930.411
    },
    {
      "epoch": 0.36795552185999497,
      "grad_norm": 0.3708415627479553,
      "learning_rate": 0.000163746835443038,
      "loss": 0.1876,
      "num_input_tokens_seen": 676866,
      "step": 364,
      "train_runtime": 727.5376,
      "train_tokens_per_second": 930.352
    },
    {
      "epoch": 0.3689663886782916,
      "grad_norm": 0.3933420181274414,
      "learning_rate": 0.00016364556962025317,
      "loss": 0.1904,
      "num_input_tokens_seen": 678574,
      "step": 365,
      "train_runtime": 729.1914,
      "train_tokens_per_second": 930.584
    },
    {
      "epoch": 0.36997725549658833,
      "grad_norm": 0.4342649281024933,
      "learning_rate": 0.00016354430379746836,
      "loss": 0.3707,
      "num_input_tokens_seen": 680838,
      "step": 366,
      "train_runtime": 731.1776,
      "train_tokens_per_second": 931.153
    },
    {
      "epoch": 0.37098812231488504,
      "grad_norm": 0.3595677316188812,
      "learning_rate": 0.00016344303797468355,
      "loss": 0.3111,
      "num_input_tokens_seen": 682828,
      "step": 367,
      "train_runtime": 732.978,
      "train_tokens_per_second": 931.58
    },
    {
      "epoch": 0.3719989891331817,
      "grad_norm": 0.48728400468826294,
      "learning_rate": 0.00016334177215189874,
      "loss": 0.2889,
      "num_input_tokens_seen": 684490,
      "step": 368,
      "train_runtime": 734.6687,
      "train_tokens_per_second": 931.699
    },
    {
      "epoch": 0.3730098559514784,
      "grad_norm": 0.47918418049812317,
      "learning_rate": 0.00016324050632911393,
      "loss": 0.211,
      "num_input_tokens_seen": 686330,
      "step": 369,
      "train_runtime": 736.4335,
      "train_tokens_per_second": 931.965
    },
    {
      "epoch": 0.3740207227697751,
      "grad_norm": 0.4765119254589081,
      "learning_rate": 0.00016313924050632912,
      "loss": 0.2642,
      "num_input_tokens_seen": 688430,
      "step": 370,
      "train_runtime": 738.2917,
      "train_tokens_per_second": 932.463
    },
    {
      "epoch": 0.37503158958807176,
      "grad_norm": 0.35585442185401917,
      "learning_rate": 0.00016303797468354431,
      "loss": 0.143,
      "num_input_tokens_seen": 690160,
      "step": 371,
      "train_runtime": 739.9946,
      "train_tokens_per_second": 932.656
    },
    {
      "epoch": 0.37604245640636846,
      "grad_norm": 0.37591856718063354,
      "learning_rate": 0.00016293670886075948,
      "loss": 0.1499,
      "num_input_tokens_seen": 691482,
      "step": 372,
      "train_runtime": 741.5746,
      "train_tokens_per_second": 932.451
    },
    {
      "epoch": 0.37705332322466517,
      "grad_norm": 0.40079328417778015,
      "learning_rate": 0.0001628354430379747,
      "loss": 0.2112,
      "num_input_tokens_seen": 693254,
      "step": 373,
      "train_runtime": 743.2675,
      "train_tokens_per_second": 932.711
    },
    {
      "epoch": 0.3780641900429618,
      "grad_norm": 0.5272864103317261,
      "learning_rate": 0.0001627341772151899,
      "loss": 0.2159,
      "num_input_tokens_seen": 694622,
      "step": 374,
      "train_runtime": 744.8578,
      "train_tokens_per_second": 932.557
    },
    {
      "epoch": 0.37907505686125853,
      "grad_norm": 0.3017108142375946,
      "learning_rate": 0.00016263291139240505,
      "loss": 0.3006,
      "num_input_tokens_seen": 697032,
      "step": 375,
      "train_runtime": 746.853,
      "train_tokens_per_second": 933.292
    },
    {
      "epoch": 0.38008592367955524,
      "grad_norm": 0.35110485553741455,
      "learning_rate": 0.00016253164556962027,
      "loss": 0.2897,
      "num_input_tokens_seen": 698714,
      "step": 376,
      "train_runtime": 748.5523,
      "train_tokens_per_second": 933.42
    },
    {
      "epoch": 0.3810967904978519,
      "grad_norm": 0.4561986029148102,
      "learning_rate": 0.00016243037974683546,
      "loss": 0.2057,
      "num_input_tokens_seen": 700308,
      "step": 377,
      "train_runtime": 750.1681,
      "train_tokens_per_second": 933.535
    },
    {
      "epoch": 0.3821076573161486,
      "grad_norm": 0.35886335372924805,
      "learning_rate": 0.00016232911392405062,
      "loss": 0.1541,
      "num_input_tokens_seen": 701994,
      "step": 378,
      "train_runtime": 751.8394,
      "train_tokens_per_second": 933.702
    },
    {
      "epoch": 0.3831185241344453,
      "grad_norm": 0.3360520899295807,
      "learning_rate": 0.00016222784810126584,
      "loss": 0.1871,
      "num_input_tokens_seen": 703676,
      "step": 379,
      "train_runtime": 753.472,
      "train_tokens_per_second": 933.911
    },
    {
      "epoch": 0.38412939095274196,
      "grad_norm": 0.4519248306751251,
      "learning_rate": 0.000162126582278481,
      "loss": 0.1392,
      "num_input_tokens_seen": 705024,
      "step": 380,
      "train_runtime": 755.0632,
      "train_tokens_per_second": 933.728
    },
    {
      "epoch": 0.38514025777103866,
      "grad_norm": 0.36681482195854187,
      "learning_rate": 0.0001620253164556962,
      "loss": 0.2327,
      "num_input_tokens_seen": 707296,
      "step": 381,
      "train_runtime": 757.0505,
      "train_tokens_per_second": 934.278
    },
    {
      "epoch": 0.38615112458933537,
      "grad_norm": 0.5280442833900452,
      "learning_rate": 0.00016192405063291141,
      "loss": 0.1413,
      "num_input_tokens_seen": 708956,
      "step": 382,
      "train_runtime": 758.7625,
      "train_tokens_per_second": 934.358
    },
    {
      "epoch": 0.387161991407632,
      "grad_norm": 0.521888792514801,
      "learning_rate": 0.00016182278481012658,
      "loss": 0.1752,
      "num_input_tokens_seen": 710732,
      "step": 383,
      "train_runtime": 760.5251,
      "train_tokens_per_second": 934.528
    },
    {
      "epoch": 0.38817285822592873,
      "grad_norm": 0.5253322720527649,
      "learning_rate": 0.00016172151898734177,
      "loss": 0.2107,
      "num_input_tokens_seen": 712236,
      "step": 384,
      "train_runtime": 762.1972,
      "train_tokens_per_second": 934.451
    },
    {
      "epoch": 0.38918372504422544,
      "grad_norm": 0.49969637393951416,
      "learning_rate": 0.000161620253164557,
      "loss": 0.1558,
      "num_input_tokens_seen": 713692,
      "step": 385,
      "train_runtime": 763.8581,
      "train_tokens_per_second": 934.325
    },
    {
      "epoch": 0.3901945918625221,
      "grad_norm": 0.4234883487224579,
      "learning_rate": 0.00016151898734177215,
      "loss": 0.2551,
      "num_input_tokens_seen": 715676,
      "step": 386,
      "train_runtime": 765.6249,
      "train_tokens_per_second": 934.761
    },
    {
      "epoch": 0.3912054586808188,
      "grad_norm": 0.40212076902389526,
      "learning_rate": 0.00016141772151898734,
      "loss": 0.2003,
      "num_input_tokens_seen": 717782,
      "step": 387,
      "train_runtime": 767.426,
      "train_tokens_per_second": 935.311
    },
    {
      "epoch": 0.3922163254991155,
      "grad_norm": 0.471465528011322,
      "learning_rate": 0.00016131645569620253,
      "loss": 0.1965,
      "num_input_tokens_seen": 719358,
      "step": 388,
      "train_runtime": 769.0566,
      "train_tokens_per_second": 935.377
    },
    {
      "epoch": 0.39322719231741216,
      "grad_norm": 0.3209170401096344,
      "learning_rate": 0.00016121518987341772,
      "loss": 0.2519,
      "num_input_tokens_seen": 721732,
      "step": 389,
      "train_runtime": 771.0593,
      "train_tokens_per_second": 936.027
    },
    {
      "epoch": 0.39423805913570886,
      "grad_norm": 0.4539428949356079,
      "learning_rate": 0.00016111392405063291,
      "loss": 0.3083,
      "num_input_tokens_seen": 723654,
      "step": 390,
      "train_runtime": 772.7774,
      "train_tokens_per_second": 936.433
    },
    {
      "epoch": 0.39524892595400557,
      "grad_norm": 0.33022862672805786,
      "learning_rate": 0.0001610126582278481,
      "loss": 0.1552,
      "num_input_tokens_seen": 725402,
      "step": 391,
      "train_runtime": 777.0095,
      "train_tokens_per_second": 933.582
    },
    {
      "epoch": 0.3962597927723022,
      "grad_norm": 0.43771734833717346,
      "learning_rate": 0.0001609113924050633,
      "loss": 0.2918,
      "num_input_tokens_seen": 727426,
      "step": 392,
      "train_runtime": 778.864,
      "train_tokens_per_second": 933.958
    },
    {
      "epoch": 0.39727065959059893,
      "grad_norm": 0.429280161857605,
      "learning_rate": 0.00016081012658227849,
      "loss": 0.2256,
      "num_input_tokens_seen": 729244,
      "step": 393,
      "train_runtime": 780.5887,
      "train_tokens_per_second": 934.223
    },
    {
      "epoch": 0.39828152640889564,
      "grad_norm": 0.4618457853794098,
      "learning_rate": 0.00016070886075949368,
      "loss": 0.2437,
      "num_input_tokens_seen": 731254,
      "step": 394,
      "train_runtime": 782.3836,
      "train_tokens_per_second": 934.649
    },
    {
      "epoch": 0.39929239322719234,
      "grad_norm": 0.43775153160095215,
      "learning_rate": 0.00016060759493670887,
      "loss": 0.3196,
      "num_input_tokens_seen": 732820,
      "step": 395,
      "train_runtime": 784.0466,
      "train_tokens_per_second": 934.664
    },
    {
      "epoch": 0.400303260045489,
      "grad_norm": 0.35830020904541016,
      "learning_rate": 0.00016050632911392406,
      "loss": 0.2727,
      "num_input_tokens_seen": 734410,
      "step": 396,
      "train_runtime": 785.7472,
      "train_tokens_per_second": 934.664
    },
    {
      "epoch": 0.400303260045489,
      "eval_loss": 0.20691320300102234,
      "eval_runtime": 61.343,
      "eval_samples_per_second": 14.329,
      "eval_steps_per_second": 7.173,
      "num_input_tokens_seen": 734410,
      "step": 396
    },
    {
      "epoch": 0.4013141268637857,
      "grad_norm": 0.2852393388748169,
      "learning_rate": 0.00016040506329113925,
      "loss": 0.1643,
      "num_input_tokens_seen": 736226,
      "step": 397,
      "train_runtime": 848.7979,
      "train_tokens_per_second": 867.375
    },
    {
      "epoch": 0.4023249936820824,
      "grad_norm": 0.40798038244247437,
      "learning_rate": 0.00016030379746835444,
      "loss": 0.2066,
      "num_input_tokens_seen": 738040,
      "step": 398,
      "train_runtime": 850.5616,
      "train_tokens_per_second": 867.709
    },
    {
      "epoch": 0.40333586050037906,
      "grad_norm": 0.36145034432411194,
      "learning_rate": 0.00016020253164556963,
      "loss": 0.1648,
      "num_input_tokens_seen": 739850,
      "step": 399,
      "train_runtime": 852.3272,
      "train_tokens_per_second": 868.035
    },
    {
      "epoch": 0.40434672731867577,
      "grad_norm": 0.34595760703086853,
      "learning_rate": 0.00016010126582278482,
      "loss": 0.2276,
      "num_input_tokens_seen": 741854,
      "step": 400,
      "train_runtime": 854.192,
      "train_tokens_per_second": 868.486
    },
    {
      "epoch": 0.4053575941369725,
      "grad_norm": 0.3657154142856598,
      "learning_rate": 0.00016,
      "loss": 0.2051,
      "num_input_tokens_seen": 743758,
      "step": 401,
      "train_runtime": 855.8502,
      "train_tokens_per_second": 869.028
    },
    {
      "epoch": 0.40636846095526913,
      "grad_norm": 0.4583825469017029,
      "learning_rate": 0.0001598987341772152,
      "loss": 0.2541,
      "num_input_tokens_seen": 745548,
      "step": 402,
      "train_runtime": 857.5461,
      "train_tokens_per_second": 869.397
    },
    {
      "epoch": 0.40737932777356584,
      "grad_norm": 0.47932156920433044,
      "learning_rate": 0.0001597974683544304,
      "loss": 0.2077,
      "num_input_tokens_seen": 747458,
      "step": 403,
      "train_runtime": 859.3099,
      "train_tokens_per_second": 869.835
    },
    {
      "epoch": 0.40839019459186254,
      "grad_norm": 0.428975909948349,
      "learning_rate": 0.00015969620253164556,
      "loss": 0.1823,
      "num_input_tokens_seen": 749572,
      "step": 404,
      "train_runtime": 861.1996,
      "train_tokens_per_second": 870.381
    },
    {
      "epoch": 0.4094010614101592,
      "grad_norm": 0.44719934463500977,
      "learning_rate": 0.00015959493670886078,
      "loss": 0.2358,
      "num_input_tokens_seen": 751616,
      "step": 405,
      "train_runtime": 863.0914,
      "train_tokens_per_second": 870.842
    },
    {
      "epoch": 0.4104119282284559,
      "grad_norm": 0.4692244231700897,
      "learning_rate": 0.00015949367088607597,
      "loss": 0.1761,
      "num_input_tokens_seen": 753438,
      "step": 406,
      "train_runtime": 864.7987,
      "train_tokens_per_second": 871.229
    },
    {
      "epoch": 0.4114227950467526,
      "grad_norm": 0.43564677238464355,
      "learning_rate": 0.00015939240506329113,
      "loss": 0.2654,
      "num_input_tokens_seen": 755350,
      "step": 407,
      "train_runtime": 866.531,
      "train_tokens_per_second": 871.694
    },
    {
      "epoch": 0.41243366186504926,
      "grad_norm": 0.4037442207336426,
      "learning_rate": 0.00015929113924050635,
      "loss": 0.1779,
      "num_input_tokens_seen": 757048,
      "step": 408,
      "train_runtime": 868.1611,
      "train_tokens_per_second": 872.013
    },
    {
      "epoch": 0.41344452868334597,
      "grad_norm": 0.5000036358833313,
      "learning_rate": 0.00015918987341772154,
      "loss": 0.3085,
      "num_input_tokens_seen": 759332,
      "step": 409,
      "train_runtime": 870.0899,
      "train_tokens_per_second": 872.705
    },
    {
      "epoch": 0.4144553955016427,
      "grad_norm": 0.446573406457901,
      "learning_rate": 0.0001590886075949367,
      "loss": 0.1944,
      "num_input_tokens_seen": 761534,
      "step": 410,
      "train_runtime": 872.027,
      "train_tokens_per_second": 873.292
    },
    {
      "epoch": 0.41546626231993933,
      "grad_norm": 0.3965902030467987,
      "learning_rate": 0.00015898734177215192,
      "loss": 0.1915,
      "num_input_tokens_seen": 763376,
      "step": 411,
      "train_runtime": 873.7386,
      "train_tokens_per_second": 873.689
    },
    {
      "epoch": 0.41647712913823604,
      "grad_norm": 0.25173959136009216,
      "learning_rate": 0.00015888607594936709,
      "loss": 0.2171,
      "num_input_tokens_seen": 766474,
      "step": 412,
      "train_runtime": 876.1057,
      "train_tokens_per_second": 874.865
    },
    {
      "epoch": 0.41748799595653274,
      "grad_norm": 0.2912166714668274,
      "learning_rate": 0.00015878481012658228,
      "loss": 0.195,
      "num_input_tokens_seen": 768510,
      "step": 413,
      "train_runtime": 877.9189,
      "train_tokens_per_second": 875.377
    },
    {
      "epoch": 0.4184988627748294,
      "grad_norm": 0.5279189944267273,
      "learning_rate": 0.0001586835443037975,
      "loss": 0.2741,
      "num_input_tokens_seen": 770568,
      "step": 414,
      "train_runtime": 879.7505,
      "train_tokens_per_second": 875.894
    },
    {
      "epoch": 0.4195097295931261,
      "grad_norm": 0.4628985822200775,
      "learning_rate": 0.00015858227848101266,
      "loss": 0.1722,
      "num_input_tokens_seen": 772174,
      "step": 415,
      "train_runtime": 881.4115,
      "train_tokens_per_second": 876.065
    },
    {
      "epoch": 0.4205205964114228,
      "grad_norm": 0.4825090765953064,
      "learning_rate": 0.00015848101265822785,
      "loss": 0.2368,
      "num_input_tokens_seen": 773982,
      "step": 416,
      "train_runtime": 883.1648,
      "train_tokens_per_second": 876.373
    },
    {
      "epoch": 0.42153146322971946,
      "grad_norm": 0.3585876524448395,
      "learning_rate": 0.00015837974683544307,
      "loss": 0.1864,
      "num_input_tokens_seen": 776102,
      "step": 417,
      "train_runtime": 885.0508,
      "train_tokens_per_second": 876.901
    },
    {
      "epoch": 0.42254233004801617,
      "grad_norm": 0.34322065114974976,
      "learning_rate": 0.00015827848101265823,
      "loss": 0.1826,
      "num_input_tokens_seen": 777932,
      "step": 418,
      "train_runtime": 886.7617,
      "train_tokens_per_second": 877.273
    },
    {
      "epoch": 0.4235531968663129,
      "grad_norm": 0.43742963671684265,
      "learning_rate": 0.00015817721518987342,
      "loss": 0.1833,
      "num_input_tokens_seen": 779846,
      "step": 419,
      "train_runtime": 888.5002,
      "train_tokens_per_second": 877.711
    },
    {
      "epoch": 0.4245640636846095,
      "grad_norm": 0.398015558719635,
      "learning_rate": 0.0001580759493670886,
      "loss": 0.2383,
      "num_input_tokens_seen": 781654,
      "step": 420,
      "train_runtime": 890.2081,
      "train_tokens_per_second": 878.058
    },
    {
      "epoch": 0.42557493050290623,
      "grad_norm": 0.4031146764755249,
      "learning_rate": 0.0001579746835443038,
      "loss": 0.1862,
      "num_input_tokens_seen": 783354,
      "step": 421,
      "train_runtime": 893.8915,
      "train_tokens_per_second": 876.341
    },
    {
      "epoch": 0.42658579732120294,
      "grad_norm": 0.35829827189445496,
      "learning_rate": 0.000157873417721519,
      "loss": 0.1866,
      "num_input_tokens_seen": 785382,
      "step": 422,
      "train_runtime": 895.7333,
      "train_tokens_per_second": 876.803
    },
    {
      "epoch": 0.4275966641394996,
      "grad_norm": 0.3397190272808075,
      "learning_rate": 0.00015777215189873418,
      "loss": 0.117,
      "num_input_tokens_seen": 787096,
      "step": 423,
      "train_runtime": 897.5052,
      "train_tokens_per_second": 876.982
    },
    {
      "epoch": 0.4286075309577963,
      "grad_norm": 0.39190611243247986,
      "learning_rate": 0.00015767088607594938,
      "loss": 0.1694,
      "num_input_tokens_seen": 788940,
      "step": 424,
      "train_runtime": 899.2529,
      "train_tokens_per_second": 877.328
    },
    {
      "epoch": 0.429618397776093,
      "grad_norm": 0.3337407112121582,
      "learning_rate": 0.00015756962025316457,
      "loss": 0.2326,
      "num_input_tokens_seen": 791164,
      "step": 425,
      "train_runtime": 901.1837,
      "train_tokens_per_second": 877.916
    },
    {
      "epoch": 0.4306292645943897,
      "grad_norm": 0.46486571431159973,
      "learning_rate": 0.00015746835443037976,
      "loss": 0.2316,
      "num_input_tokens_seen": 792874,
      "step": 426,
      "train_runtime": 902.8159,
      "train_tokens_per_second": 878.223
    },
    {
      "epoch": 0.43164013141268637,
      "grad_norm": 0.4039671719074249,
      "learning_rate": 0.00015736708860759495,
      "loss": 0.1757,
      "num_input_tokens_seen": 794330,
      "step": 427,
      "train_runtime": 904.4436,
      "train_tokens_per_second": 878.253
    },
    {
      "epoch": 0.4326509982309831,
      "grad_norm": 0.36065515875816345,
      "learning_rate": 0.0001572658227848101,
      "loss": 0.1402,
      "num_input_tokens_seen": 796094,
      "step": 428,
      "train_runtime": 906.134,
      "train_tokens_per_second": 878.561
    },
    {
      "epoch": 0.4336618650492798,
      "grad_norm": 0.3220762014389038,
      "learning_rate": 0.00015716455696202533,
      "loss": 0.1747,
      "num_input_tokens_seen": 798060,
      "step": 429,
      "train_runtime": 907.8011,
      "train_tokens_per_second": 879.113
    },
    {
      "epoch": 0.43467273186757643,
      "grad_norm": 0.42454615235328674,
      "learning_rate": 0.00015706329113924052,
      "loss": 0.2688,
      "num_input_tokens_seen": 799862,
      "step": 430,
      "train_runtime": 909.4977,
      "train_tokens_per_second": 879.455
    },
    {
      "epoch": 0.43568359868587314,
      "grad_norm": 0.5222974419593811,
      "learning_rate": 0.00015696202531645568,
      "loss": 0.2158,
      "num_input_tokens_seen": 801506,
      "step": 431,
      "train_runtime": 911.1798,
      "train_tokens_per_second": 879.635
    },
    {
      "epoch": 0.43669446550416985,
      "grad_norm": 0.38617977499961853,
      "learning_rate": 0.0001568607594936709,
      "loss": 0.1557,
      "num_input_tokens_seen": 804264,
      "step": 432,
      "train_runtime": 913.3571,
      "train_tokens_per_second": 880.558
    },
    {
      "epoch": 0.4377053323224665,
      "grad_norm": 0.40649381279945374,
      "learning_rate": 0.0001567594936708861,
      "loss": 0.194,
      "num_input_tokens_seen": 806084,
      "step": 433,
      "train_runtime": 915.0544,
      "train_tokens_per_second": 880.914
    },
    {
      "epoch": 0.4387161991407632,
      "grad_norm": 0.5631338357925415,
      "learning_rate": 0.00015665822784810126,
      "loss": 0.2402,
      "num_input_tokens_seen": 807584,
      "step": 434,
      "train_runtime": 916.6591,
      "train_tokens_per_second": 881.008
    },
    {
      "epoch": 0.4397270659590599,
      "grad_norm": 0.4309951663017273,
      "learning_rate": 0.00015655696202531647,
      "loss": 0.3269,
      "num_input_tokens_seen": 809506,
      "step": 435,
      "train_runtime": 918.4248,
      "train_tokens_per_second": 881.407
    },
    {
      "epoch": 0.44073793277735657,
      "grad_norm": 0.35738056898117065,
      "learning_rate": 0.00015645569620253164,
      "loss": 0.1802,
      "num_input_tokens_seen": 811374,
      "step": 436,
      "train_runtime": 920.0792,
      "train_tokens_per_second": 881.852
    },
    {
      "epoch": 0.4417487995956533,
      "grad_norm": 0.462649405002594,
      "learning_rate": 0.00015635443037974683,
      "loss": 0.2103,
      "num_input_tokens_seen": 813486,
      "step": 437,
      "train_runtime": 921.9783,
      "train_tokens_per_second": 882.327
    },
    {
      "epoch": 0.44275966641395,
      "grad_norm": 0.29911482334136963,
      "learning_rate": 0.00015625316455696205,
      "loss": 0.1508,
      "num_input_tokens_seen": 815772,
      "step": 438,
      "train_runtime": 923.9162,
      "train_tokens_per_second": 882.95
    },
    {
      "epoch": 0.44377053323224663,
      "grad_norm": 0.4340844452381134,
      "learning_rate": 0.0001561518987341772,
      "loss": 0.2338,
      "num_input_tokens_seen": 818164,
      "step": 439,
      "train_runtime": 925.9139,
      "train_tokens_per_second": 883.629
    },
    {
      "epoch": 0.44478140005054334,
      "grad_norm": 0.4801517426967621,
      "learning_rate": 0.0001560506329113924,
      "loss": 0.185,
      "num_input_tokens_seen": 820120,
      "step": 440,
      "train_runtime": 927.7473,
      "train_tokens_per_second": 883.991
    },
    {
      "epoch": 0.44579226686884005,
      "grad_norm": 0.4134570062160492,
      "learning_rate": 0.00015594936708860762,
      "loss": 0.3063,
      "num_input_tokens_seen": 821738,
      "step": 441,
      "train_runtime": 929.4355,
      "train_tokens_per_second": 884.126
    },
    {
      "epoch": 0.4468031336871367,
      "grad_norm": 0.3586715757846832,
      "learning_rate": 0.00015584810126582278,
      "loss": 0.2401,
      "num_input_tokens_seen": 823640,
      "step": 442,
      "train_runtime": 931.1808,
      "train_tokens_per_second": 884.511
    },
    {
      "epoch": 0.4478140005054334,
      "grad_norm": 0.4239972233772278,
      "learning_rate": 0.00015574683544303797,
      "loss": 0.2167,
      "num_input_tokens_seen": 825460,
      "step": 443,
      "train_runtime": 933.0131,
      "train_tokens_per_second": 884.725
    },
    {
      "epoch": 0.4488248673237301,
      "grad_norm": 0.3098328113555908,
      "learning_rate": 0.00015564556962025317,
      "loss": 0.2958,
      "num_input_tokens_seen": 827636,
      "step": 444,
      "train_runtime": 934.8758,
      "train_tokens_per_second": 885.29
    },
    {
      "epoch": 0.44983573414202677,
      "grad_norm": 0.3603465259075165,
      "learning_rate": 0.00015554430379746836,
      "loss": 0.2206,
      "num_input_tokens_seen": 829770,
      "step": 445,
      "train_runtime": 936.6778,
      "train_tokens_per_second": 885.865
    },
    {
      "epoch": 0.4508466009603235,
      "grad_norm": 0.31367743015289307,
      "learning_rate": 0.00015544303797468355,
      "loss": 0.2102,
      "num_input_tokens_seen": 831986,
      "step": 446,
      "train_runtime": 938.6113,
      "train_tokens_per_second": 886.401
    },
    {
      "epoch": 0.4518574677786202,
      "grad_norm": 0.4784441888332367,
      "learning_rate": 0.00015534177215189874,
      "loss": 0.2591,
      "num_input_tokens_seen": 833484,
      "step": 447,
      "train_runtime": 940.2268,
      "train_tokens_per_second": 886.471
    },
    {
      "epoch": 0.45286833459691683,
      "grad_norm": 0.3459112346172333,
      "learning_rate": 0.00015524050632911393,
      "loss": 0.3216,
      "num_input_tokens_seen": 835392,
      "step": 448,
      "train_runtime": 942.015,
      "train_tokens_per_second": 886.814
    },
    {
      "epoch": 0.45387920141521354,
      "grad_norm": 0.302502304315567,
      "learning_rate": 0.00015513924050632912,
      "loss": 0.141,
      "num_input_tokens_seen": 837616,
      "step": 449,
      "train_runtime": 943.9455,
      "train_tokens_per_second": 887.356
    },
    {
      "epoch": 0.45489006823351025,
      "grad_norm": 0.33230262994766235,
      "learning_rate": 0.0001550379746835443,
      "loss": 0.2712,
      "num_input_tokens_seen": 840092,
      "step": 450,
      "train_runtime": 945.9858,
      "train_tokens_per_second": 888.06
    },
    {
      "epoch": 0.4559009350518069,
      "grad_norm": 0.2827284634113312,
      "learning_rate": 0.0001549367088607595,
      "loss": 0.1218,
      "num_input_tokens_seen": 842154,
      "step": 451,
      "train_runtime": 950.0303,
      "train_tokens_per_second": 886.45
    },
    {
      "epoch": 0.4569118018701036,
      "grad_norm": 0.35446277260780334,
      "learning_rate": 0.0001548354430379747,
      "loss": 0.238,
      "num_input_tokens_seen": 844514,
      "step": 452,
      "train_runtime": 952.0426,
      "train_tokens_per_second": 887.055
    },
    {
      "epoch": 0.4579226686884003,
      "grad_norm": 0.3898954689502716,
      "learning_rate": 0.00015473417721518988,
      "loss": 0.2343,
      "num_input_tokens_seen": 846476,
      "step": 453,
      "train_runtime": 953.8182,
      "train_tokens_per_second": 887.461
    },
    {
      "epoch": 0.45893353550669697,
      "grad_norm": 0.5028534531593323,
      "learning_rate": 0.00015463291139240507,
      "loss": 0.2647,
      "num_input_tokens_seen": 848444,
      "step": 454,
      "train_runtime": 955.6586,
      "train_tokens_per_second": 887.811
    },
    {
      "epoch": 0.4599444023249937,
      "grad_norm": 0.45474883913993835,
      "learning_rate": 0.00015453164556962026,
      "loss": 0.2045,
      "num_input_tokens_seen": 850126,
      "step": 455,
      "train_runtime": 957.3945,
      "train_tokens_per_second": 887.958
    },
    {
      "epoch": 0.4609552691432904,
      "grad_norm": 0.3534106910228729,
      "learning_rate": 0.00015443037974683546,
      "loss": 0.143,
      "num_input_tokens_seen": 851994,
      "step": 456,
      "train_runtime": 959.161,
      "train_tokens_per_second": 888.27
    },
    {
      "epoch": 0.4619661359615871,
      "grad_norm": 0.392352819442749,
      "learning_rate": 0.00015432911392405065,
      "loss": 0.2864,
      "num_input_tokens_seen": 854360,
      "step": 457,
      "train_runtime": 961.1696,
      "train_tokens_per_second": 888.875
    },
    {
      "epoch": 0.46297700277988374,
      "grad_norm": 0.481840044260025,
      "learning_rate": 0.00015422784810126584,
      "loss": 0.3185,
      "num_input_tokens_seen": 856634,
      "step": 458,
      "train_runtime": 963.1613,
      "train_tokens_per_second": 889.398
    },
    {
      "epoch": 0.46398786959818045,
      "grad_norm": 0.4527128040790558,
      "learning_rate": 0.00015412658227848103,
      "loss": 0.1872,
      "num_input_tokens_seen": 858330,
      "step": 459,
      "train_runtime": 964.7923,
      "train_tokens_per_second": 889.653
    },
    {
      "epoch": 0.46499873641647715,
      "grad_norm": 0.5048484206199646,
      "learning_rate": 0.0001540253164556962,
      "loss": 0.2681,
      "num_input_tokens_seen": 860108,
      "step": 460,
      "train_runtime": 966.4711,
      "train_tokens_per_second": 889.947
    },
    {
      "epoch": 0.4660096032347738,
      "grad_norm": 0.49292078614234924,
      "learning_rate": 0.0001539240506329114,
      "loss": 0.3287,
      "num_input_tokens_seen": 861860,
      "step": 461,
      "train_runtime": 968.1427,
      "train_tokens_per_second": 890.22
    },
    {
      "epoch": 0.4670204700530705,
      "grad_norm": 0.41087356209754944,
      "learning_rate": 0.0001538227848101266,
      "loss": 0.1956,
      "num_input_tokens_seen": 863356,
      "step": 462,
      "train_runtime": 969.7452,
      "train_tokens_per_second": 890.292
    },
    {
      "epoch": 0.4680313368713672,
      "grad_norm": 0.3939993977546692,
      "learning_rate": 0.00015372151898734176,
      "loss": 0.2667,
      "num_input_tokens_seen": 865252,
      "step": 463,
      "train_runtime": 971.5792,
      "train_tokens_per_second": 890.563
    },
    {
      "epoch": 0.4690422036896639,
      "grad_norm": 0.3936285972595215,
      "learning_rate": 0.00015362025316455698,
      "loss": 0.2168,
      "num_input_tokens_seen": 867254,
      "step": 464,
      "train_runtime": 973.3133,
      "train_tokens_per_second": 891.033
    },
    {
      "epoch": 0.4700530705079606,
      "grad_norm": 0.4447113871574402,
      "learning_rate": 0.00015351898734177217,
      "loss": 0.2208,
      "num_input_tokens_seen": 868816,
      "step": 465,
      "train_runtime": 974.9772,
      "train_tokens_per_second": 891.114
    },
    {
      "epoch": 0.4710639373262573,
      "grad_norm": 0.32850441336631775,
      "learning_rate": 0.00015341772151898734,
      "loss": 0.238,
      "num_input_tokens_seen": 870760,
      "step": 466,
      "train_runtime": 976.7589,
      "train_tokens_per_second": 891.479
    },
    {
      "epoch": 0.47207480414455394,
      "grad_norm": 0.4373947083950043,
      "learning_rate": 0.00015331645569620255,
      "loss": 0.1596,
      "num_input_tokens_seen": 872386,
      "step": 467,
      "train_runtime": 978.3926,
      "train_tokens_per_second": 891.652
    },
    {
      "epoch": 0.47308567096285065,
      "grad_norm": 0.34838488698005676,
      "learning_rate": 0.00015321518987341772,
      "loss": 0.23,
      "num_input_tokens_seen": 874220,
      "step": 468,
      "train_runtime": 980.1526,
      "train_tokens_per_second": 891.922
    },
    {
      "epoch": 0.47409653778114735,
      "grad_norm": 0.3535575866699219,
      "learning_rate": 0.0001531139240506329,
      "loss": 0.1748,
      "num_input_tokens_seen": 876046,
      "step": 469,
      "train_runtime": 981.9579,
      "train_tokens_per_second": 892.142
    },
    {
      "epoch": 0.475107404599444,
      "grad_norm": 0.4254951477050781,
      "learning_rate": 0.00015301265822784813,
      "loss": 0.1954,
      "num_input_tokens_seen": 877734,
      "step": 470,
      "train_runtime": 983.6426,
      "train_tokens_per_second": 892.33
    },
    {
      "epoch": 0.4761182714177407,
      "grad_norm": 0.4093538522720337,
      "learning_rate": 0.0001529113924050633,
      "loss": 0.2269,
      "num_input_tokens_seen": 879312,
      "step": 471,
      "train_runtime": 985.3279,
      "train_tokens_per_second": 892.406
    },
    {
      "epoch": 0.4771291382360374,
      "grad_norm": 0.42421266436576843,
      "learning_rate": 0.00015281012658227848,
      "loss": 0.1913,
      "num_input_tokens_seen": 881290,
      "step": 472,
      "train_runtime": 987.2085,
      "train_tokens_per_second": 892.709
    },
    {
      "epoch": 0.47814000505433407,
      "grad_norm": 0.47526732087135315,
      "learning_rate": 0.0001527088607594937,
      "loss": 0.2237,
      "num_input_tokens_seen": 882800,
      "step": 473,
      "train_runtime": 988.825,
      "train_tokens_per_second": 892.777
    },
    {
      "epoch": 0.4791508718726308,
      "grad_norm": 0.42744213342666626,
      "learning_rate": 0.00015260759493670886,
      "loss": 0.1693,
      "num_input_tokens_seen": 884462,
      "step": 474,
      "train_runtime": 990.4888,
      "train_tokens_per_second": 892.955
    },
    {
      "epoch": 0.4801617386909275,
      "grad_norm": 0.4059623181819916,
      "learning_rate": 0.00015250632911392405,
      "loss": 0.172,
      "num_input_tokens_seen": 886444,
      "step": 475,
      "train_runtime": 992.2762,
      "train_tokens_per_second": 893.344
    },
    {
      "epoch": 0.48117260550922414,
      "grad_norm": 0.4145307242870331,
      "learning_rate": 0.00015240506329113924,
      "loss": 0.2311,
      "num_input_tokens_seen": 888522,
      "step": 476,
      "train_runtime": 994.1315,
      "train_tokens_per_second": 893.767
    },
    {
      "epoch": 0.48218347232752085,
      "grad_norm": 0.4267613887786865,
      "learning_rate": 0.00015230379746835444,
      "loss": 0.1474,
      "num_input_tokens_seen": 890008,
      "step": 477,
      "train_runtime": 995.7988,
      "train_tokens_per_second": 893.763
    },
    {
      "epoch": 0.48319433914581755,
      "grad_norm": 0.30390217900276184,
      "learning_rate": 0.00015220253164556963,
      "loss": 0.3236,
      "num_input_tokens_seen": 892472,
      "step": 478,
      "train_runtime": 997.83,
      "train_tokens_per_second": 894.413
    },
    {
      "epoch": 0.4842052059641142,
      "grad_norm": 0.4135509729385376,
      "learning_rate": 0.00015210126582278482,
      "loss": 0.1404,
      "num_input_tokens_seen": 894160,
      "step": 479,
      "train_runtime": 999.5224,
      "train_tokens_per_second": 894.587
    },
    {
      "epoch": 0.4852160727824109,
      "grad_norm": 0.422590047121048,
      "learning_rate": 0.000152,
      "loss": 0.216,
      "num_input_tokens_seen": 896218,
      "step": 480,
      "train_runtime": 1001.3637,
      "train_tokens_per_second": 894.998
    },
    {
      "epoch": 0.4862269396007076,
      "grad_norm": 0.45324161648750305,
      "learning_rate": 0.0001518987341772152,
      "loss": 0.1877,
      "num_input_tokens_seen": 898014,
      "step": 481,
      "train_runtime": 1005.1959,
      "train_tokens_per_second": 893.372
    },
    {
      "epoch": 0.48723780641900427,
      "grad_norm": 0.3673161566257477,
      "learning_rate": 0.0001517974683544304,
      "loss": 0.1691,
      "num_input_tokens_seen": 900284,
      "step": 482,
      "train_runtime": 1007.1287,
      "train_tokens_per_second": 893.912
    },
    {
      "epoch": 0.488248673237301,
      "grad_norm": 0.3802099823951721,
      "learning_rate": 0.00015169620253164558,
      "loss": 0.2305,
      "num_input_tokens_seen": 902464,
      "step": 483,
      "train_runtime": 1009.0475,
      "train_tokens_per_second": 894.372
    },
    {
      "epoch": 0.4892595400555977,
      "grad_norm": 0.4809756577014923,
      "learning_rate": 0.00015159493670886074,
      "loss": 0.2301,
      "num_input_tokens_seen": 904160,
      "step": 484,
      "train_runtime": 1010.6663,
      "train_tokens_per_second": 894.618
    },
    {
      "epoch": 0.49027040687389434,
      "grad_norm": 0.2805626392364502,
      "learning_rate": 0.00015149367088607596,
      "loss": 0.2022,
      "num_input_tokens_seen": 905996,
      "step": 485,
      "train_runtime": 1012.3667,
      "train_tokens_per_second": 894.929
    },
    {
      "epoch": 0.49128127369219105,
      "grad_norm": 0.37523695826530457,
      "learning_rate": 0.00015139240506329115,
      "loss": 0.1751,
      "num_input_tokens_seen": 907368,
      "step": 486,
      "train_runtime": 1013.9526,
      "train_tokens_per_second": 894.882
    },
    {
      "epoch": 0.49229214051048775,
      "grad_norm": 0.4012540876865387,
      "learning_rate": 0.00015129113924050632,
      "loss": 0.1721,
      "num_input_tokens_seen": 909380,
      "step": 487,
      "train_runtime": 1015.7416,
      "train_tokens_per_second": 895.287
    },
    {
      "epoch": 0.49330300732878446,
      "grad_norm": 0.6578808426856995,
      "learning_rate": 0.00015118987341772154,
      "loss": 0.3186,
      "num_input_tokens_seen": 911348,
      "step": 488,
      "train_runtime": 1017.5278,
      "train_tokens_per_second": 895.649
    },
    {
      "epoch": 0.4943138741470811,
      "grad_norm": 0.3889693319797516,
      "learning_rate": 0.00015108860759493673,
      "loss": 0.1356,
      "num_input_tokens_seen": 913306,
      "step": 489,
      "train_runtime": 1019.2522,
      "train_tokens_per_second": 896.055
    },
    {
      "epoch": 0.4953247409653778,
      "grad_norm": 0.6702691912651062,
      "learning_rate": 0.0001509873417721519,
      "loss": 0.1721,
      "num_input_tokens_seen": 915058,
      "step": 490,
      "train_runtime": 1020.9492,
      "train_tokens_per_second": 896.282
    },
    {
      "epoch": 0.4963356077836745,
      "grad_norm": 0.4354226589202881,
      "learning_rate": 0.0001508860759493671,
      "loss": 0.1891,
      "num_input_tokens_seen": 916702,
      "step": 491,
      "train_runtime": 1022.6346,
      "train_tokens_per_second": 896.412
    },
    {
      "epoch": 0.4973464746019712,
      "grad_norm": 0.5096386671066284,
      "learning_rate": 0.00015078481012658227,
      "loss": 0.1441,
      "num_input_tokens_seen": 918266,
      "step": 492,
      "train_runtime": 1024.3221,
      "train_tokens_per_second": 896.462
    },
    {
      "epoch": 0.4983573414202679,
      "grad_norm": 0.45088422298431396,
      "learning_rate": 0.00015068354430379746,
      "loss": 0.2398,
      "num_input_tokens_seen": 920190,
      "step": 493,
      "train_runtime": 1026.1133,
      "train_tokens_per_second": 896.772
    },
    {
      "epoch": 0.4993682082385646,
      "grad_norm": 0.4552615284919739,
      "learning_rate": 0.00015058227848101268,
      "loss": 0.2524,
      "num_input_tokens_seen": 921704,
      "step": 494,
      "train_runtime": 1027.7338,
      "train_tokens_per_second": 896.831
    },
    {
      "epoch": 0.5003790750568613,
      "grad_norm": 0.4810153543949127,
      "learning_rate": 0.00015048101265822784,
      "loss": 0.2148,
      "num_input_tokens_seen": 923178,
      "step": 495,
      "train_runtime": 1029.3314,
      "train_tokens_per_second": 896.871
    },
    {
      "epoch": 0.501389941875158,
      "grad_norm": 0.4851894974708557,
      "learning_rate": 0.00015037974683544303,
      "loss": 0.3121,
      "num_input_tokens_seen": 925008,
      "step": 496,
      "train_runtime": 1031.111,
      "train_tokens_per_second": 897.098
    },
    {
      "epoch": 0.5024008086934546,
      "grad_norm": 0.4135304391384125,
      "learning_rate": 0.00015027848101265825,
      "loss": 0.2264,
      "num_input_tokens_seen": 926962,
      "step": 497,
      "train_runtime": 1032.8891,
      "train_tokens_per_second": 897.446
    },
    {
      "epoch": 0.5034116755117514,
      "grad_norm": 0.3830285966396332,
      "learning_rate": 0.00015017721518987342,
      "loss": 0.1614,
      "num_input_tokens_seen": 928622,
      "step": 498,
      "train_runtime": 1034.5809,
      "train_tokens_per_second": 897.583
    },
    {
      "epoch": 0.504422542330048,
      "grad_norm": 0.36824968457221985,
      "learning_rate": 0.0001500759493670886,
      "loss": 0.1645,
      "num_input_tokens_seen": 930538,
      "step": 499,
      "train_runtime": 1036.2947,
      "train_tokens_per_second": 897.947
    },
    {
      "epoch": 0.5054334091483447,
      "grad_norm": 0.32907575368881226,
      "learning_rate": 0.0001499746835443038,
      "loss": 0.1928,
      "num_input_tokens_seen": 932480,
      "step": 500,
      "train_runtime": 1038.091,
      "train_tokens_per_second": 898.264
    },
    {
      "epoch": 0.5064442759666414,
      "grad_norm": 0.43304961919784546,
      "learning_rate": 0.000149873417721519,
      "loss": 0.1576,
      "num_input_tokens_seen": 934090,
      "step": 501,
      "train_runtime": 1039.7844,
      "train_tokens_per_second": 898.35
    },
    {
      "epoch": 0.5074551427849381,
      "grad_norm": 0.45645779371261597,
      "learning_rate": 0.00014977215189873418,
      "loss": 0.1772,
      "num_input_tokens_seen": 935766,
      "step": 502,
      "train_runtime": 1041.4731,
      "train_tokens_per_second": 898.502
    },
    {
      "epoch": 0.5084660096032347,
      "grad_norm": 0.38893789052963257,
      "learning_rate": 0.00014967088607594937,
      "loss": 0.1828,
      "num_input_tokens_seen": 937732,
      "step": 503,
      "train_runtime": 1043.2656,
      "train_tokens_per_second": 898.843
    },
    {
      "epoch": 0.5094768764215315,
      "grad_norm": 0.3733377456665039,
      "learning_rate": 0.00014956962025316456,
      "loss": 0.2014,
      "num_input_tokens_seen": 939214,
      "step": 504,
      "train_runtime": 1044.871,
      "train_tokens_per_second": 898.88
    },
    {
      "epoch": 0.5104877432398282,
      "grad_norm": 0.3456538915634155,
      "learning_rate": 0.00014946835443037975,
      "loss": 0.1535,
      "num_input_tokens_seen": 941428,
      "step": 505,
      "train_runtime": 1046.752,
      "train_tokens_per_second": 899.38
    },
    {
      "epoch": 0.5114986100581248,
      "grad_norm": 0.4836580157279968,
      "learning_rate": 0.00014936708860759494,
      "loss": 0.2011,
      "num_input_tokens_seen": 943318,
      "step": 506,
      "train_runtime": 1048.5512,
      "train_tokens_per_second": 899.639
    },
    {
      "epoch": 0.5125094768764216,
      "grad_norm": 0.41217049956321716,
      "learning_rate": 0.00014926582278481013,
      "loss": 0.1694,
      "num_input_tokens_seen": 945338,
      "step": 507,
      "train_runtime": 1050.2881,
      "train_tokens_per_second": 900.075
    },
    {
      "epoch": 0.5135203436947182,
      "grad_norm": 0.38519519567489624,
      "learning_rate": 0.00014916455696202532,
      "loss": 0.1512,
      "num_input_tokens_seen": 947502,
      "step": 508,
      "train_runtime": 1052.1162,
      "train_tokens_per_second": 900.568
    },
    {
      "epoch": 0.5145312105130149,
      "grad_norm": 0.45736628770828247,
      "learning_rate": 0.00014906329113924052,
      "loss": 0.3374,
      "num_input_tokens_seen": 949186,
      "step": 509,
      "train_runtime": 1053.8219,
      "train_tokens_per_second": 900.708
    },
    {
      "epoch": 0.5155420773313116,
      "grad_norm": 0.3352411389350891,
      "learning_rate": 0.0001489620253164557,
      "loss": 0.1256,
      "num_input_tokens_seen": 951300,
      "step": 510,
      "train_runtime": 1055.6844,
      "train_tokens_per_second": 901.122
    },
    {
      "epoch": 0.5165529441496083,
      "grad_norm": 0.36706387996673584,
      "learning_rate": 0.0001488607594936709,
      "loss": 0.1673,
      "num_input_tokens_seen": 953168,
      "step": 511,
      "train_runtime": 1059.1365,
      "train_tokens_per_second": 899.948
    },
    {
      "epoch": 0.5175638109679049,
      "grad_norm": 0.3382917642593384,
      "learning_rate": 0.0001487594936708861,
      "loss": 0.1971,
      "num_input_tokens_seen": 955434,
      "step": 512,
      "train_runtime": 1061.1152,
      "train_tokens_per_second": 900.406
    },
    {
      "epoch": 0.5185746777862017,
      "grad_norm": 0.48139604926109314,
      "learning_rate": 0.00014865822784810128,
      "loss": 0.2688,
      "num_input_tokens_seen": 957034,
      "step": 513,
      "train_runtime": 1062.777,
      "train_tokens_per_second": 900.503
    },
    {
      "epoch": 0.5195855446044984,
      "grad_norm": 0.39226430654525757,
      "learning_rate": 0.00014855696202531647,
      "loss": 0.2352,
      "num_input_tokens_seen": 958986,
      "step": 514,
      "train_runtime": 1064.6216,
      "train_tokens_per_second": 900.776
    },
    {
      "epoch": 0.520596411422795,
      "grad_norm": 0.45412585139274597,
      "learning_rate": 0.00014845569620253166,
      "loss": 0.1738,
      "num_input_tokens_seen": 961100,
      "step": 515,
      "train_runtime": 1066.5289,
      "train_tokens_per_second": 901.148
    },
    {
      "epoch": 0.5216072782410918,
      "grad_norm": 0.38501042127609253,
      "learning_rate": 0.00014835443037974682,
      "loss": 0.1626,
      "num_input_tokens_seen": 963022,
      "step": 516,
      "train_runtime": 1068.2401,
      "train_tokens_per_second": 901.503
    },
    {
      "epoch": 0.5226181450593884,
      "grad_norm": 0.384539932012558,
      "learning_rate": 0.00014825316455696204,
      "loss": 0.2145,
      "num_input_tokens_seen": 964910,
      "step": 517,
      "train_runtime": 1070.0265,
      "train_tokens_per_second": 901.763
    },
    {
      "epoch": 0.5236290118776851,
      "grad_norm": 0.6114786863327026,
      "learning_rate": 0.00014815189873417723,
      "loss": 0.2917,
      "num_input_tokens_seen": 966458,
      "step": 518,
      "train_runtime": 1071.6494,
      "train_tokens_per_second": 901.842
    },
    {
      "epoch": 0.5246398786959818,
      "grad_norm": 0.3773212432861328,
      "learning_rate": 0.0001480506329113924,
      "loss": 0.2422,
      "num_input_tokens_seen": 968034,
      "step": 519,
      "train_runtime": 1073.256,
      "train_tokens_per_second": 901.96
    },
    {
      "epoch": 0.5256507455142785,
      "grad_norm": 0.3456310033798218,
      "learning_rate": 0.00014794936708860761,
      "loss": 0.2735,
      "num_input_tokens_seen": 970254,
      "step": 520,
      "train_runtime": 1075.1839,
      "train_tokens_per_second": 902.408
    },
    {
      "epoch": 0.5266616123325751,
      "grad_norm": 0.47020086646080017,
      "learning_rate": 0.0001478481012658228,
      "loss": 0.1886,
      "num_input_tokens_seen": 971846,
      "step": 521,
      "train_runtime": 1076.8399,
      "train_tokens_per_second": 902.498
    },
    {
      "epoch": 0.5276724791508719,
      "grad_norm": 0.4325953722000122,
      "learning_rate": 0.00014774683544303797,
      "loss": 0.2944,
      "num_input_tokens_seen": 973748,
      "step": 522,
      "train_runtime": 1078.6153,
      "train_tokens_per_second": 902.776
    },
    {
      "epoch": 0.5286833459691685,
      "grad_norm": 0.47092968225479126,
      "learning_rate": 0.0001476455696202532,
      "loss": 0.2113,
      "num_input_tokens_seen": 975376,
      "step": 523,
      "train_runtime": 1080.3551,
      "train_tokens_per_second": 902.829
    },
    {
      "epoch": 0.5296942127874652,
      "grad_norm": 0.44826313853263855,
      "learning_rate": 0.00014754430379746835,
      "loss": 0.2931,
      "num_input_tokens_seen": 977280,
      "step": 524,
      "train_runtime": 1082.1186,
      "train_tokens_per_second": 903.117
    },
    {
      "epoch": 0.530705079605762,
      "grad_norm": 0.3668995201587677,
      "learning_rate": 0.00014744303797468354,
      "loss": 0.1651,
      "num_input_tokens_seen": 979808,
      "step": 525,
      "train_runtime": 1084.2343,
      "train_tokens_per_second": 903.687
    },
    {
      "epoch": 0.5317159464240586,
      "grad_norm": 0.40047508478164673,
      "learning_rate": 0.00014734177215189876,
      "loss": 0.2185,
      "num_input_tokens_seen": 981994,
      "step": 526,
      "train_runtime": 1086.1076,
      "train_tokens_per_second": 904.141
    },
    {
      "epoch": 0.5327268132423553,
      "grad_norm": 0.2596887946128845,
      "learning_rate": 0.00014724050632911392,
      "loss": 0.176,
      "num_input_tokens_seen": 983680,
      "step": 527,
      "train_runtime": 1087.7293,
      "train_tokens_per_second": 904.343
    },
    {
      "epoch": 0.533737680060652,
      "grad_norm": 0.36929258704185486,
      "learning_rate": 0.00014713924050632911,
      "loss": 0.1758,
      "num_input_tokens_seen": 985244,
      "step": 528,
      "train_runtime": 1089.3627,
      "train_tokens_per_second": 904.422
    },
    {
      "epoch": 0.5347485468789487,
      "grad_norm": 0.4168088138103485,
      "learning_rate": 0.00014703797468354433,
      "loss": 0.1534,
      "num_input_tokens_seen": 987174,
      "step": 529,
      "train_runtime": 1091.1598,
      "train_tokens_per_second": 904.702
    },
    {
      "epoch": 0.5357594136972453,
      "grad_norm": 0.43357694149017334,
      "learning_rate": 0.0001469367088607595,
      "loss": 0.2082,
      "num_input_tokens_seen": 988916,
      "step": 530,
      "train_runtime": 1092.8046,
      "train_tokens_per_second": 904.934
    },
    {
      "epoch": 0.5367702805155421,
      "grad_norm": 0.4053344428539276,
      "learning_rate": 0.0001468354430379747,
      "loss": 0.1625,
      "num_input_tokens_seen": 990246,
      "step": 531,
      "train_runtime": 1094.3918,
      "train_tokens_per_second": 904.837
    },
    {
      "epoch": 0.5377811473338387,
      "grad_norm": 0.3580296039581299,
      "learning_rate": 0.00014673417721518988,
      "loss": 0.2563,
      "num_input_tokens_seen": 992026,
      "step": 532,
      "train_runtime": 1096.0996,
      "train_tokens_per_second": 905.051
    },
    {
      "epoch": 0.5387920141521354,
      "grad_norm": 0.38738858699798584,
      "learning_rate": 0.00014663291139240507,
      "loss": 0.1965,
      "num_input_tokens_seen": 993518,
      "step": 533,
      "train_runtime": 1097.7914,
      "train_tokens_per_second": 905.015
    },
    {
      "epoch": 0.5398028809704322,
      "grad_norm": 0.4551422894001007,
      "learning_rate": 0.00014653164556962026,
      "loss": 0.3162,
      "num_input_tokens_seen": 995702,
      "step": 534,
      "train_runtime": 1099.6768,
      "train_tokens_per_second": 905.45
    },
    {
      "epoch": 0.5408137477887288,
      "grad_norm": 0.37893468141555786,
      "learning_rate": 0.00014643037974683545,
      "loss": 0.1811,
      "num_input_tokens_seen": 997600,
      "step": 535,
      "train_runtime": 1101.4888,
      "train_tokens_per_second": 905.683
    },
    {
      "epoch": 0.5418246146070256,
      "grad_norm": 0.4151322543621063,
      "learning_rate": 0.00014632911392405064,
      "loss": 0.1891,
      "num_input_tokens_seen": 999420,
      "step": 536,
      "train_runtime": 1103.1944,
      "train_tokens_per_second": 905.933
    },
    {
      "epoch": 0.5428354814253222,
      "grad_norm": 0.4174366295337677,
      "learning_rate": 0.00014622784810126583,
      "loss": 0.1765,
      "num_input_tokens_seen": 1001196,
      "step": 537,
      "train_runtime": 1104.9331,
      "train_tokens_per_second": 906.115
    },
    {
      "epoch": 0.5438463482436189,
      "grad_norm": 0.29842501878738403,
      "learning_rate": 0.00014612658227848102,
      "loss": 0.2105,
      "num_input_tokens_seen": 1003184,
      "step": 538,
      "train_runtime": 1106.726,
      "train_tokens_per_second": 906.443
    },
    {
      "epoch": 0.5448572150619156,
      "grad_norm": 0.4382702112197876,
      "learning_rate": 0.00014602531645569621,
      "loss": 0.2244,
      "num_input_tokens_seen": 1005212,
      "step": 539,
      "train_runtime": 1108.5599,
      "train_tokens_per_second": 906.773
    },
    {
      "epoch": 0.5458680818802123,
      "grad_norm": 0.5053007006645203,
      "learning_rate": 0.00014592405063291138,
      "loss": 0.2209,
      "num_input_tokens_seen": 1007078,
      "step": 540,
      "train_runtime": 1110.3033,
      "train_tokens_per_second": 907.03
    },
    {
      "epoch": 0.546878948698509,
      "grad_norm": 0.4378385543823242,
      "learning_rate": 0.0001458227848101266,
      "loss": 0.1674,
      "num_input_tokens_seen": 1008832,
      "step": 541,
      "train_runtime": 1113.699,
      "train_tokens_per_second": 905.839
    },
    {
      "epoch": 0.5478898155168057,
      "grad_norm": 0.31743279099464417,
      "learning_rate": 0.00014572151898734179,
      "loss": 0.2119,
      "num_input_tokens_seen": 1011414,
      "step": 542,
      "train_runtime": 1115.7944,
      "train_tokens_per_second": 906.452
    },
    {
      "epoch": 0.5489006823351024,
      "grad_norm": 0.4299280643463135,
      "learning_rate": 0.00014562025316455695,
      "loss": 0.2422,
      "num_input_tokens_seen": 1013356,
      "step": 543,
      "train_runtime": 1117.5224,
      "train_tokens_per_second": 906.788
    },
    {
      "epoch": 0.549911549153399,
      "grad_norm": 0.37672558426856995,
      "learning_rate": 0.00014551898734177217,
      "loss": 0.1441,
      "num_input_tokens_seen": 1015100,
      "step": 544,
      "train_runtime": 1119.2239,
      "train_tokens_per_second": 906.968
    },
    {
      "epoch": 0.5509224159716958,
      "grad_norm": 0.4549933969974518,
      "learning_rate": 0.00014541772151898733,
      "loss": 0.2114,
      "num_input_tokens_seen": 1017160,
      "step": 545,
      "train_runtime": 1121.0675,
      "train_tokens_per_second": 907.314
    },
    {
      "epoch": 0.5519332827899924,
      "grad_norm": 0.3631333112716675,
      "learning_rate": 0.00014531645569620252,
      "loss": 0.1708,
      "num_input_tokens_seen": 1019486,
      "step": 546,
      "train_runtime": 1123.0554,
      "train_tokens_per_second": 907.779
    },
    {
      "epoch": 0.5529441496082891,
      "grad_norm": 0.38967546820640564,
      "learning_rate": 0.00014521518987341774,
      "loss": 0.1762,
      "num_input_tokens_seen": 1021352,
      "step": 547,
      "train_runtime": 1124.7761,
      "train_tokens_per_second": 908.049
    },
    {
      "epoch": 0.5539550164265858,
      "grad_norm": 0.41608816385269165,
      "learning_rate": 0.0001451139240506329,
      "loss": 0.2531,
      "num_input_tokens_seen": 1023204,
      "step": 548,
      "train_runtime": 1126.5952,
      "train_tokens_per_second": 908.227
    },
    {
      "epoch": 0.5549658832448825,
      "grad_norm": 0.38755542039871216,
      "learning_rate": 0.0001450126582278481,
      "loss": 0.1732,
      "num_input_tokens_seen": 1025652,
      "step": 549,
      "train_runtime": 1128.604,
      "train_tokens_per_second": 908.779
    },
    {
      "epoch": 0.5559767500631791,
      "grad_norm": 0.528978168964386,
      "learning_rate": 0.0001449113924050633,
      "loss": 0.2496,
      "num_input_tokens_seen": 1027620,
      "step": 550,
      "train_runtime": 1130.3782,
      "train_tokens_per_second": 909.094
    },
    {
      "epoch": 0.5569876168814759,
      "grad_norm": 0.422991544008255,
      "learning_rate": 0.00014481012658227848,
      "loss": 0.1859,
      "num_input_tokens_seen": 1029384,
      "step": 551,
      "train_runtime": 1132.146,
      "train_tokens_per_second": 909.233
    },
    {
      "epoch": 0.5579984836997726,
      "grad_norm": 0.3693796396255493,
      "learning_rate": 0.00014470886075949367,
      "loss": 0.183,
      "num_input_tokens_seen": 1031376,
      "step": 552,
      "train_runtime": 1133.9194,
      "train_tokens_per_second": 909.567
    },
    {
      "epoch": 0.5590093505180692,
      "grad_norm": 0.4130207598209381,
      "learning_rate": 0.00014460759493670886,
      "loss": 0.1545,
      "num_input_tokens_seen": 1033068,
      "step": 553,
      "train_runtime": 1135.6172,
      "train_tokens_per_second": 909.697
    },
    {
      "epoch": 0.560020217336366,
      "grad_norm": 0.4068583548069,
      "learning_rate": 0.00014450632911392405,
      "loss": 0.1753,
      "num_input_tokens_seen": 1034514,
      "step": 554,
      "train_runtime": 1137.2561,
      "train_tokens_per_second": 909.658
    },
    {
      "epoch": 0.5610310841546626,
      "grad_norm": 0.37618324160575867,
      "learning_rate": 0.00014440506329113924,
      "loss": 0.1619,
      "num_input_tokens_seen": 1036280,
      "step": 555,
      "train_runtime": 1138.9987,
      "train_tokens_per_second": 909.817
    },
    {
      "epoch": 0.5620419509729593,
      "grad_norm": 0.5306529402732849,
      "learning_rate": 0.00014430379746835443,
      "loss": 0.1499,
      "num_input_tokens_seen": 1038124,
      "step": 556,
      "train_runtime": 1140.7426,
      "train_tokens_per_second": 910.042
    },
    {
      "epoch": 0.563052817791256,
      "grad_norm": 0.41581252217292786,
      "learning_rate": 0.00014420253164556962,
      "loss": 0.2099,
      "num_input_tokens_seen": 1039450,
      "step": 557,
      "train_runtime": 1142.3274,
      "train_tokens_per_second": 909.941
    },
    {
      "epoch": 0.5640636846095527,
      "grad_norm": 0.4545428454875946,
      "learning_rate": 0.0001441012658227848,
      "loss": 0.2257,
      "num_input_tokens_seen": 1041268,
      "step": 558,
      "train_runtime": 1144.031,
      "train_tokens_per_second": 910.175
    },
    {
      "epoch": 0.5650745514278493,
      "grad_norm": 0.4114132821559906,
      "learning_rate": 0.000144,
      "loss": 0.1979,
      "num_input_tokens_seen": 1043102,
      "step": 559,
      "train_runtime": 1145.7788,
      "train_tokens_per_second": 910.387
    },
    {
      "epoch": 0.5660854182461461,
      "grad_norm": 0.4922008216381073,
      "learning_rate": 0.0001438987341772152,
      "loss": 0.1924,
      "num_input_tokens_seen": 1044816,
      "step": 560,
      "train_runtime": 1147.4845,
      "train_tokens_per_second": 910.527
    },
    {
      "epoch": 0.5670962850644428,
      "grad_norm": 0.390349417924881,
      "learning_rate": 0.00014379746835443039,
      "loss": 0.1824,
      "num_input_tokens_seen": 1046774,
      "step": 561,
      "train_runtime": 1149.2201,
      "train_tokens_per_second": 910.856
    },
    {
      "epoch": 0.5681071518827394,
      "grad_norm": 0.5121869444847107,
      "learning_rate": 0.00014369620253164558,
      "loss": 0.2223,
      "num_input_tokens_seen": 1048378,
      "step": 562,
      "train_runtime": 1150.8973,
      "train_tokens_per_second": 910.922
    },
    {
      "epoch": 0.5691180187010362,
      "grad_norm": 0.3791031241416931,
      "learning_rate": 0.00014359493670886077,
      "loss": 0.1767,
      "num_input_tokens_seen": 1050576,
      "step": 563,
      "train_runtime": 1152.8289,
      "train_tokens_per_second": 911.303
    },
    {
      "epoch": 0.5701288855193328,
      "grad_norm": 0.4632573425769806,
      "learning_rate": 0.00014349367088607596,
      "loss": 0.16,
      "num_input_tokens_seen": 1052246,
      "step": 564,
      "train_runtime": 1154.5757,
      "train_tokens_per_second": 911.37
    },
    {
      "epoch": 0.5711397523376295,
      "grad_norm": 0.5109297037124634,
      "learning_rate": 0.00014339240506329115,
      "loss": 0.1945,
      "num_input_tokens_seen": 1053862,
      "step": 565,
      "train_runtime": 1156.1996,
      "train_tokens_per_second": 911.488
    },
    {
      "epoch": 0.5721506191559262,
      "grad_norm": 0.4783368706703186,
      "learning_rate": 0.00014329113924050634,
      "loss": 0.2149,
      "num_input_tokens_seen": 1055614,
      "step": 566,
      "train_runtime": 1157.8965,
      "train_tokens_per_second": 911.665
    },
    {
      "epoch": 0.5731614859742229,
      "grad_norm": 0.3857126533985138,
      "learning_rate": 0.00014318987341772153,
      "loss": 0.2684,
      "num_input_tokens_seen": 1057856,
      "step": 567,
      "train_runtime": 1159.8785,
      "train_tokens_per_second": 912.04
    },
    {
      "epoch": 0.5741723527925195,
      "grad_norm": 0.49782082438468933,
      "learning_rate": 0.00014308860759493672,
      "loss": 0.2032,
      "num_input_tokens_seen": 1059884,
      "step": 568,
      "train_runtime": 1161.7286,
      "train_tokens_per_second": 912.334
    },
    {
      "epoch": 0.5751832196108163,
      "grad_norm": 0.4673866927623749,
      "learning_rate": 0.00014298734177215188,
      "loss": 0.1962,
      "num_input_tokens_seen": 1061424,
      "step": 569,
      "train_runtime": 1163.3897,
      "train_tokens_per_second": 912.355
    },
    {
      "epoch": 0.576194086429113,
      "grad_norm": 0.360312283039093,
      "learning_rate": 0.0001428860759493671,
      "loss": 0.2023,
      "num_input_tokens_seen": 1063634,
      "step": 570,
      "train_runtime": 1165.2763,
      "train_tokens_per_second": 912.774
    },
    {
      "epoch": 0.5772049532474096,
      "grad_norm": 0.371200293302536,
      "learning_rate": 0.0001427848101265823,
      "loss": 0.2356,
      "num_input_tokens_seen": 1066160,
      "step": 571,
      "train_runtime": 1169.0693,
      "train_tokens_per_second": 911.973
    },
    {
      "epoch": 0.5782158200657064,
      "grad_norm": 0.4121630787849426,
      "learning_rate": 0.00014268354430379746,
      "loss": 0.1596,
      "num_input_tokens_seen": 1068052,
      "step": 572,
      "train_runtime": 1170.8747,
      "train_tokens_per_second": 912.183
    },
    {
      "epoch": 0.579226686884003,
      "grad_norm": 0.5687708258628845,
      "learning_rate": 0.00014258227848101268,
      "loss": 0.3351,
      "num_input_tokens_seen": 1069696,
      "step": 573,
      "train_runtime": 1172.4881,
      "train_tokens_per_second": 912.33
    },
    {
      "epoch": 0.5802375537022997,
      "grad_norm": 0.2903662323951721,
      "learning_rate": 0.00014248101265822787,
      "loss": 0.1987,
      "num_input_tokens_seen": 1071820,
      "step": 574,
      "train_runtime": 1174.2988,
      "train_tokens_per_second": 912.732
    },
    {
      "epoch": 0.5812484205205964,
      "grad_norm": 0.306836873292923,
      "learning_rate": 0.00014237974683544303,
      "loss": 0.1047,
      "num_input_tokens_seen": 1074216,
      "step": 575,
      "train_runtime": 1176.2553,
      "train_tokens_per_second": 913.251
    },
    {
      "epoch": 0.5822592873388931,
      "grad_norm": 0.4702828526496887,
      "learning_rate": 0.00014227848101265825,
      "loss": 0.3093,
      "num_input_tokens_seen": 1075822,
      "step": 576,
      "train_runtime": 1177.9242,
      "train_tokens_per_second": 913.32
    },
    {
      "epoch": 0.5832701541571897,
      "grad_norm": 0.47218936681747437,
      "learning_rate": 0.0001421772151898734,
      "loss": 0.2416,
      "num_input_tokens_seen": 1078104,
      "step": 577,
      "train_runtime": 1179.8512,
      "train_tokens_per_second": 913.763
    },
    {
      "epoch": 0.5842810209754865,
      "grad_norm": 0.331085205078125,
      "learning_rate": 0.0001420759493670886,
      "loss": 0.2187,
      "num_input_tokens_seen": 1079748,
      "step": 578,
      "train_runtime": 1181.5279,
      "train_tokens_per_second": 913.857
    },
    {
      "epoch": 0.5852918877937832,
      "grad_norm": 0.4153466820716858,
      "learning_rate": 0.00014197468354430382,
      "loss": 0.2032,
      "num_input_tokens_seen": 1081338,
      "step": 579,
      "train_runtime": 1183.1568,
      "train_tokens_per_second": 913.943
    },
    {
      "epoch": 0.5863027546120798,
      "grad_norm": 0.3859248757362366,
      "learning_rate": 0.00014187341772151898,
      "loss": 0.203,
      "num_input_tokens_seen": 1083466,
      "step": 580,
      "train_runtime": 1185.0429,
      "train_tokens_per_second": 914.284
    },
    {
      "epoch": 0.5873136214303766,
      "grad_norm": 0.463846892118454,
      "learning_rate": 0.00014177215189873418,
      "loss": 0.2193,
      "num_input_tokens_seen": 1085340,
      "step": 581,
      "train_runtime": 1186.765,
      "train_tokens_per_second": 914.537
    },
    {
      "epoch": 0.5883244882486732,
      "grad_norm": 0.3924486041069031,
      "learning_rate": 0.0001416708860759494,
      "loss": 0.2138,
      "num_input_tokens_seen": 1087066,
      "step": 582,
      "train_runtime": 1188.3924,
      "train_tokens_per_second": 914.737
    },
    {
      "epoch": 0.5893353550669699,
      "grad_norm": 0.3305489122867584,
      "learning_rate": 0.00014156962025316456,
      "loss": 0.225,
      "num_input_tokens_seen": 1089186,
      "step": 583,
      "train_runtime": 1190.2641,
      "train_tokens_per_second": 915.079
    },
    {
      "epoch": 0.5903462218852666,
      "grad_norm": 0.3427446484565735,
      "learning_rate": 0.00014146835443037975,
      "loss": 0.3415,
      "num_input_tokens_seen": 1091036,
      "step": 584,
      "train_runtime": 1192.0466,
      "train_tokens_per_second": 915.263
    },
    {
      "epoch": 0.5913570887035633,
      "grad_norm": 0.47477230429649353,
      "learning_rate": 0.00014136708860759494,
      "loss": 0.322,
      "num_input_tokens_seen": 1092928,
      "step": 585,
      "train_runtime": 1193.7624,
      "train_tokens_per_second": 915.532
    },
    {
      "epoch": 0.59236795552186,
      "grad_norm": 0.32592275738716125,
      "learning_rate": 0.00014126582278481013,
      "loss": 0.2132,
      "num_input_tokens_seen": 1094898,
      "step": 586,
      "train_runtime": 1195.5584,
      "train_tokens_per_second": 915.805
    },
    {
      "epoch": 0.5933788223401567,
      "grad_norm": 0.45266860723495483,
      "learning_rate": 0.00014116455696202532,
      "loss": 0.2134,
      "num_input_tokens_seen": 1096502,
      "step": 587,
      "train_runtime": 1197.2439,
      "train_tokens_per_second": 915.855
    },
    {
      "epoch": 0.5943896891584534,
      "grad_norm": 0.4829668402671814,
      "learning_rate": 0.0001410632911392405,
      "loss": 0.1822,
      "num_input_tokens_seen": 1097890,
      "step": 588,
      "train_runtime": 1198.8313,
      "train_tokens_per_second": 915.8
    },
    {
      "epoch": 0.59540055597675,
      "grad_norm": 0.46501827239990234,
      "learning_rate": 0.0001409620253164557,
      "loss": 0.1881,
      "num_input_tokens_seen": 1099700,
      "step": 589,
      "train_runtime": 1200.5457,
      "train_tokens_per_second": 916.0
    },
    {
      "epoch": 0.5964114227950468,
      "grad_norm": 0.3549115061759949,
      "learning_rate": 0.0001408607594936709,
      "loss": 0.2193,
      "num_input_tokens_seen": 1102062,
      "step": 590,
      "train_runtime": 1202.555,
      "train_tokens_per_second": 916.434
    },
    {
      "epoch": 0.5974222896133434,
      "grad_norm": 0.520986795425415,
      "learning_rate": 0.00014075949367088608,
      "loss": 0.2858,
      "num_input_tokens_seen": 1103904,
      "step": 591,
      "train_runtime": 1204.3327,
      "train_tokens_per_second": 916.611
    },
    {
      "epoch": 0.5984331564316401,
      "grad_norm": 0.5083928108215332,
      "learning_rate": 0.00014065822784810127,
      "loss": 0.2111,
      "num_input_tokens_seen": 1105536,
      "step": 592,
      "train_runtime": 1205.9703,
      "train_tokens_per_second": 916.719
    },
    {
      "epoch": 0.5994440232499368,
      "grad_norm": 0.4003691077232361,
      "learning_rate": 0.00014055696202531647,
      "loss": 0.1758,
      "num_input_tokens_seen": 1107264,
      "step": 593,
      "train_runtime": 1207.6622,
      "train_tokens_per_second": 916.866
    },
    {
      "epoch": 0.6004548900682335,
      "grad_norm": 0.3368241786956787,
      "learning_rate": 0.00014045569620253166,
      "loss": 0.1774,
      "num_input_tokens_seen": 1109584,
      "step": 594,
      "train_runtime": 1209.6215,
      "train_tokens_per_second": 917.298
    },
    {
      "epoch": 0.6004548900682335,
      "eval_loss": 0.1999172419309616,
      "eval_runtime": 61.4282,
      "eval_samples_per_second": 14.309,
      "eval_steps_per_second": 7.163,
      "num_input_tokens_seen": 1109584,
      "step": 594
    },
    {
      "epoch": 0.6014657568865301,
      "grad_norm": 0.5142544507980347,
      "learning_rate": 0.00014035443037974685,
      "loss": 0.2682,
      "num_input_tokens_seen": 1111106,
      "step": 595,
      "train_runtime": 1272.7308,
      "train_tokens_per_second": 873.009
    },
    {
      "epoch": 0.6024766237048269,
      "grad_norm": 0.42380717396736145,
      "learning_rate": 0.00014025316455696204,
      "loss": 0.1566,
      "num_input_tokens_seen": 1113288,
      "step": 596,
      "train_runtime": 1274.5414,
      "train_tokens_per_second": 873.481
    },
    {
      "epoch": 0.6034874905231236,
      "grad_norm": 0.452054888010025,
      "learning_rate": 0.00014015189873417723,
      "loss": 0.1474,
      "num_input_tokens_seen": 1115252,
      "step": 597,
      "train_runtime": 1276.3727,
      "train_tokens_per_second": 873.767
    },
    {
      "epoch": 0.6044983573414203,
      "grad_norm": 0.535305380821228,
      "learning_rate": 0.00014005063291139242,
      "loss": 0.2835,
      "num_input_tokens_seen": 1116960,
      "step": 598,
      "train_runtime": 1278.0142,
      "train_tokens_per_second": 873.981
    },
    {
      "epoch": 0.605509224159717,
      "grad_norm": 0.5015350580215454,
      "learning_rate": 0.0001399493670886076,
      "loss": 0.1981,
      "num_input_tokens_seen": 1118466,
      "step": 599,
      "train_runtime": 1279.5965,
      "train_tokens_per_second": 874.077
    },
    {
      "epoch": 0.6065200909780136,
      "grad_norm": 0.44699451327323914,
      "learning_rate": 0.0001398481012658228,
      "loss": 0.2412,
      "num_input_tokens_seen": 1120110,
      "step": 600,
      "train_runtime": 1281.2265,
      "train_tokens_per_second": 874.248
    },
    {
      "epoch": 0.6075309577963104,
      "grad_norm": 0.42712441086769104,
      "learning_rate": 0.00013974683544303796,
      "loss": 0.2404,
      "num_input_tokens_seen": 1122134,
      "step": 601,
      "train_runtime": 1285.133,
      "train_tokens_per_second": 873.166
    },
    {
      "epoch": 0.608541824614607,
      "grad_norm": 0.5717500448226929,
      "learning_rate": 0.00013964556962025318,
      "loss": 0.2623,
      "num_input_tokens_seen": 1123944,
      "step": 602,
      "train_runtime": 1286.9072,
      "train_tokens_per_second": 873.368
    },
    {
      "epoch": 0.6095526914329037,
      "grad_norm": 0.4131810963153839,
      "learning_rate": 0.00013954430379746837,
      "loss": 0.1334,
      "num_input_tokens_seen": 1125774,
      "step": 603,
      "train_runtime": 1288.6145,
      "train_tokens_per_second": 873.631
    },
    {
      "epoch": 0.6105635582512005,
      "grad_norm": 0.42398935556411743,
      "learning_rate": 0.00013944303797468354,
      "loss": 0.1776,
      "num_input_tokens_seen": 1127612,
      "step": 604,
      "train_runtime": 1290.4229,
      "train_tokens_per_second": 873.831
    },
    {
      "epoch": 0.6115744250694971,
      "grad_norm": 0.47160011529922485,
      "learning_rate": 0.00013934177215189876,
      "loss": 0.2297,
      "num_input_tokens_seen": 1129478,
      "step": 605,
      "train_runtime": 1292.2006,
      "train_tokens_per_second": 874.073
    },
    {
      "epoch": 0.6125852918877938,
      "grad_norm": 0.3654675781726837,
      "learning_rate": 0.00013924050632911395,
      "loss": 0.1374,
      "num_input_tokens_seen": 1131272,
      "step": 606,
      "train_runtime": 1293.841,
      "train_tokens_per_second": 874.352
    },
    {
      "epoch": 0.6135961587060905,
      "grad_norm": 0.42554304003715515,
      "learning_rate": 0.0001391392405063291,
      "loss": 0.2403,
      "num_input_tokens_seen": 1133002,
      "step": 607,
      "train_runtime": 1295.582,
      "train_tokens_per_second": 874.512
    },
    {
      "epoch": 0.6146070255243872,
      "grad_norm": 0.38294222950935364,
      "learning_rate": 0.0001390379746835443,
      "loss": 0.2169,
      "num_input_tokens_seen": 1135134,
      "step": 608,
      "train_runtime": 1297.3831,
      "train_tokens_per_second": 874.941
    },
    {
      "epoch": 0.6156178923426838,
      "grad_norm": 0.419821172952652,
      "learning_rate": 0.0001389367088607595,
      "loss": 0.1488,
      "num_input_tokens_seen": 1136798,
      "step": 609,
      "train_runtime": 1299.0132,
      "train_tokens_per_second": 875.124
    },
    {
      "epoch": 0.6166287591609806,
      "grad_norm": 0.4196281433105469,
      "learning_rate": 0.00013883544303797468,
      "loss": 0.15,
      "num_input_tokens_seen": 1138694,
      "step": 610,
      "train_runtime": 1300.7687,
      "train_tokens_per_second": 875.401
    },
    {
      "epoch": 0.6176396259792772,
      "grad_norm": 0.4162880778312683,
      "learning_rate": 0.00013873417721518987,
      "loss": 0.1656,
      "num_input_tokens_seen": 1140184,
      "step": 611,
      "train_runtime": 1302.4433,
      "train_tokens_per_second": 875.419
    },
    {
      "epoch": 0.6186504927975739,
      "grad_norm": 0.47770410776138306,
      "learning_rate": 0.00013863291139240506,
      "loss": 0.3258,
      "num_input_tokens_seen": 1142572,
      "step": 612,
      "train_runtime": 1304.386,
      "train_tokens_per_second": 875.946
    },
    {
      "epoch": 0.6196613596158707,
      "grad_norm": 0.49336400628089905,
      "learning_rate": 0.00013853164556962025,
      "loss": 0.2225,
      "num_input_tokens_seen": 1144346,
      "step": 613,
      "train_runtime": 1306.0177,
      "train_tokens_per_second": 876.21
    },
    {
      "epoch": 0.6206722264341673,
      "grad_norm": 0.45543521642684937,
      "learning_rate": 0.00013843037974683545,
      "loss": 0.1773,
      "num_input_tokens_seen": 1145922,
      "step": 614,
      "train_runtime": 1307.6453,
      "train_tokens_per_second": 876.325
    },
    {
      "epoch": 0.621683093252464,
      "grad_norm": 0.3972449004650116,
      "learning_rate": 0.00013832911392405064,
      "loss": 0.1437,
      "num_input_tokens_seen": 1147932,
      "step": 615,
      "train_runtime": 1309.4836,
      "train_tokens_per_second": 876.63
    },
    {
      "epoch": 0.6226939600707607,
      "grad_norm": 0.5032971501350403,
      "learning_rate": 0.00013822784810126583,
      "loss": 0.2274,
      "num_input_tokens_seen": 1149966,
      "step": 616,
      "train_runtime": 1311.3488,
      "train_tokens_per_second": 876.934
    },
    {
      "epoch": 0.6237048268890574,
      "grad_norm": 0.5544525980949402,
      "learning_rate": 0.00013812658227848102,
      "loss": 0.2774,
      "num_input_tokens_seen": 1151980,
      "step": 617,
      "train_runtime": 1313.1947,
      "train_tokens_per_second": 877.235
    },
    {
      "epoch": 0.624715693707354,
      "grad_norm": 0.4054725170135498,
      "learning_rate": 0.0001380253164556962,
      "loss": 0.1844,
      "num_input_tokens_seen": 1154506,
      "step": 618,
      "train_runtime": 1315.2554,
      "train_tokens_per_second": 877.781
    },
    {
      "epoch": 0.6257265605256508,
      "grad_norm": 0.4948369860649109,
      "learning_rate": 0.0001379240506329114,
      "loss": 0.2104,
      "num_input_tokens_seen": 1156236,
      "step": 619,
      "train_runtime": 1317.0121,
      "train_tokens_per_second": 877.924
    },
    {
      "epoch": 0.6267374273439474,
      "grad_norm": 0.4355620741844177,
      "learning_rate": 0.0001378227848101266,
      "loss": 0.2392,
      "num_input_tokens_seen": 1158250,
      "step": 620,
      "train_runtime": 1318.8694,
      "train_tokens_per_second": 878.214
    },
    {
      "epoch": 0.6277482941622441,
      "grad_norm": 0.49952054023742676,
      "learning_rate": 0.00013772151898734178,
      "loss": 0.257,
      "num_input_tokens_seen": 1159834,
      "step": 621,
      "train_runtime": 1320.4982,
      "train_tokens_per_second": 878.331
    },
    {
      "epoch": 0.6287591609805409,
      "grad_norm": 0.4113228917121887,
      "learning_rate": 0.00013762025316455697,
      "loss": 0.1428,
      "num_input_tokens_seen": 1161566,
      "step": 622,
      "train_runtime": 1322.2029,
      "train_tokens_per_second": 878.508
    },
    {
      "epoch": 0.6297700277988375,
      "grad_norm": 0.40772753953933716,
      "learning_rate": 0.00013751898734177216,
      "loss": 0.2294,
      "num_input_tokens_seen": 1163734,
      "step": 623,
      "train_runtime": 1324.0694,
      "train_tokens_per_second": 878.907
    },
    {
      "epoch": 0.6307808946171342,
      "grad_norm": 0.36806434392929077,
      "learning_rate": 0.00013741772151898735,
      "loss": 0.2122,
      "num_input_tokens_seen": 1165512,
      "step": 624,
      "train_runtime": 1325.8261,
      "train_tokens_per_second": 879.084
    },
    {
      "epoch": 0.6317917614354309,
      "grad_norm": 0.3921598494052887,
      "learning_rate": 0.00013731645569620252,
      "loss": 0.1499,
      "num_input_tokens_seen": 1167786,
      "step": 625,
      "train_runtime": 1327.7854,
      "train_tokens_per_second": 879.499
    },
    {
      "epoch": 0.6328026282537276,
      "grad_norm": 0.5765639543533325,
      "learning_rate": 0.00013721518987341774,
      "loss": 0.1781,
      "num_input_tokens_seen": 1169068,
      "step": 626,
      "train_runtime": 1329.3517,
      "train_tokens_per_second": 879.427
    },
    {
      "epoch": 0.6338134950720242,
      "grad_norm": 0.4361734092235565,
      "learning_rate": 0.00013711392405063293,
      "loss": 0.1977,
      "num_input_tokens_seen": 1170292,
      "step": 627,
      "train_runtime": 1330.9128,
      "train_tokens_per_second": 879.315
    },
    {
      "epoch": 0.634824361890321,
      "grad_norm": 0.4898892045021057,
      "learning_rate": 0.0001370126582278481,
      "loss": 0.1883,
      "num_input_tokens_seen": 1172330,
      "step": 628,
      "train_runtime": 1332.7698,
      "train_tokens_per_second": 879.619
    },
    {
      "epoch": 0.6358352287086176,
      "grad_norm": 0.39373868703842163,
      "learning_rate": 0.0001369113924050633,
      "loss": 0.2619,
      "num_input_tokens_seen": 1174324,
      "step": 629,
      "train_runtime": 1334.4959,
      "train_tokens_per_second": 879.976
    },
    {
      "epoch": 0.6368460955269143,
      "grad_norm": 0.5396379828453064,
      "learning_rate": 0.0001368101265822785,
      "loss": 0.2839,
      "num_input_tokens_seen": 1176344,
      "step": 630,
      "train_runtime": 1336.2338,
      "train_tokens_per_second": 880.343
    },
    {
      "epoch": 0.637856962345211,
      "grad_norm": 0.47252243757247925,
      "learning_rate": 0.00013670886075949366,
      "loss": 0.1488,
      "num_input_tokens_seen": 1178350,
      "step": 631,
      "train_runtime": 1340.0088,
      "train_tokens_per_second": 879.36
    },
    {
      "epoch": 0.6388678291635077,
      "grad_norm": 0.45499494671821594,
      "learning_rate": 0.00013660759493670888,
      "loss": 0.2265,
      "num_input_tokens_seen": 1180318,
      "step": 632,
      "train_runtime": 1341.8151,
      "train_tokens_per_second": 879.643
    },
    {
      "epoch": 0.6398786959818044,
      "grad_norm": 0.45699816942214966,
      "learning_rate": 0.00013650632911392404,
      "loss": 0.2249,
      "num_input_tokens_seen": 1182472,
      "step": 633,
      "train_runtime": 1343.6785,
      "train_tokens_per_second": 880.026
    },
    {
      "epoch": 0.6408895628001011,
      "grad_norm": 0.4991965591907501,
      "learning_rate": 0.00013640506329113924,
      "loss": 0.2452,
      "num_input_tokens_seen": 1184080,
      "step": 634,
      "train_runtime": 1345.3167,
      "train_tokens_per_second": 880.15
    },
    {
      "epoch": 0.6419004296183978,
      "grad_norm": 0.4307124614715576,
      "learning_rate": 0.00013630379746835445,
      "loss": 0.1796,
      "num_input_tokens_seen": 1186314,
      "step": 635,
      "train_runtime": 1347.2777,
      "train_tokens_per_second": 880.527
    },
    {
      "epoch": 0.6429112964366944,
      "grad_norm": 0.41934290528297424,
      "learning_rate": 0.00013620253164556962,
      "loss": 0.2379,
      "num_input_tokens_seen": 1188320,
      "step": 636,
      "train_runtime": 1349.0601,
      "train_tokens_per_second": 880.85
    },
    {
      "epoch": 0.6439221632549912,
      "grad_norm": 0.3485095798969269,
      "learning_rate": 0.0001361012658227848,
      "loss": 0.2734,
      "num_input_tokens_seen": 1189978,
      "step": 637,
      "train_runtime": 1350.7358,
      "train_tokens_per_second": 880.985
    },
    {
      "epoch": 0.6449330300732878,
      "grad_norm": 0.4292319715023041,
      "learning_rate": 0.00013600000000000003,
      "loss": 0.158,
      "num_input_tokens_seen": 1191680,
      "step": 638,
      "train_runtime": 1352.3584,
      "train_tokens_per_second": 881.187
    },
    {
      "epoch": 0.6459438968915845,
      "grad_norm": 0.3399895131587982,
      "learning_rate": 0.0001358987341772152,
      "loss": 0.1869,
      "num_input_tokens_seen": 1194270,
      "step": 639,
      "train_runtime": 1354.4558,
      "train_tokens_per_second": 881.734
    },
    {
      "epoch": 0.6469547637098813,
      "grad_norm": 0.4121042490005493,
      "learning_rate": 0.00013579746835443038,
      "loss": 0.1869,
      "num_input_tokens_seen": 1196906,
      "step": 640,
      "train_runtime": 1356.5211,
      "train_tokens_per_second": 882.335
    },
    {
      "epoch": 0.6479656305281779,
      "grad_norm": 0.4975435137748718,
      "learning_rate": 0.00013569620253164557,
      "loss": 0.2162,
      "num_input_tokens_seen": 1198600,
      "step": 641,
      "train_runtime": 1358.246,
      "train_tokens_per_second": 882.462
    },
    {
      "epoch": 0.6489764973464746,
      "grad_norm": 0.5656388998031616,
      "learning_rate": 0.00013559493670886076,
      "loss": 0.3482,
      "num_input_tokens_seen": 1200188,
      "step": 642,
      "train_runtime": 1359.8671,
      "train_tokens_per_second": 882.577
    },
    {
      "epoch": 0.6499873641647713,
      "grad_norm": 0.45611894130706787,
      "learning_rate": 0.00013549367088607595,
      "loss": 0.1694,
      "num_input_tokens_seen": 1202518,
      "step": 643,
      "train_runtime": 1361.8079,
      "train_tokens_per_second": 883.031
    },
    {
      "epoch": 0.650998230983068,
      "grad_norm": 0.5768261551856995,
      "learning_rate": 0.00013539240506329114,
      "loss": 0.2855,
      "num_input_tokens_seen": 1204106,
      "step": 644,
      "train_runtime": 1363.4386,
      "train_tokens_per_second": 883.139
    },
    {
      "epoch": 0.6520090978013646,
      "grad_norm": 0.4159179627895355,
      "learning_rate": 0.00013529113924050633,
      "loss": 0.2193,
      "num_input_tokens_seen": 1206452,
      "step": 645,
      "train_runtime": 1365.3405,
      "train_tokens_per_second": 883.627
    },
    {
      "epoch": 0.6530199646196614,
      "grad_norm": 0.4470064043998718,
      "learning_rate": 0.00013518987341772153,
      "loss": 0.2135,
      "num_input_tokens_seen": 1208614,
      "step": 646,
      "train_runtime": 1367.2625,
      "train_tokens_per_second": 883.966
    },
    {
      "epoch": 0.654030831437958,
      "grad_norm": 0.41663551330566406,
      "learning_rate": 0.00013508860759493672,
      "loss": 0.2353,
      "num_input_tokens_seen": 1209990,
      "step": 647,
      "train_runtime": 1368.861,
      "train_tokens_per_second": 883.939
    },
    {
      "epoch": 0.6550416982562547,
      "grad_norm": 0.3976331651210785,
      "learning_rate": 0.0001349873417721519,
      "loss": 0.1853,
      "num_input_tokens_seen": 1211572,
      "step": 648,
      "train_runtime": 1370.5441,
      "train_tokens_per_second": 884.008
    },
    {
      "epoch": 0.6560525650745515,
      "grad_norm": 0.3913930356502533,
      "learning_rate": 0.0001348860759493671,
      "loss": 0.1631,
      "num_input_tokens_seen": 1213344,
      "step": 649,
      "train_runtime": 1372.2407,
      "train_tokens_per_second": 884.206
    },
    {
      "epoch": 0.6570634318928481,
      "grad_norm": 0.4002862274646759,
      "learning_rate": 0.0001347848101265823,
      "loss": 0.2661,
      "num_input_tokens_seen": 1215340,
      "step": 650,
      "train_runtime": 1374.0117,
      "train_tokens_per_second": 884.519
    },
    {
      "epoch": 0.6580742987111448,
      "grad_norm": 0.3866746723651886,
      "learning_rate": 0.00013468354430379748,
      "loss": 0.2703,
      "num_input_tokens_seen": 1217024,
      "step": 651,
      "train_runtime": 1375.7145,
      "train_tokens_per_second": 884.649
    },
    {
      "epoch": 0.6590851655294415,
      "grad_norm": 0.4973559081554413,
      "learning_rate": 0.00013458227848101267,
      "loss": 0.1676,
      "num_input_tokens_seen": 1218734,
      "step": 652,
      "train_runtime": 1377.3536,
      "train_tokens_per_second": 884.837
    },
    {
      "epoch": 0.6600960323477382,
      "grad_norm": 0.5597884058952332,
      "learning_rate": 0.00013448101265822786,
      "loss": 0.3641,
      "num_input_tokens_seen": 1220580,
      "step": 653,
      "train_runtime": 1379.0588,
      "train_tokens_per_second": 885.082
    },
    {
      "epoch": 0.6611068991660348,
      "grad_norm": 0.46370941400527954,
      "learning_rate": 0.00013437974683544305,
      "loss": 0.2083,
      "num_input_tokens_seen": 1222446,
      "step": 654,
      "train_runtime": 1380.894,
      "train_tokens_per_second": 885.257
    },
    {
      "epoch": 0.6621177659843316,
      "grad_norm": 0.5185815095901489,
      "learning_rate": 0.00013427848101265824,
      "loss": 0.1914,
      "num_input_tokens_seen": 1224226,
      "step": 655,
      "train_runtime": 1382.6009,
      "train_tokens_per_second": 885.451
    },
    {
      "epoch": 0.6631286328026282,
      "grad_norm": 0.452104777097702,
      "learning_rate": 0.00013417721518987343,
      "loss": 0.2125,
      "num_input_tokens_seen": 1226252,
      "step": 656,
      "train_runtime": 1384.3852,
      "train_tokens_per_second": 885.774
    },
    {
      "epoch": 0.6641394996209249,
      "grad_norm": 0.353582501411438,
      "learning_rate": 0.0001340759493670886,
      "loss": 0.1754,
      "num_input_tokens_seen": 1228886,
      "step": 657,
      "train_runtime": 1386.5012,
      "train_tokens_per_second": 886.322
    },
    {
      "epoch": 0.6651503664392217,
      "grad_norm": 0.42549431324005127,
      "learning_rate": 0.00013397468354430382,
      "loss": 0.155,
      "num_input_tokens_seen": 1231158,
      "step": 658,
      "train_runtime": 1388.4567,
      "train_tokens_per_second": 886.71
    },
    {
      "epoch": 0.6661612332575183,
      "grad_norm": 0.4215696454048157,
      "learning_rate": 0.000133873417721519,
      "loss": 0.2138,
      "num_input_tokens_seen": 1232656,
      "step": 659,
      "train_runtime": 1390.1232,
      "train_tokens_per_second": 886.724
    },
    {
      "epoch": 0.6671721000758151,
      "grad_norm": 0.3711097836494446,
      "learning_rate": 0.00013377215189873417,
      "loss": 0.1722,
      "num_input_tokens_seen": 1234186,
      "step": 660,
      "train_runtime": 1391.7882,
      "train_tokens_per_second": 886.763
    },
    {
      "epoch": 0.6681829668941117,
      "grad_norm": 0.4210507869720459,
      "learning_rate": 0.0001336708860759494,
      "loss": 0.1515,
      "num_input_tokens_seen": 1236296,
      "step": 661,
      "train_runtime": 1395.7178,
      "train_tokens_per_second": 885.778
    },
    {
      "epoch": 0.6691938337124084,
      "grad_norm": 0.3783086836338043,
      "learning_rate": 0.00013356962025316458,
      "loss": 0.1955,
      "num_input_tokens_seen": 1238166,
      "step": 662,
      "train_runtime": 1397.438,
      "train_tokens_per_second": 886.026
    },
    {
      "epoch": 0.6702047005307051,
      "grad_norm": 0.4010985195636749,
      "learning_rate": 0.00013346835443037974,
      "loss": 0.2455,
      "num_input_tokens_seen": 1240276,
      "step": 663,
      "train_runtime": 1399.2322,
      "train_tokens_per_second": 886.398
    },
    {
      "epoch": 0.6712155673490018,
      "grad_norm": 0.43554016947746277,
      "learning_rate": 0.00013336708860759496,
      "loss": 0.1425,
      "num_input_tokens_seen": 1241846,
      "step": 664,
      "train_runtime": 1400.9068,
      "train_tokens_per_second": 886.459
    },
    {
      "epoch": 0.6722264341672984,
      "grad_norm": 0.3944196403026581,
      "learning_rate": 0.00013326582278481012,
      "loss": 0.2067,
      "num_input_tokens_seen": 1243514,
      "step": 665,
      "train_runtime": 1402.5471,
      "train_tokens_per_second": 886.611
    },
    {
      "epoch": 0.6732373009855952,
      "grad_norm": 0.4447316825389862,
      "learning_rate": 0.00013316455696202532,
      "loss": 0.2011,
      "num_input_tokens_seen": 1245576,
      "step": 666,
      "train_runtime": 1404.3991,
      "train_tokens_per_second": 886.91
    },
    {
      "epoch": 0.6742481678038919,
      "grad_norm": 0.4452163279056549,
      "learning_rate": 0.00013306329113924053,
      "loss": 0.2443,
      "num_input_tokens_seen": 1247542,
      "step": 667,
      "train_runtime": 1406.2335,
      "train_tokens_per_second": 887.151
    },
    {
      "epoch": 0.6752590346221885,
      "grad_norm": 0.45511099696159363,
      "learning_rate": 0.0001329620253164557,
      "loss": 0.19,
      "num_input_tokens_seen": 1249648,
      "step": 668,
      "train_runtime": 1408.1208,
      "train_tokens_per_second": 887.458
    },
    {
      "epoch": 0.6762699014404853,
      "grad_norm": 0.2935425639152527,
      "learning_rate": 0.0001328607594936709,
      "loss": 0.258,
      "num_input_tokens_seen": 1251566,
      "step": 669,
      "train_runtime": 1409.8916,
      "train_tokens_per_second": 887.704
    },
    {
      "epoch": 0.6772807682587819,
      "grad_norm": 0.4612777531147003,
      "learning_rate": 0.0001327594936708861,
      "loss": 0.2456,
      "num_input_tokens_seen": 1253414,
      "step": 670,
      "train_runtime": 1411.6445,
      "train_tokens_per_second": 887.911
    },
    {
      "epoch": 0.6782916350770786,
      "grad_norm": 0.38125815987586975,
      "learning_rate": 0.00013265822784810127,
      "loss": 0.1427,
      "num_input_tokens_seen": 1255404,
      "step": 671,
      "train_runtime": 1413.484,
      "train_tokens_per_second": 888.163
    },
    {
      "epoch": 0.6793025018953753,
      "grad_norm": 0.4688827395439148,
      "learning_rate": 0.00013255696202531646,
      "loss": 0.2247,
      "num_input_tokens_seen": 1257308,
      "step": 672,
      "train_runtime": 1415.2099,
      "train_tokens_per_second": 888.425
    },
    {
      "epoch": 0.680313368713672,
      "grad_norm": 0.2908744513988495,
      "learning_rate": 0.00013245569620253165,
      "loss": 0.1254,
      "num_input_tokens_seen": 1258656,
      "step": 673,
      "train_runtime": 1416.7941,
      "train_tokens_per_second": 888.383
    },
    {
      "epoch": 0.6813242355319686,
      "grad_norm": 0.43420150876045227,
      "learning_rate": 0.00013235443037974684,
      "loss": 0.2506,
      "num_input_tokens_seen": 1260934,
      "step": 674,
      "train_runtime": 1418.7324,
      "train_tokens_per_second": 888.775
    },
    {
      "epoch": 0.6823351023502654,
      "grad_norm": 0.5073977112770081,
      "learning_rate": 0.00013225316455696203,
      "loss": 0.1734,
      "num_input_tokens_seen": 1262612,
      "step": 675,
      "train_runtime": 1420.4166,
      "train_tokens_per_second": 888.903
    },
    {
      "epoch": 0.683345969168562,
      "grad_norm": 0.5391412973403931,
      "learning_rate": 0.00013215189873417722,
      "loss": 0.2277,
      "num_input_tokens_seen": 1263840,
      "step": 676,
      "train_runtime": 1421.9759,
      "train_tokens_per_second": 888.791
    },
    {
      "epoch": 0.6843568359868587,
      "grad_norm": 0.41563281416893005,
      "learning_rate": 0.00013205063291139241,
      "loss": 0.1413,
      "num_input_tokens_seen": 1266208,
      "step": 677,
      "train_runtime": 1424.0361,
      "train_tokens_per_second": 889.168
    },
    {
      "epoch": 0.6853677028051555,
      "grad_norm": 0.3735334873199463,
      "learning_rate": 0.0001319493670886076,
      "loss": 0.3,
      "num_input_tokens_seen": 1267706,
      "step": 678,
      "train_runtime": 1425.7347,
      "train_tokens_per_second": 889.16
    },
    {
      "epoch": 0.6863785696234521,
      "grad_norm": 0.5222294330596924,
      "learning_rate": 0.0001318481012658228,
      "loss": 0.2447,
      "num_input_tokens_seen": 1269792,
      "step": 679,
      "train_runtime": 1427.5909,
      "train_tokens_per_second": 889.465
    },
    {
      "epoch": 0.6873894364417488,
      "grad_norm": 0.47381308674812317,
      "learning_rate": 0.000131746835443038,
      "loss": 0.2701,
      "num_input_tokens_seen": 1272086,
      "step": 680,
      "train_runtime": 1429.5168,
      "train_tokens_per_second": 889.871
    },
    {
      "epoch": 0.6884003032600455,
      "grad_norm": 0.3976532816886902,
      "learning_rate": 0.00013164556962025315,
      "loss": 0.1387,
      "num_input_tokens_seen": 1273896,
      "step": 681,
      "train_runtime": 1431.2722,
      "train_tokens_per_second": 890.045
    },
    {
      "epoch": 0.6894111700783422,
      "grad_norm": 0.5570082068443298,
      "learning_rate": 0.00013154430379746837,
      "loss": 0.2635,
      "num_input_tokens_seen": 1276032,
      "step": 682,
      "train_runtime": 1433.1965,
      "train_tokens_per_second": 890.34
    },
    {
      "epoch": 0.6904220368966388,
      "grad_norm": 0.5047719478607178,
      "learning_rate": 0.00013144303797468356,
      "loss": 0.3581,
      "num_input_tokens_seen": 1277888,
      "step": 683,
      "train_runtime": 1434.9191,
      "train_tokens_per_second": 890.565
    },
    {
      "epoch": 0.6914329037149356,
      "grad_norm": 0.5290535688400269,
      "learning_rate": 0.00013134177215189872,
      "loss": 0.1986,
      "num_input_tokens_seen": 1279506,
      "step": 684,
      "train_runtime": 1436.5453,
      "train_tokens_per_second": 890.683
    },
    {
      "epoch": 0.6924437705332323,
      "grad_norm": 0.49076324701309204,
      "learning_rate": 0.00013124050632911394,
      "loss": 0.1877,
      "num_input_tokens_seen": 1281218,
      "step": 685,
      "train_runtime": 1438.2277,
      "train_tokens_per_second": 890.831
    },
    {
      "epoch": 0.6934546373515289,
      "grad_norm": 0.4349224269390106,
      "learning_rate": 0.00013113924050632913,
      "loss": 0.1873,
      "num_input_tokens_seen": 1282962,
      "step": 686,
      "train_runtime": 1439.9793,
      "train_tokens_per_second": 890.959
    },
    {
      "epoch": 0.6944655041698257,
      "grad_norm": 0.3889138698577881,
      "learning_rate": 0.0001310379746835443,
      "loss": 0.2006,
      "num_input_tokens_seen": 1284568,
      "step": 687,
      "train_runtime": 1441.6789,
      "train_tokens_per_second": 891.022
    },
    {
      "epoch": 0.6954763709881223,
      "grad_norm": 0.40236109495162964,
      "learning_rate": 0.00013093670886075951,
      "loss": 0.2049,
      "num_input_tokens_seen": 1286224,
      "step": 688,
      "train_runtime": 1443.3211,
      "train_tokens_per_second": 891.156
    },
    {
      "epoch": 0.696487237806419,
      "grad_norm": 0.3458422124385834,
      "learning_rate": 0.00013083544303797468,
      "loss": 0.2041,
      "num_input_tokens_seen": 1288048,
      "step": 689,
      "train_runtime": 1445.0887,
      "train_tokens_per_second": 891.328
    },
    {
      "epoch": 0.6974981046247157,
      "grad_norm": 0.45814043283462524,
      "learning_rate": 0.00013073417721518987,
      "loss": 0.1965,
      "num_input_tokens_seen": 1289712,
      "step": 690,
      "train_runtime": 1446.7832,
      "train_tokens_per_second": 891.434
    },
    {
      "epoch": 0.6985089714430124,
      "grad_norm": 0.36505913734436035,
      "learning_rate": 0.0001306329113924051,
      "loss": 0.1506,
      "num_input_tokens_seen": 1291352,
      "step": 691,
      "train_runtime": 1450.5428,
      "train_tokens_per_second": 890.254
    },
    {
      "epoch": 0.699519838261309,
      "grad_norm": 0.5061760544776917,
      "learning_rate": 0.00013053164556962025,
      "loss": 0.2407,
      "num_input_tokens_seen": 1293044,
      "step": 692,
      "train_runtime": 1452.1997,
      "train_tokens_per_second": 890.404
    },
    {
      "epoch": 0.7005307050796058,
      "grad_norm": 0.36819201707839966,
      "learning_rate": 0.00013043037974683544,
      "loss": 0.4679,
      "num_input_tokens_seen": 1294774,
      "step": 693,
      "train_runtime": 1453.891,
      "train_tokens_per_second": 890.558
    },
    {
      "epoch": 0.7015415718979024,
      "grad_norm": 0.3914896249771118,
      "learning_rate": 0.00013032911392405066,
      "loss": 0.1971,
      "num_input_tokens_seen": 1296520,
      "step": 694,
      "train_runtime": 1455.5855,
      "train_tokens_per_second": 890.721
    },
    {
      "epoch": 0.7025524387161991,
      "grad_norm": 0.37987568974494934,
      "learning_rate": 0.00013022784810126582,
      "loss": 0.2146,
      "num_input_tokens_seen": 1298658,
      "step": 695,
      "train_runtime": 1457.4366,
      "train_tokens_per_second": 891.056
    },
    {
      "epoch": 0.7035633055344959,
      "grad_norm": 0.44168776273727417,
      "learning_rate": 0.00013012658227848101,
      "loss": 0.2155,
      "num_input_tokens_seen": 1300618,
      "step": 696,
      "train_runtime": 1459.261,
      "train_tokens_per_second": 891.285
    },
    {
      "epoch": 0.7045741723527925,
      "grad_norm": 0.3115847110748291,
      "learning_rate": 0.0001300253164556962,
      "loss": 0.1736,
      "num_input_tokens_seen": 1302984,
      "step": 697,
      "train_runtime": 1461.215,
      "train_tokens_per_second": 891.713
    },
    {
      "epoch": 0.7055850391710892,
      "grad_norm": 0.4177860617637634,
      "learning_rate": 0.0001299240506329114,
      "loss": 0.1859,
      "num_input_tokens_seen": 1304498,
      "step": 698,
      "train_runtime": 1462.8754,
      "train_tokens_per_second": 891.736
    },
    {
      "epoch": 0.7065959059893859,
      "grad_norm": 0.5578729510307312,
      "learning_rate": 0.00012982278481012659,
      "loss": 0.2147,
      "num_input_tokens_seen": 1306128,
      "step": 699,
      "train_runtime": 1464.5811,
      "train_tokens_per_second": 891.81
    },
    {
      "epoch": 0.7076067728076826,
      "grad_norm": 0.6117987632751465,
      "learning_rate": 0.00012972151898734178,
      "loss": 0.2632,
      "num_input_tokens_seen": 1307666,
      "step": 700,
      "train_runtime": 1466.2011,
      "train_tokens_per_second": 891.874
    },
    {
      "epoch": 0.7086176396259792,
      "grad_norm": 0.39309704303741455,
      "learning_rate": 0.00012962025316455697,
      "loss": 0.2293,
      "num_input_tokens_seen": 1309742,
      "step": 701,
      "train_runtime": 1468.051,
      "train_tokens_per_second": 892.164
    },
    {
      "epoch": 0.709628506444276,
      "grad_norm": 0.43305808305740356,
      "learning_rate": 0.00012951898734177216,
      "loss": 0.2504,
      "num_input_tokens_seen": 1311894,
      "step": 702,
      "train_runtime": 1469.9164,
      "train_tokens_per_second": 892.496
    },
    {
      "epoch": 0.7106393732625726,
      "grad_norm": 0.46049821376800537,
      "learning_rate": 0.00012941772151898735,
      "loss": 0.1748,
      "num_input_tokens_seen": 1313542,
      "step": 703,
      "train_runtime": 1471.6117,
      "train_tokens_per_second": 892.587
    },
    {
      "epoch": 0.7116502400808693,
      "grad_norm": 0.5031542778015137,
      "learning_rate": 0.00012931645569620254,
      "loss": 0.1891,
      "num_input_tokens_seen": 1315136,
      "step": 704,
      "train_runtime": 1473.2394,
      "train_tokens_per_second": 892.683
    },
    {
      "epoch": 0.7126611068991661,
      "grad_norm": 0.43103334307670593,
      "learning_rate": 0.00012921518987341773,
      "loss": 0.1379,
      "num_input_tokens_seen": 1316974,
      "step": 705,
      "train_runtime": 1474.9401,
      "train_tokens_per_second": 892.9
    },
    {
      "epoch": 0.7136719737174627,
      "grad_norm": 0.39274096488952637,
      "learning_rate": 0.00012911392405063292,
      "loss": 0.1556,
      "num_input_tokens_seen": 1318902,
      "step": 706,
      "train_runtime": 1476.7012,
      "train_tokens_per_second": 893.141
    },
    {
      "epoch": 0.7146828405357594,
      "grad_norm": 0.4736999571323395,
      "learning_rate": 0.0001290126582278481,
      "loss": 0.2823,
      "num_input_tokens_seen": 1320710,
      "step": 707,
      "train_runtime": 1478.3976,
      "train_tokens_per_second": 893.339
    },
    {
      "epoch": 0.7156937073540561,
      "grad_norm": 0.5220576524734497,
      "learning_rate": 0.0001289113924050633,
      "loss": 0.2178,
      "num_input_tokens_seen": 1322216,
      "step": 708,
      "train_runtime": 1480.0601,
      "train_tokens_per_second": 893.353
    },
    {
      "epoch": 0.7167045741723528,
      "grad_norm": 0.4497409164905548,
      "learning_rate": 0.0001288101265822785,
      "loss": 0.2251,
      "num_input_tokens_seen": 1324022,
      "step": 709,
      "train_runtime": 1481.775,
      "train_tokens_per_second": 893.538
    },
    {
      "epoch": 0.7177154409906494,
      "grad_norm": 0.47773632407188416,
      "learning_rate": 0.00012870886075949369,
      "loss": 0.2139,
      "num_input_tokens_seen": 1325794,
      "step": 710,
      "train_runtime": 1483.4324,
      "train_tokens_per_second": 893.734
    },
    {
      "epoch": 0.7187263078089462,
      "grad_norm": 0.47331786155700684,
      "learning_rate": 0.00012860759493670888,
      "loss": 0.1787,
      "num_input_tokens_seen": 1328058,
      "step": 711,
      "train_runtime": 1485.3072,
      "train_tokens_per_second": 894.13
    },
    {
      "epoch": 0.7197371746272428,
      "grad_norm": 0.4962442219257355,
      "learning_rate": 0.00012850632911392407,
      "loss": 0.1936,
      "num_input_tokens_seen": 1329944,
      "step": 712,
      "train_runtime": 1487.0413,
      "train_tokens_per_second": 894.356
    },
    {
      "epoch": 0.7207480414455395,
      "grad_norm": 0.47528332471847534,
      "learning_rate": 0.00012840506329113923,
      "loss": 0.1644,
      "num_input_tokens_seen": 1331844,
      "step": 713,
      "train_runtime": 1488.7594,
      "train_tokens_per_second": 894.6
    },
    {
      "epoch": 0.7217589082638363,
      "grad_norm": 0.4774949550628662,
      "learning_rate": 0.00012830379746835445,
      "loss": 0.1908,
      "num_input_tokens_seen": 1333816,
      "step": 714,
      "train_runtime": 1490.5403,
      "train_tokens_per_second": 894.854
    },
    {
      "epoch": 0.7227697750821329,
      "grad_norm": 0.43645143508911133,
      "learning_rate": 0.00012820253164556964,
      "loss": 0.149,
      "num_input_tokens_seen": 1335350,
      "step": 715,
      "train_runtime": 1492.2145,
      "train_tokens_per_second": 894.878
    },
    {
      "epoch": 0.7237806419004296,
      "grad_norm": 0.38800936937332153,
      "learning_rate": 0.0001281012658227848,
      "loss": 0.2146,
      "num_input_tokens_seen": 1337112,
      "step": 716,
      "train_runtime": 1493.9178,
      "train_tokens_per_second": 895.037
    },
    {
      "epoch": 0.7247915087187263,
      "grad_norm": 0.49082088470458984,
      "learning_rate": 0.00012800000000000002,
      "loss": 0.1886,
      "num_input_tokens_seen": 1339114,
      "step": 717,
      "train_runtime": 1495.7185,
      "train_tokens_per_second": 895.298
    },
    {
      "epoch": 0.725802375537023,
      "grad_norm": 0.4170313775539398,
      "learning_rate": 0.0001278987341772152,
      "loss": 0.1569,
      "num_input_tokens_seen": 1341268,
      "step": 718,
      "train_runtime": 1497.6402,
      "train_tokens_per_second": 895.588
    },
    {
      "epoch": 0.7268132423553196,
      "grad_norm": 0.46090570092201233,
      "learning_rate": 0.00012779746835443038,
      "loss": 0.1811,
      "num_input_tokens_seen": 1343108,
      "step": 719,
      "train_runtime": 1499.4014,
      "train_tokens_per_second": 895.763
    },
    {
      "epoch": 0.7278241091736164,
      "grad_norm": 0.4252725839614868,
      "learning_rate": 0.0001276962025316456,
      "loss": 0.2131,
      "num_input_tokens_seen": 1344944,
      "step": 720,
      "train_runtime": 1501.1569,
      "train_tokens_per_second": 895.938
    },
    {
      "epoch": 0.728834975991913,
      "grad_norm": 0.44955697655677795,
      "learning_rate": 0.00012759493670886076,
      "loss": 0.2252,
      "num_input_tokens_seen": 1346962,
      "step": 721,
      "train_runtime": 1504.7453,
      "train_tokens_per_second": 895.143
    },
    {
      "epoch": 0.7298458428102097,
      "grad_norm": 0.47491171956062317,
      "learning_rate": 0.00012749367088607595,
      "loss": 0.1559,
      "num_input_tokens_seen": 1348754,
      "step": 722,
      "train_runtime": 1506.5053,
      "train_tokens_per_second": 895.287
    },
    {
      "epoch": 0.7308567096285065,
      "grad_norm": 0.4124668836593628,
      "learning_rate": 0.00012739240506329117,
      "loss": 0.2047,
      "num_input_tokens_seen": 1350670,
      "step": 723,
      "train_runtime": 1508.3267,
      "train_tokens_per_second": 895.476
    },
    {
      "epoch": 0.7318675764468031,
      "grad_norm": 0.30529698729515076,
      "learning_rate": 0.00012729113924050633,
      "loss": 0.1088,
      "num_input_tokens_seen": 1353106,
      "step": 724,
      "train_runtime": 1510.3529,
      "train_tokens_per_second": 895.887
    },
    {
      "epoch": 0.7328784432650999,
      "grad_norm": 0.3167610764503479,
      "learning_rate": 0.00012718987341772152,
      "loss": 0.1771,
      "num_input_tokens_seen": 1355056,
      "step": 725,
      "train_runtime": 1512.0713,
      "train_tokens_per_second": 896.159
    },
    {
      "epoch": 0.7338893100833965,
      "grad_norm": 0.36973923444747925,
      "learning_rate": 0.0001270886075949367,
      "loss": 0.2096,
      "num_input_tokens_seen": 1356638,
      "step": 726,
      "train_runtime": 1513.6977,
      "train_tokens_per_second": 896.241
    },
    {
      "epoch": 0.7349001769016932,
      "grad_norm": 0.5283274054527283,
      "learning_rate": 0.0001269873417721519,
      "loss": 0.1622,
      "num_input_tokens_seen": 1358056,
      "step": 727,
      "train_runtime": 1515.2781,
      "train_tokens_per_second": 896.242
    },
    {
      "epoch": 0.7359110437199899,
      "grad_norm": 0.44161418080329895,
      "learning_rate": 0.0001268860759493671,
      "loss": 0.1642,
      "num_input_tokens_seen": 1359776,
      "step": 728,
      "train_runtime": 1516.9189,
      "train_tokens_per_second": 896.407
    },
    {
      "epoch": 0.7369219105382866,
      "grad_norm": 0.4106193780899048,
      "learning_rate": 0.00012678481012658228,
      "loss": 0.1819,
      "num_input_tokens_seen": 1361576,
      "step": 729,
      "train_runtime": 1518.632,
      "train_tokens_per_second": 896.581
    },
    {
      "epoch": 0.7379327773565832,
      "grad_norm": 0.3663170039653778,
      "learning_rate": 0.00012668354430379748,
      "loss": 0.1841,
      "num_input_tokens_seen": 1363414,
      "step": 730,
      "train_runtime": 1520.3902,
      "train_tokens_per_second": 896.753
    },
    {
      "epoch": 0.73894364417488,
      "grad_norm": 0.45822158455848694,
      "learning_rate": 0.00012658227848101267,
      "loss": 0.1727,
      "num_input_tokens_seen": 1365166,
      "step": 731,
      "train_runtime": 1522.1061,
      "train_tokens_per_second": 896.893
    },
    {
      "epoch": 0.7399545109931767,
      "grad_norm": 0.44998377561569214,
      "learning_rate": 0.00012648101265822786,
      "loss": 0.1803,
      "num_input_tokens_seen": 1366760,
      "step": 732,
      "train_runtime": 1523.7404,
      "train_tokens_per_second": 896.977
    },
    {
      "epoch": 0.7409653778114733,
      "grad_norm": 0.38428956270217896,
      "learning_rate": 0.00012637974683544305,
      "loss": 0.1601,
      "num_input_tokens_seen": 1368492,
      "step": 733,
      "train_runtime": 1525.3789,
      "train_tokens_per_second": 897.149
    },
    {
      "epoch": 0.7419762446297701,
      "grad_norm": 0.41664576530456543,
      "learning_rate": 0.0001262784810126582,
      "loss": 0.1784,
      "num_input_tokens_seen": 1370530,
      "step": 734,
      "train_runtime": 1527.1619,
      "train_tokens_per_second": 897.436
    },
    {
      "epoch": 0.7429871114480667,
      "grad_norm": 0.4322846531867981,
      "learning_rate": 0.00012617721518987343,
      "loss": 0.1484,
      "num_input_tokens_seen": 1372698,
      "step": 735,
      "train_runtime": 1529.0505,
      "train_tokens_per_second": 897.745
    },
    {
      "epoch": 0.7439979782663634,
      "grad_norm": 0.49796196818351746,
      "learning_rate": 0.00012607594936708862,
      "loss": 0.2135,
      "num_input_tokens_seen": 1374798,
      "step": 736,
      "train_runtime": 1530.9322,
      "train_tokens_per_second": 898.014
    },
    {
      "epoch": 0.7450088450846601,
      "grad_norm": 0.40537935495376587,
      "learning_rate": 0.00012597468354430378,
      "loss": 0.251,
      "num_input_tokens_seen": 1376982,
      "step": 737,
      "train_runtime": 1532.8699,
      "train_tokens_per_second": 898.303
    },
    {
      "epoch": 0.7460197119029568,
      "grad_norm": 0.48897862434387207,
      "learning_rate": 0.000125873417721519,
      "loss": 0.2775,
      "num_input_tokens_seen": 1378818,
      "step": 738,
      "train_runtime": 1534.5843,
      "train_tokens_per_second": 898.496
    },
    {
      "epoch": 0.7470305787212534,
      "grad_norm": 0.5688496828079224,
      "learning_rate": 0.0001257721518987342,
      "loss": 0.227,
      "num_input_tokens_seen": 1380380,
      "step": 739,
      "train_runtime": 1536.2047,
      "train_tokens_per_second": 898.565
    },
    {
      "epoch": 0.7480414455395502,
      "grad_norm": 0.390685498714447,
      "learning_rate": 0.00012567088607594936,
      "loss": 0.1397,
      "num_input_tokens_seen": 1382458,
      "step": 740,
      "train_runtime": 1538.0695,
      "train_tokens_per_second": 898.827
    },
    {
      "epoch": 0.7490523123578469,
      "grad_norm": 0.5368296504020691,
      "learning_rate": 0.00012556962025316457,
      "loss": 0.1804,
      "num_input_tokens_seen": 1384230,
      "step": 741,
      "train_runtime": 1539.7765,
      "train_tokens_per_second": 898.981
    },
    {
      "epoch": 0.7500631791761435,
      "grad_norm": 0.4591348469257355,
      "learning_rate": 0.00012546835443037974,
      "loss": 0.1778,
      "num_input_tokens_seen": 1386176,
      "step": 742,
      "train_runtime": 1541.5473,
      "train_tokens_per_second": 899.211
    },
    {
      "epoch": 0.7510740459944403,
      "grad_norm": 0.4368685781955719,
      "learning_rate": 0.00012536708860759493,
      "loss": 0.156,
      "num_input_tokens_seen": 1387826,
      "step": 743,
      "train_runtime": 1543.243,
      "train_tokens_per_second": 899.292
    },
    {
      "epoch": 0.7520849128127369,
      "grad_norm": 0.29662302136421204,
      "learning_rate": 0.00012526582278481015,
      "loss": 0.2113,
      "num_input_tokens_seen": 1390232,
      "step": 744,
      "train_runtime": 1545.2493,
      "train_tokens_per_second": 899.681
    },
    {
      "epoch": 0.7530957796310336,
      "grad_norm": 0.4854196012020111,
      "learning_rate": 0.0001251645569620253,
      "loss": 0.1335,
      "num_input_tokens_seen": 1392116,
      "step": 745,
      "train_runtime": 1547.0203,
      "train_tokens_per_second": 899.869
    },
    {
      "epoch": 0.7541066464493303,
      "grad_norm": 0.6670404076576233,
      "learning_rate": 0.0001250632911392405,
      "loss": 0.2498,
      "num_input_tokens_seen": 1393572,
      "step": 746,
      "train_runtime": 1548.6386,
      "train_tokens_per_second": 899.869
    },
    {
      "epoch": 0.755117513267627,
      "grad_norm": 0.4352905750274658,
      "learning_rate": 0.00012496202531645572,
      "loss": 0.1736,
      "num_input_tokens_seen": 1395382,
      "step": 747,
      "train_runtime": 1550.3484,
      "train_tokens_per_second": 900.044
    },
    {
      "epoch": 0.7561283800859236,
      "grad_norm": 0.475554496049881,
      "learning_rate": 0.00012486075949367088,
      "loss": 0.1525,
      "num_input_tokens_seen": 1397442,
      "step": 748,
      "train_runtime": 1552.1008,
      "train_tokens_per_second": 900.355
    },
    {
      "epoch": 0.7571392469042204,
      "grad_norm": 0.43190351128578186,
      "learning_rate": 0.00012475949367088607,
      "loss": 0.1808,
      "num_input_tokens_seen": 1399870,
      "step": 749,
      "train_runtime": 1554.0692,
      "train_tokens_per_second": 900.777
    },
    {
      "epoch": 0.7581501137225171,
      "grad_norm": 0.4478033781051636,
      "learning_rate": 0.00012465822784810126,
      "loss": 0.2439,
      "num_input_tokens_seen": 1402104,
      "step": 750,
      "train_runtime": 1555.9526,
      "train_tokens_per_second": 901.123
    },
    {
      "epoch": 0.7591609805408137,
      "grad_norm": 0.4629665017127991,
      "learning_rate": 0.00012455696202531646,
      "loss": 0.2138,
      "num_input_tokens_seen": 1403906,
      "step": 751,
      "train_runtime": 1559.4725,
      "train_tokens_per_second": 900.244
    },
    {
      "epoch": 0.7601718473591105,
      "grad_norm": 0.39230385422706604,
      "learning_rate": 0.00012445569620253165,
      "loss": 0.1478,
      "num_input_tokens_seen": 1405970,
      "step": 752,
      "train_runtime": 1561.3892,
      "train_tokens_per_second": 900.461
    },
    {
      "epoch": 0.7611827141774071,
      "grad_norm": 0.5002062320709229,
      "learning_rate": 0.00012435443037974684,
      "loss": 0.2248,
      "num_input_tokens_seen": 1407986,
      "step": 753,
      "train_runtime": 1563.1658,
      "train_tokens_per_second": 900.727
    },
    {
      "epoch": 0.7621935809957038,
      "grad_norm": 0.3833402097225189,
      "learning_rate": 0.00012425316455696203,
      "loss": 0.1683,
      "num_input_tokens_seen": 1409718,
      "step": 754,
      "train_runtime": 1564.9194,
      "train_tokens_per_second": 900.825
    },
    {
      "epoch": 0.7632044478140005,
      "grad_norm": 0.530303418636322,
      "learning_rate": 0.00012415189873417722,
      "loss": 0.2111,
      "num_input_tokens_seen": 1411278,
      "step": 755,
      "train_runtime": 1566.5443,
      "train_tokens_per_second": 900.886
    },
    {
      "epoch": 0.7642153146322972,
      "grad_norm": 0.3445185422897339,
      "learning_rate": 0.0001240506329113924,
      "loss": 0.1959,
      "num_input_tokens_seen": 1413560,
      "step": 756,
      "train_runtime": 1568.5295,
      "train_tokens_per_second": 901.201
    },
    {
      "epoch": 0.7652261814505938,
      "grad_norm": 0.5255396366119385,
      "learning_rate": 0.0001239493670886076,
      "loss": 0.2558,
      "num_input_tokens_seen": 1415670,
      "step": 757,
      "train_runtime": 1570.3826,
      "train_tokens_per_second": 901.481
    },
    {
      "epoch": 0.7662370482688906,
      "grad_norm": 0.3746929168701172,
      "learning_rate": 0.0001238481012658228,
      "loss": 0.1443,
      "num_input_tokens_seen": 1417504,
      "step": 758,
      "train_runtime": 1572.121,
      "train_tokens_per_second": 901.651
    },
    {
      "epoch": 0.7672479150871873,
      "grad_norm": 0.4821506142616272,
      "learning_rate": 0.00012374683544303798,
      "loss": 0.235,
      "num_input_tokens_seen": 1419526,
      "step": 759,
      "train_runtime": 1573.9626,
      "train_tokens_per_second": 901.88
    },
    {
      "epoch": 0.7682587819054839,
      "grad_norm": 0.4334051012992859,
      "learning_rate": 0.00012364556962025317,
      "loss": 0.1309,
      "num_input_tokens_seen": 1421724,
      "step": 760,
      "train_runtime": 1575.8729,
      "train_tokens_per_second": 902.182
    },
    {
      "epoch": 0.7692696487237807,
      "grad_norm": 0.38763320446014404,
      "learning_rate": 0.00012354430379746836,
      "loss": 0.2261,
      "num_input_tokens_seen": 1423704,
      "step": 761,
      "train_runtime": 1577.6764,
      "train_tokens_per_second": 902.406
    },
    {
      "epoch": 0.7702805155420773,
      "grad_norm": 0.400185227394104,
      "learning_rate": 0.00012344303797468356,
      "loss": 0.1233,
      "num_input_tokens_seen": 1425830,
      "step": 762,
      "train_runtime": 1579.5407,
      "train_tokens_per_second": 902.686
    },
    {
      "epoch": 0.771291382360374,
      "grad_norm": 0.47351810336112976,
      "learning_rate": 0.00012334177215189875,
      "loss": 0.1551,
      "num_input_tokens_seen": 1427436,
      "step": 763,
      "train_runtime": 1581.2153,
      "train_tokens_per_second": 902.746
    },
    {
      "epoch": 0.7723022491786707,
      "grad_norm": 0.40420597791671753,
      "learning_rate": 0.00012324050632911394,
      "loss": 0.2703,
      "num_input_tokens_seen": 1429530,
      "step": 764,
      "train_runtime": 1583.0653,
      "train_tokens_per_second": 903.014
    },
    {
      "epoch": 0.7733131159969674,
      "grad_norm": 0.4711056649684906,
      "learning_rate": 0.00012313924050632913,
      "loss": 0.1923,
      "num_input_tokens_seen": 1431692,
      "step": 765,
      "train_runtime": 1584.942,
      "train_tokens_per_second": 903.309
    },
    {
      "epoch": 0.774323982815264,
      "grad_norm": 0.4936767518520355,
      "learning_rate": 0.0001230379746835443,
      "loss": 0.1866,
      "num_input_tokens_seen": 1433506,
      "step": 766,
      "train_runtime": 1586.6431,
      "train_tokens_per_second": 903.484
    },
    {
      "epoch": 0.7753348496335608,
      "grad_norm": 0.4346882402896881,
      "learning_rate": 0.0001229367088607595,
      "loss": 0.0903,
      "num_input_tokens_seen": 1435216,
      "step": 767,
      "train_runtime": 1588.3444,
      "train_tokens_per_second": 903.592
    },
    {
      "epoch": 0.7763457164518575,
      "grad_norm": 0.47223055362701416,
      "learning_rate": 0.0001228354430379747,
      "loss": 0.2075,
      "num_input_tokens_seen": 1437180,
      "step": 768,
      "train_runtime": 1590.0702,
      "train_tokens_per_second": 903.847
    },
    {
      "epoch": 0.7773565832701541,
      "grad_norm": 0.40802159905433655,
      "learning_rate": 0.00012273417721518986,
      "loss": 0.1695,
      "num_input_tokens_seen": 1439174,
      "step": 769,
      "train_runtime": 1591.9123,
      "train_tokens_per_second": 904.054
    },
    {
      "epoch": 0.7783674500884509,
      "grad_norm": 0.5417159795761108,
      "learning_rate": 0.00012263291139240508,
      "loss": 0.259,
      "num_input_tokens_seen": 1440818,
      "step": 770,
      "train_runtime": 1593.5367,
      "train_tokens_per_second": 904.164
    },
    {
      "epoch": 0.7793783169067475,
      "grad_norm": 0.47632548213005066,
      "learning_rate": 0.00012253164556962027,
      "loss": 0.2161,
      "num_input_tokens_seen": 1442724,
      "step": 771,
      "train_runtime": 1595.3854,
      "train_tokens_per_second": 904.311
    },
    {
      "epoch": 0.7803891837250442,
      "grad_norm": 0.4863998293876648,
      "learning_rate": 0.00012243037974683544,
      "loss": 0.1861,
      "num_input_tokens_seen": 1444634,
      "step": 772,
      "train_runtime": 1597.1699,
      "train_tokens_per_second": 904.496
    },
    {
      "epoch": 0.7814000505433409,
      "grad_norm": 0.4289018213748932,
      "learning_rate": 0.00012232911392405065,
      "loss": 0.3262,
      "num_input_tokens_seen": 1446376,
      "step": 773,
      "train_runtime": 1598.8017,
      "train_tokens_per_second": 904.663
    },
    {
      "epoch": 0.7824109173616376,
      "grad_norm": 0.48690351843833923,
      "learning_rate": 0.00012222784810126582,
      "loss": 0.1836,
      "num_input_tokens_seen": 1447954,
      "step": 774,
      "train_runtime": 1600.4292,
      "train_tokens_per_second": 904.729
    },
    {
      "epoch": 0.7834217841799342,
      "grad_norm": 0.45947712659835815,
      "learning_rate": 0.000122126582278481,
      "loss": 0.1649,
      "num_input_tokens_seen": 1449790,
      "step": 775,
      "train_runtime": 1602.1404,
      "train_tokens_per_second": 904.908
    },
    {
      "epoch": 0.784432650998231,
      "grad_norm": 0.44784119725227356,
      "learning_rate": 0.00012202531645569621,
      "loss": 0.1773,
      "num_input_tokens_seen": 1451420,
      "step": 776,
      "train_runtime": 1603.7562,
      "train_tokens_per_second": 905.013
    },
    {
      "epoch": 0.7854435178165277,
      "grad_norm": 0.5048577785491943,
      "learning_rate": 0.00012192405063291139,
      "loss": 0.4025,
      "num_input_tokens_seen": 1453310,
      "step": 777,
      "train_runtime": 1605.526,
      "train_tokens_per_second": 905.192
    },
    {
      "epoch": 0.7864543846348243,
      "grad_norm": 0.37859445810317993,
      "learning_rate": 0.0001218227848101266,
      "loss": 0.2033,
      "num_input_tokens_seen": 1455086,
      "step": 778,
      "train_runtime": 1607.2298,
      "train_tokens_per_second": 905.338
    },
    {
      "epoch": 0.7874652514531211,
      "grad_norm": 0.5107361078262329,
      "learning_rate": 0.00012172151898734179,
      "loss": 0.1705,
      "num_input_tokens_seen": 1456632,
      "step": 779,
      "train_runtime": 1608.8468,
      "train_tokens_per_second": 905.389
    },
    {
      "epoch": 0.7884761182714177,
      "grad_norm": 0.44889283180236816,
      "learning_rate": 0.00012162025316455696,
      "loss": 0.1574,
      "num_input_tokens_seen": 1458412,
      "step": 780,
      "train_runtime": 1610.6019,
      "train_tokens_per_second": 905.507
    },
    {
      "epoch": 0.7894869850897144,
      "grad_norm": 0.3823074698448181,
      "learning_rate": 0.00012151898734177217,
      "loss": 0.2528,
      "num_input_tokens_seen": 1460750,
      "step": 781,
      "train_runtime": 1614.3246,
      "train_tokens_per_second": 904.868
    },
    {
      "epoch": 0.7904978519080111,
      "grad_norm": 0.4243546426296234,
      "learning_rate": 0.00012141772151898733,
      "loss": 0.2872,
      "num_input_tokens_seen": 1463018,
      "step": 782,
      "train_runtime": 1616.2762,
      "train_tokens_per_second": 905.178
    },
    {
      "epoch": 0.7915087187263078,
      "grad_norm": 0.34533020853996277,
      "learning_rate": 0.00012131645569620254,
      "loss": 0.2406,
      "num_input_tokens_seen": 1464998,
      "step": 783,
      "train_runtime": 1618.0801,
      "train_tokens_per_second": 905.393
    },
    {
      "epoch": 0.7925195855446044,
      "grad_norm": 0.33816954493522644,
      "learning_rate": 0.00012121518987341774,
      "loss": 0.1489,
      "num_input_tokens_seen": 1467092,
      "step": 784,
      "train_runtime": 1620.0305,
      "train_tokens_per_second": 905.595
    },
    {
      "epoch": 0.7935304523629012,
      "grad_norm": 0.5090389251708984,
      "learning_rate": 0.0001211139240506329,
      "loss": 0.3019,
      "num_input_tokens_seen": 1468734,
      "step": 785,
      "train_runtime": 1621.7232,
      "train_tokens_per_second": 905.663
    },
    {
      "epoch": 0.7945413191811979,
      "grad_norm": 0.46344900131225586,
      "learning_rate": 0.00012101265822784811,
      "loss": 0.1949,
      "num_input_tokens_seen": 1470256,
      "step": 786,
      "train_runtime": 1623.3271,
      "train_tokens_per_second": 905.705
    },
    {
      "epoch": 0.7955521859994946,
      "grad_norm": 0.5146666169166565,
      "learning_rate": 0.00012091139240506331,
      "loss": 0.2443,
      "num_input_tokens_seen": 1471828,
      "step": 787,
      "train_runtime": 1624.9462,
      "train_tokens_per_second": 905.77
    },
    {
      "epoch": 0.7965630528177913,
      "grad_norm": 0.43645817041397095,
      "learning_rate": 0.00012081012658227848,
      "loss": 0.2064,
      "num_input_tokens_seen": 1473878,
      "step": 788,
      "train_runtime": 1626.7993,
      "train_tokens_per_second": 905.999
    },
    {
      "epoch": 0.7975739196360879,
      "grad_norm": 0.44168621301651,
      "learning_rate": 0.00012070886075949368,
      "loss": 0.2864,
      "num_input_tokens_seen": 1476312,
      "step": 789,
      "train_runtime": 1628.7774,
      "train_tokens_per_second": 906.393
    },
    {
      "epoch": 0.7985847864543847,
      "grad_norm": 0.5032877326011658,
      "learning_rate": 0.00012060759493670886,
      "loss": 0.2168,
      "num_input_tokens_seen": 1478126,
      "step": 790,
      "train_runtime": 1630.5462,
      "train_tokens_per_second": 906.522
    },
    {
      "epoch": 0.7995956532726813,
      "grad_norm": 0.3511337637901306,
      "learning_rate": 0.00012050632911392405,
      "loss": 0.2164,
      "num_input_tokens_seen": 1480092,
      "step": 791,
      "train_runtime": 1632.3919,
      "train_tokens_per_second": 906.701
    },
    {
      "epoch": 0.800606520090978,
      "grad_norm": 0.558936357498169,
      "learning_rate": 0.00012040506329113925,
      "loss": 0.2207,
      "num_input_tokens_seen": 1481796,
      "step": 792,
      "train_runtime": 1634.0818,
      "train_tokens_per_second": 906.807
    },
    {
      "epoch": 0.800606520090978,
      "eval_loss": 0.19605906307697296,
      "eval_runtime": 61.4416,
      "eval_samples_per_second": 14.306,
      "eval_steps_per_second": 7.161,
      "num_input_tokens_seen": 1481796,
      "step": 792
    },
    {
      "epoch": 0.8016173869092748,
      "grad_norm": 0.48004817962646484,
      "learning_rate": 0.00012030379746835443,
      "loss": 0.2437,
      "num_input_tokens_seen": 1483304,
      "step": 793,
      "train_runtime": 1697.1412,
      "train_tokens_per_second": 874.002
    },
    {
      "epoch": 0.8026282537275714,
      "grad_norm": 0.37774133682250977,
      "learning_rate": 0.00012020253164556962,
      "loss": 0.1135,
      "num_input_tokens_seen": 1484762,
      "step": 794,
      "train_runtime": 1698.7761,
      "train_tokens_per_second": 874.019
    },
    {
      "epoch": 0.8036391205458681,
      "grad_norm": 0.4917088449001312,
      "learning_rate": 0.00012010126582278483,
      "loss": 0.2255,
      "num_input_tokens_seen": 1486508,
      "step": 795,
      "train_runtime": 1700.5321,
      "train_tokens_per_second": 874.143
    },
    {
      "epoch": 0.8046499873641648,
      "grad_norm": 0.4402884244918823,
      "learning_rate": 0.00012,
      "loss": 0.1763,
      "num_input_tokens_seen": 1488478,
      "step": 796,
      "train_runtime": 1702.2929,
      "train_tokens_per_second": 874.396
    },
    {
      "epoch": 0.8056608541824615,
      "grad_norm": 0.37884581089019775,
      "learning_rate": 0.0001198987341772152,
      "loss": 0.193,
      "num_input_tokens_seen": 1490572,
      "step": 797,
      "train_runtime": 1704.0914,
      "train_tokens_per_second": 874.702
    },
    {
      "epoch": 0.8066717210007581,
      "grad_norm": 0.4645121991634369,
      "learning_rate": 0.00011979746835443037,
      "loss": 0.2041,
      "num_input_tokens_seen": 1492388,
      "step": 798,
      "train_runtime": 1705.7437,
      "train_tokens_per_second": 874.919
    },
    {
      "epoch": 0.8076825878190549,
      "grad_norm": 0.4440683126449585,
      "learning_rate": 0.00011969620253164558,
      "loss": 0.2565,
      "num_input_tokens_seen": 1494258,
      "step": 799,
      "train_runtime": 1707.4461,
      "train_tokens_per_second": 875.142
    },
    {
      "epoch": 0.8086934546373515,
      "grad_norm": 0.3711960017681122,
      "learning_rate": 0.00011959493670886077,
      "loss": 0.1793,
      "num_input_tokens_seen": 1496008,
      "step": 800,
      "train_runtime": 1709.1545,
      "train_tokens_per_second": 875.291
    },
    {
      "epoch": 0.8097043214556482,
      "grad_norm": 0.5635051727294922,
      "learning_rate": 0.00011949367088607594,
      "loss": 0.2713,
      "num_input_tokens_seen": 1497806,
      "step": 801,
      "train_runtime": 1710.8502,
      "train_tokens_per_second": 875.475
    },
    {
      "epoch": 0.810715188273945,
      "grad_norm": 0.3368736505508423,
      "learning_rate": 0.00011939240506329115,
      "loss": 0.2824,
      "num_input_tokens_seen": 1499654,
      "step": 802,
      "train_runtime": 1712.5673,
      "train_tokens_per_second": 875.676
    },
    {
      "epoch": 0.8117260550922416,
      "grad_norm": 0.35640907287597656,
      "learning_rate": 0.00011929113924050634,
      "loss": 0.1959,
      "num_input_tokens_seen": 1502054,
      "step": 803,
      "train_runtime": 1714.6426,
      "train_tokens_per_second": 876.016
    },
    {
      "epoch": 0.8127369219105383,
      "grad_norm": 0.42300862073898315,
      "learning_rate": 0.00011918987341772152,
      "loss": 0.1734,
      "num_input_tokens_seen": 1503636,
      "step": 804,
      "train_runtime": 1716.2663,
      "train_tokens_per_second": 876.109
    },
    {
      "epoch": 0.813747788728835,
      "grad_norm": 0.4077221155166626,
      "learning_rate": 0.00011908860759493672,
      "loss": 0.249,
      "num_input_tokens_seen": 1505588,
      "step": 805,
      "train_runtime": 1718.1438,
      "train_tokens_per_second": 876.288
    },
    {
      "epoch": 0.8147586555471317,
      "grad_norm": 0.5087118148803711,
      "learning_rate": 0.0001189873417721519,
      "loss": 0.2233,
      "num_input_tokens_seen": 1507126,
      "step": 806,
      "train_runtime": 1719.7569,
      "train_tokens_per_second": 876.36
    },
    {
      "epoch": 0.8157695223654283,
      "grad_norm": 0.31490835547447205,
      "learning_rate": 0.00011888607594936709,
      "loss": 0.1315,
      "num_input_tokens_seen": 1509118,
      "step": 807,
      "train_runtime": 1721.5542,
      "train_tokens_per_second": 876.602
    },
    {
      "epoch": 0.8167803891837251,
      "grad_norm": 0.485928475856781,
      "learning_rate": 0.00011878481012658229,
      "loss": 0.254,
      "num_input_tokens_seen": 1510842,
      "step": 808,
      "train_runtime": 1723.2535,
      "train_tokens_per_second": 876.738
    },
    {
      "epoch": 0.8177912560020217,
      "grad_norm": 0.5121615529060364,
      "learning_rate": 0.00011868354430379747,
      "loss": 0.1921,
      "num_input_tokens_seen": 1512260,
      "step": 809,
      "train_runtime": 1724.8624,
      "train_tokens_per_second": 876.742
    },
    {
      "epoch": 0.8188021228203184,
      "grad_norm": 0.5394084453582764,
      "learning_rate": 0.00011858227848101266,
      "loss": 0.2328,
      "num_input_tokens_seen": 1514086,
      "step": 810,
      "train_runtime": 1726.572,
      "train_tokens_per_second": 876.932
    },
    {
      "epoch": 0.8198129896386152,
      "grad_norm": 0.46057066321372986,
      "learning_rate": 0.00011848101265822787,
      "loss": 0.2042,
      "num_input_tokens_seen": 1515896,
      "step": 811,
      "train_runtime": 1730.3617,
      "train_tokens_per_second": 876.057
    },
    {
      "epoch": 0.8208238564569118,
      "grad_norm": 0.428923100233078,
      "learning_rate": 0.00011837974683544304,
      "loss": 0.2546,
      "num_input_tokens_seen": 1518066,
      "step": 812,
      "train_runtime": 1732.237,
      "train_tokens_per_second": 876.362
    },
    {
      "epoch": 0.8218347232752085,
      "grad_norm": 0.43719565868377686,
      "learning_rate": 0.00011827848101265823,
      "loss": 0.1651,
      "num_input_tokens_seen": 1519894,
      "step": 813,
      "train_runtime": 1733.9345,
      "train_tokens_per_second": 876.558
    },
    {
      "epoch": 0.8228455900935052,
      "grad_norm": 0.41401392221450806,
      "learning_rate": 0.00011817721518987341,
      "loss": 0.1753,
      "num_input_tokens_seen": 1521416,
      "step": 814,
      "train_runtime": 1735.5397,
      "train_tokens_per_second": 876.624
    },
    {
      "epoch": 0.8238564569118019,
      "grad_norm": 0.447737455368042,
      "learning_rate": 0.00011807594936708862,
      "loss": 0.1801,
      "num_input_tokens_seen": 1523152,
      "step": 815,
      "train_runtime": 1737.3043,
      "train_tokens_per_second": 876.733
    },
    {
      "epoch": 0.8248673237300985,
      "grad_norm": 0.32408955693244934,
      "learning_rate": 0.0001179746835443038,
      "loss": 0.1166,
      "num_input_tokens_seen": 1525124,
      "step": 816,
      "train_runtime": 1739.0719,
      "train_tokens_per_second": 876.976
    },
    {
      "epoch": 0.8258781905483953,
      "grad_norm": 0.3348199129104614,
      "learning_rate": 0.00011787341772151898,
      "loss": 0.2547,
      "num_input_tokens_seen": 1526764,
      "step": 817,
      "train_runtime": 1740.7565,
      "train_tokens_per_second": 877.069
    },
    {
      "epoch": 0.8268890573666919,
      "grad_norm": 0.35868316888809204,
      "learning_rate": 0.00011777215189873419,
      "loss": 0.2158,
      "num_input_tokens_seen": 1528806,
      "step": 818,
      "train_runtime": 1742.5511,
      "train_tokens_per_second": 877.338
    },
    {
      "epoch": 0.8278999241849886,
      "grad_norm": 0.46953269839286804,
      "learning_rate": 0.00011767088607594938,
      "loss": 0.1645,
      "num_input_tokens_seen": 1530316,
      "step": 819,
      "train_runtime": 1744.1514,
      "train_tokens_per_second": 877.399
    },
    {
      "epoch": 0.8289107910032854,
      "grad_norm": 0.49311771988868713,
      "learning_rate": 0.00011756962025316456,
      "loss": 0.2754,
      "num_input_tokens_seen": 1532088,
      "step": 820,
      "train_runtime": 1745.8603,
      "train_tokens_per_second": 877.555
    },
    {
      "epoch": 0.829921657821582,
      "grad_norm": 0.5105012059211731,
      "learning_rate": 0.00011746835443037976,
      "loss": 0.2313,
      "num_input_tokens_seen": 1534130,
      "step": 821,
      "train_runtime": 1747.6999,
      "train_tokens_per_second": 877.799
    },
    {
      "epoch": 0.8309325246398787,
      "grad_norm": 0.5455875992774963,
      "learning_rate": 0.00011736708860759494,
      "loss": 0.1658,
      "num_input_tokens_seen": 1535566,
      "step": 822,
      "train_runtime": 1749.297,
      "train_tokens_per_second": 877.819
    },
    {
      "epoch": 0.8319433914581754,
      "grad_norm": 0.33632540702819824,
      "learning_rate": 0.00011726582278481013,
      "loss": 0.1832,
      "num_input_tokens_seen": 1537676,
      "step": 823,
      "train_runtime": 1751.2157,
      "train_tokens_per_second": 878.062
    },
    {
      "epoch": 0.8329542582764721,
      "grad_norm": 0.417454332113266,
      "learning_rate": 0.00011716455696202533,
      "loss": 0.2208,
      "num_input_tokens_seen": 1540372,
      "step": 824,
      "train_runtime": 1753.4293,
      "train_tokens_per_second": 878.491
    },
    {
      "epoch": 0.8339651250947687,
      "grad_norm": 0.49717605113983154,
      "learning_rate": 0.00011706329113924051,
      "loss": 0.2049,
      "num_input_tokens_seen": 1542294,
      "step": 825,
      "train_runtime": 1755.2116,
      "train_tokens_per_second": 878.694
    },
    {
      "epoch": 0.8349759919130655,
      "grad_norm": 0.5155264735221863,
      "learning_rate": 0.0001169620253164557,
      "loss": 0.297,
      "num_input_tokens_seen": 1543944,
      "step": 826,
      "train_runtime": 1756.8393,
      "train_tokens_per_second": 878.819
    },
    {
      "epoch": 0.8359868587313621,
      "grad_norm": 0.5160174369812012,
      "learning_rate": 0.0001168607594936709,
      "loss": 0.3208,
      "num_input_tokens_seen": 1545834,
      "step": 827,
      "train_runtime": 1758.5485,
      "train_tokens_per_second": 879.04
    },
    {
      "epoch": 0.8369977255496588,
      "grad_norm": 0.5705915689468384,
      "learning_rate": 0.00011675949367088608,
      "loss": 0.3488,
      "num_input_tokens_seen": 1547776,
      "step": 828,
      "train_runtime": 1760.2679,
      "train_tokens_per_second": 879.284
    },
    {
      "epoch": 0.8380085923679556,
      "grad_norm": 0.48469799757003784,
      "learning_rate": 0.00011665822784810127,
      "loss": 0.2096,
      "num_input_tokens_seen": 1549604,
      "step": 829,
      "train_runtime": 1762.0434,
      "train_tokens_per_second": 879.436
    },
    {
      "epoch": 0.8390194591862522,
      "grad_norm": 0.4165559411048889,
      "learning_rate": 0.00011655696202531645,
      "loss": 0.2704,
      "num_input_tokens_seen": 1551412,
      "step": 830,
      "train_runtime": 1763.7986,
      "train_tokens_per_second": 879.586
    },
    {
      "epoch": 0.8400303260045489,
      "grad_norm": 0.43866249918937683,
      "learning_rate": 0.00011645569620253166,
      "loss": 0.2007,
      "num_input_tokens_seen": 1553322,
      "step": 831,
      "train_runtime": 1765.6164,
      "train_tokens_per_second": 879.762
    },
    {
      "epoch": 0.8410411928228456,
      "grad_norm": 0.36295077204704285,
      "learning_rate": 0.00011635443037974685,
      "loss": 0.2139,
      "num_input_tokens_seen": 1555272,
      "step": 832,
      "train_runtime": 1767.3998,
      "train_tokens_per_second": 879.977
    },
    {
      "epoch": 0.8420520596411423,
      "grad_norm": 0.4608427882194519,
      "learning_rate": 0.00011625316455696202,
      "loss": 0.2087,
      "num_input_tokens_seen": 1557374,
      "step": 833,
      "train_runtime": 1769.25,
      "train_tokens_per_second": 880.245
    },
    {
      "epoch": 0.8430629264594389,
      "grad_norm": 0.3304494321346283,
      "learning_rate": 0.00011615189873417723,
      "loss": 0.1256,
      "num_input_tokens_seen": 1559176,
      "step": 834,
      "train_runtime": 1770.9643,
      "train_tokens_per_second": 880.411
    },
    {
      "epoch": 0.8440737932777357,
      "grad_norm": 0.3980521559715271,
      "learning_rate": 0.00011605063291139242,
      "loss": 0.1561,
      "num_input_tokens_seen": 1561254,
      "step": 835,
      "train_runtime": 1772.8262,
      "train_tokens_per_second": 880.658
    },
    {
      "epoch": 0.8450846600960323,
      "grad_norm": 0.49653252959251404,
      "learning_rate": 0.0001159493670886076,
      "loss": 0.2097,
      "num_input_tokens_seen": 1562938,
      "step": 836,
      "train_runtime": 1774.4581,
      "train_tokens_per_second": 880.797
    },
    {
      "epoch": 0.846095526914329,
      "grad_norm": 0.37524303793907166,
      "learning_rate": 0.0001158481012658228,
      "loss": 0.1431,
      "num_input_tokens_seen": 1564634,
      "step": 837,
      "train_runtime": 1776.0892,
      "train_tokens_per_second": 880.943
    },
    {
      "epoch": 0.8471063937326258,
      "grad_norm": 0.41417282819747925,
      "learning_rate": 0.00011574683544303796,
      "loss": 0.1756,
      "num_input_tokens_seen": 1566696,
      "step": 838,
      "train_runtime": 1777.8811,
      "train_tokens_per_second": 881.215
    },
    {
      "epoch": 0.8481172605509224,
      "grad_norm": 0.32036110758781433,
      "learning_rate": 0.00011564556962025317,
      "loss": 0.3549,
      "num_input_tokens_seen": 1568400,
      "step": 839,
      "train_runtime": 1779.6076,
      "train_tokens_per_second": 881.318
    },
    {
      "epoch": 0.849128127369219,
      "grad_norm": 0.4309079945087433,
      "learning_rate": 0.00011554430379746837,
      "loss": 0.2009,
      "num_input_tokens_seen": 1570560,
      "step": 840,
      "train_runtime": 1781.5342,
      "train_tokens_per_second": 881.577
    },
    {
      "epoch": 0.8501389941875158,
      "grad_norm": 0.36666759848594666,
      "learning_rate": 0.00011544303797468354,
      "loss": 0.1309,
      "num_input_tokens_seen": 1572078,
      "step": 841,
      "train_runtime": 1784.9546,
      "train_tokens_per_second": 880.738
    },
    {
      "epoch": 0.8511498610058125,
      "grad_norm": 0.4371661841869354,
      "learning_rate": 0.00011534177215189874,
      "loss": 0.1629,
      "num_input_tokens_seen": 1573858,
      "step": 842,
      "train_runtime": 1786.6094,
      "train_tokens_per_second": 880.919
    },
    {
      "epoch": 0.8521607278241091,
      "grad_norm": 0.4695513844490051,
      "learning_rate": 0.00011524050632911395,
      "loss": 0.2096,
      "num_input_tokens_seen": 1575802,
      "step": 843,
      "train_runtime": 1788.4411,
      "train_tokens_per_second": 881.104
    },
    {
      "epoch": 0.8531715946424059,
      "grad_norm": 0.5117560029029846,
      "learning_rate": 0.00011513924050632911,
      "loss": 0.1798,
      "num_input_tokens_seen": 1577894,
      "step": 844,
      "train_runtime": 1790.3082,
      "train_tokens_per_second": 881.353
    },
    {
      "epoch": 0.8541824614607025,
      "grad_norm": 0.46726614236831665,
      "learning_rate": 0.00011503797468354431,
      "loss": 0.1902,
      "num_input_tokens_seen": 1579838,
      "step": 845,
      "train_runtime": 1792.0991,
      "train_tokens_per_second": 881.557
    },
    {
      "epoch": 0.8551933282789992,
      "grad_norm": 0.396850049495697,
      "learning_rate": 0.00011493670886075949,
      "loss": 0.1679,
      "num_input_tokens_seen": 1581532,
      "step": 846,
      "train_runtime": 1793.8163,
      "train_tokens_per_second": 881.658
    },
    {
      "epoch": 0.856204195097296,
      "grad_norm": 0.4181012213230133,
      "learning_rate": 0.00011483544303797468,
      "loss": 0.2638,
      "num_input_tokens_seen": 1584082,
      "step": 847,
      "train_runtime": 1795.8959,
      "train_tokens_per_second": 882.057
    },
    {
      "epoch": 0.8572150619155926,
      "grad_norm": 0.5064924359321594,
      "learning_rate": 0.00011473417721518989,
      "loss": 0.244,
      "num_input_tokens_seen": 1585974,
      "step": 848,
      "train_runtime": 1797.6786,
      "train_tokens_per_second": 882.234
    },
    {
      "epoch": 0.8582259287338894,
      "grad_norm": 0.3508239984512329,
      "learning_rate": 0.00011463291139240506,
      "loss": 0.2887,
      "num_input_tokens_seen": 1588014,
      "step": 849,
      "train_runtime": 1799.5745,
      "train_tokens_per_second": 882.439
    },
    {
      "epoch": 0.859236795552186,
      "grad_norm": 0.43819907307624817,
      "learning_rate": 0.00011453164556962025,
      "loss": 0.1741,
      "num_input_tokens_seen": 1589460,
      "step": 850,
      "train_runtime": 1801.1844,
      "train_tokens_per_second": 882.453
    },
    {
      "epoch": 0.8602476623704827,
      "grad_norm": 0.5680981278419495,
      "learning_rate": 0.00011443037974683546,
      "loss": 0.2763,
      "num_input_tokens_seen": 1591552,
      "step": 851,
      "train_runtime": 1803.0431,
      "train_tokens_per_second": 882.703
    },
    {
      "epoch": 0.8612585291887794,
      "grad_norm": 0.47815704345703125,
      "learning_rate": 0.00011432911392405064,
      "loss": 0.1554,
      "num_input_tokens_seen": 1593368,
      "step": 852,
      "train_runtime": 1804.8094,
      "train_tokens_per_second": 882.846
    },
    {
      "epoch": 0.8622693960070761,
      "grad_norm": 0.43788331747055054,
      "learning_rate": 0.00011422784810126583,
      "loss": 0.2106,
      "num_input_tokens_seen": 1595052,
      "step": 853,
      "train_runtime": 1806.4393,
      "train_tokens_per_second": 882.981
    },
    {
      "epoch": 0.8632802628253727,
      "grad_norm": 0.5186537504196167,
      "learning_rate": 0.000114126582278481,
      "loss": 0.2235,
      "num_input_tokens_seen": 1596774,
      "step": 854,
      "train_runtime": 1808.1298,
      "train_tokens_per_second": 883.108
    },
    {
      "epoch": 0.8642911296436695,
      "grad_norm": 0.4398152530193329,
      "learning_rate": 0.00011402531645569621,
      "loss": 0.2736,
      "num_input_tokens_seen": 1598976,
      "step": 855,
      "train_runtime": 1810.0631,
      "train_tokens_per_second": 883.381
    },
    {
      "epoch": 0.8653019964619661,
      "grad_norm": 0.42215922474861145,
      "learning_rate": 0.0001139240506329114,
      "loss": 0.1749,
      "num_input_tokens_seen": 1600644,
      "step": 856,
      "train_runtime": 1811.7673,
      "train_tokens_per_second": 883.471
    },
    {
      "epoch": 0.8663128632802628,
      "grad_norm": 0.44031965732574463,
      "learning_rate": 0.00011382278481012658,
      "loss": 0.1371,
      "num_input_tokens_seen": 1602544,
      "step": 857,
      "train_runtime": 1813.5741,
      "train_tokens_per_second": 883.639
    },
    {
      "epoch": 0.8673237300985596,
      "grad_norm": 0.3823223412036896,
      "learning_rate": 0.00011372151898734178,
      "loss": 0.2225,
      "num_input_tokens_seen": 1604252,
      "step": 858,
      "train_runtime": 1815.314,
      "train_tokens_per_second": 883.733
    },
    {
      "epoch": 0.8683345969168562,
      "grad_norm": 0.5721985697746277,
      "learning_rate": 0.00011362025316455697,
      "loss": 0.2417,
      "num_input_tokens_seen": 1605764,
      "step": 859,
      "train_runtime": 1816.9033,
      "train_tokens_per_second": 883.792
    },
    {
      "epoch": 0.8693454637351529,
      "grad_norm": 0.44026222825050354,
      "learning_rate": 0.00011351898734177215,
      "loss": 0.182,
      "num_input_tokens_seen": 1607886,
      "step": 860,
      "train_runtime": 1818.7565,
      "train_tokens_per_second": 884.058
    },
    {
      "epoch": 0.8703563305534496,
      "grad_norm": 0.440901517868042,
      "learning_rate": 0.00011341772151898735,
      "loss": 0.1252,
      "num_input_tokens_seen": 1609856,
      "step": 861,
      "train_runtime": 1820.4279,
      "train_tokens_per_second": 884.328
    },
    {
      "epoch": 0.8713671973717463,
      "grad_norm": 0.49819907546043396,
      "learning_rate": 0.00011331645569620253,
      "loss": 0.2334,
      "num_input_tokens_seen": 1611762,
      "step": 862,
      "train_runtime": 1822.2147,
      "train_tokens_per_second": 884.507
    },
    {
      "epoch": 0.8723780641900429,
      "grad_norm": 0.5641105771064758,
      "learning_rate": 0.00011321518987341772,
      "loss": 0.193,
      "num_input_tokens_seen": 1613372,
      "step": 863,
      "train_runtime": 1823.8516,
      "train_tokens_per_second": 884.596
    },
    {
      "epoch": 0.8733889310083397,
      "grad_norm": 0.3912726044654846,
      "learning_rate": 0.00011311392405063293,
      "loss": 0.1918,
      "num_input_tokens_seen": 1615290,
      "step": 864,
      "train_runtime": 1825.6272,
      "train_tokens_per_second": 884.786
    },
    {
      "epoch": 0.8743997978266363,
      "grad_norm": 0.4572635591030121,
      "learning_rate": 0.0001130126582278481,
      "loss": 0.2682,
      "num_input_tokens_seen": 1617398,
      "step": 865,
      "train_runtime": 1827.4296,
      "train_tokens_per_second": 885.067
    },
    {
      "epoch": 0.875410664644933,
      "grad_norm": 0.4834804832935333,
      "learning_rate": 0.0001129113924050633,
      "loss": 0.1872,
      "num_input_tokens_seen": 1619110,
      "step": 866,
      "train_runtime": 1829.1819,
      "train_tokens_per_second": 885.155
    },
    {
      "epoch": 0.8764215314632298,
      "grad_norm": 0.4358219504356384,
      "learning_rate": 0.0001128101265822785,
      "loss": 0.1598,
      "num_input_tokens_seen": 1621110,
      "step": 867,
      "train_runtime": 1830.9625,
      "train_tokens_per_second": 885.387
    },
    {
      "epoch": 0.8774323982815264,
      "grad_norm": 0.5084313750267029,
      "learning_rate": 0.00011270886075949368,
      "loss": 0.2373,
      "num_input_tokens_seen": 1623138,
      "step": 868,
      "train_runtime": 1832.8883,
      "train_tokens_per_second": 885.563
    },
    {
      "epoch": 0.8784432650998231,
      "grad_norm": 0.43396979570388794,
      "learning_rate": 0.00011260759493670887,
      "loss": 0.1147,
      "num_input_tokens_seen": 1624772,
      "step": 869,
      "train_runtime": 1834.5083,
      "train_tokens_per_second": 885.672
    },
    {
      "epoch": 0.8794541319181198,
      "grad_norm": 0.4286526143550873,
      "learning_rate": 0.00011250632911392404,
      "loss": 0.1809,
      "num_input_tokens_seen": 1627124,
      "step": 870,
      "train_runtime": 1836.56,
      "train_tokens_per_second": 885.963
    },
    {
      "epoch": 0.8804649987364165,
      "grad_norm": 0.3508733808994293,
      "learning_rate": 0.00011240506329113925,
      "loss": 0.2037,
      "num_input_tokens_seen": 1628594,
      "step": 871,
      "train_runtime": 1840.2075,
      "train_tokens_per_second": 885.006
    },
    {
      "epoch": 0.8814758655547131,
      "grad_norm": 0.45954668521881104,
      "learning_rate": 0.00011230379746835444,
      "loss": 0.1671,
      "num_input_tokens_seen": 1630380,
      "step": 872,
      "train_runtime": 1841.9197,
      "train_tokens_per_second": 885.153
    },
    {
      "epoch": 0.8824867323730099,
      "grad_norm": 0.6832849979400635,
      "learning_rate": 0.00011220253164556962,
      "loss": 0.2442,
      "num_input_tokens_seen": 1632128,
      "step": 873,
      "train_runtime": 1843.5621,
      "train_tokens_per_second": 885.312
    },
    {
      "epoch": 0.8834975991913065,
      "grad_norm": 0.48650091886520386,
      "learning_rate": 0.00011210126582278482,
      "loss": 0.33,
      "num_input_tokens_seen": 1634194,
      "step": 874,
      "train_runtime": 1845.4676,
      "train_tokens_per_second": 885.518
    },
    {
      "epoch": 0.8845084660096032,
      "grad_norm": 0.565884530544281,
      "learning_rate": 0.00011200000000000001,
      "loss": 0.2553,
      "num_input_tokens_seen": 1635922,
      "step": 875,
      "train_runtime": 1847.2051,
      "train_tokens_per_second": 885.62
    },
    {
      "epoch": 0.8855193328279,
      "grad_norm": 0.445090651512146,
      "learning_rate": 0.00011189873417721519,
      "loss": 0.1394,
      "num_input_tokens_seen": 1638068,
      "step": 876,
      "train_runtime": 1849.1313,
      "train_tokens_per_second": 885.858
    },
    {
      "epoch": 0.8865301996461966,
      "grad_norm": 0.5211514830589294,
      "learning_rate": 0.0001117974683544304,
      "loss": 0.2277,
      "num_input_tokens_seen": 1639642,
      "step": 877,
      "train_runtime": 1850.7581,
      "train_tokens_per_second": 885.93
    },
    {
      "epoch": 0.8875410664644933,
      "grad_norm": 0.4003089964389801,
      "learning_rate": 0.00011169620253164557,
      "loss": 0.2932,
      "num_input_tokens_seen": 1641240,
      "step": 878,
      "train_runtime": 1852.3803,
      "train_tokens_per_second": 886.017
    },
    {
      "epoch": 0.88855193328279,
      "grad_norm": 0.40674495697021484,
      "learning_rate": 0.00011159493670886076,
      "loss": 0.1979,
      "num_input_tokens_seen": 1643092,
      "step": 879,
      "train_runtime": 1854.1722,
      "train_tokens_per_second": 886.159
    },
    {
      "epoch": 0.8895628001010867,
      "grad_norm": 0.5176137089729309,
      "learning_rate": 0.00011149367088607597,
      "loss": 0.2057,
      "num_input_tokens_seen": 1644710,
      "step": 880,
      "train_runtime": 1855.8244,
      "train_tokens_per_second": 886.242
    },
    {
      "epoch": 0.8905736669193833,
      "grad_norm": 0.38269540667533875,
      "learning_rate": 0.00011139240506329114,
      "loss": 0.2282,
      "num_input_tokens_seen": 1646118,
      "step": 881,
      "train_runtime": 1857.4097,
      "train_tokens_per_second": 886.244
    },
    {
      "epoch": 0.8915845337376801,
      "grad_norm": 0.38651415705680847,
      "learning_rate": 0.00011129113924050633,
      "loss": 0.1238,
      "num_input_tokens_seen": 1647746,
      "step": 882,
      "train_runtime": 1859.0404,
      "train_tokens_per_second": 886.342
    },
    {
      "epoch": 0.8925954005559767,
      "grad_norm": 0.4340237081050873,
      "learning_rate": 0.00011118987341772154,
      "loss": 0.163,
      "num_input_tokens_seen": 1649270,
      "step": 883,
      "train_runtime": 1860.6619,
      "train_tokens_per_second": 886.389
    },
    {
      "epoch": 0.8936062673742734,
      "grad_norm": 0.34248411655426025,
      "learning_rate": 0.00011108860759493672,
      "loss": 0.1758,
      "num_input_tokens_seen": 1652040,
      "step": 884,
      "train_runtime": 1862.8963,
      "train_tokens_per_second": 886.813
    },
    {
      "epoch": 0.8946171341925702,
      "grad_norm": 0.36963558197021484,
      "learning_rate": 0.0001109873417721519,
      "loss": 0.1487,
      "num_input_tokens_seen": 1654126,
      "step": 885,
      "train_runtime": 1864.6948,
      "train_tokens_per_second": 887.076
    },
    {
      "epoch": 0.8956280010108668,
      "grad_norm": 0.6001124382019043,
      "learning_rate": 0.00011088607594936708,
      "loss": 0.2062,
      "num_input_tokens_seen": 1655560,
      "step": 886,
      "train_runtime": 1866.3026,
      "train_tokens_per_second": 887.08
    },
    {
      "epoch": 0.8966388678291635,
      "grad_norm": 0.47406721115112305,
      "learning_rate": 0.00011078481012658229,
      "loss": 0.1831,
      "num_input_tokens_seen": 1657012,
      "step": 887,
      "train_runtime": 1867.9057,
      "train_tokens_per_second": 887.096
    },
    {
      "epoch": 0.8976497346474602,
      "grad_norm": 0.5258237719535828,
      "learning_rate": 0.00011068354430379748,
      "loss": 0.2067,
      "num_input_tokens_seen": 1658512,
      "step": 888,
      "train_runtime": 1869.5274,
      "train_tokens_per_second": 887.129
    },
    {
      "epoch": 0.8986606014657569,
      "grad_norm": 0.3512861132621765,
      "learning_rate": 0.00011058227848101266,
      "loss": 0.1172,
      "num_input_tokens_seen": 1660652,
      "step": 889,
      "train_runtime": 1871.3802,
      "train_tokens_per_second": 887.394
    },
    {
      "epoch": 0.8996714682840535,
      "grad_norm": 0.5059172511100769,
      "learning_rate": 0.00011048101265822786,
      "loss": 0.2965,
      "num_input_tokens_seen": 1662352,
      "step": 890,
      "train_runtime": 1873.0933,
      "train_tokens_per_second": 887.49
    },
    {
      "epoch": 0.9006823351023503,
      "grad_norm": 0.3939543068408966,
      "learning_rate": 0.00011037974683544305,
      "loss": 0.1308,
      "num_input_tokens_seen": 1664158,
      "step": 891,
      "train_runtime": 1874.8841,
      "train_tokens_per_second": 887.606
    },
    {
      "epoch": 0.901693201920647,
      "grad_norm": 0.4661056399345398,
      "learning_rate": 0.00011027848101265823,
      "loss": 0.2556,
      "num_input_tokens_seen": 1666046,
      "step": 892,
      "train_runtime": 1876.6867,
      "train_tokens_per_second": 887.759
    },
    {
      "epoch": 0.9027040687389436,
      "grad_norm": 0.3504481017589569,
      "learning_rate": 0.00011017721518987343,
      "loss": 0.1371,
      "num_input_tokens_seen": 1667806,
      "step": 893,
      "train_runtime": 1878.402,
      "train_tokens_per_second": 887.886
    },
    {
      "epoch": 0.9037149355572404,
      "grad_norm": 0.4064381718635559,
      "learning_rate": 0.00011007594936708861,
      "loss": 0.1452,
      "num_input_tokens_seen": 1669604,
      "step": 894,
      "train_runtime": 1880.1036,
      "train_tokens_per_second": 888.038
    },
    {
      "epoch": 0.904725802375537,
      "grad_norm": 0.4747428297996521,
      "learning_rate": 0.0001099746835443038,
      "loss": 0.1422,
      "num_input_tokens_seen": 1671732,
      "step": 895,
      "train_runtime": 1881.9667,
      "train_tokens_per_second": 888.29
    },
    {
      "epoch": 0.9057366691938337,
      "grad_norm": 0.5185026526451111,
      "learning_rate": 0.000109873417721519,
      "loss": 0.165,
      "num_input_tokens_seen": 1673544,
      "step": 896,
      "train_runtime": 1883.6717,
      "train_tokens_per_second": 888.448
    },
    {
      "epoch": 0.9067475360121304,
      "grad_norm": 0.490278959274292,
      "learning_rate": 0.00010977215189873418,
      "loss": 0.1568,
      "num_input_tokens_seen": 1675252,
      "step": 897,
      "train_runtime": 1885.2941,
      "train_tokens_per_second": 888.589
    },
    {
      "epoch": 0.9077584028304271,
      "grad_norm": 0.5024723410606384,
      "learning_rate": 0.00010967088607594937,
      "loss": 0.1858,
      "num_input_tokens_seen": 1676740,
      "step": 898,
      "train_runtime": 1886.951,
      "train_tokens_per_second": 888.598
    },
    {
      "epoch": 0.9087692696487237,
      "grad_norm": 0.4894983470439911,
      "learning_rate": 0.00010956962025316458,
      "loss": 0.2375,
      "num_input_tokens_seen": 1678430,
      "step": 899,
      "train_runtime": 1888.6525,
      "train_tokens_per_second": 888.692
    },
    {
      "epoch": 0.9097801364670205,
      "grad_norm": 0.6311760544776917,
      "learning_rate": 0.00010946835443037976,
      "loss": 0.3262,
      "num_input_tokens_seen": 1679904,
      "step": 900,
      "train_runtime": 1890.2784,
      "train_tokens_per_second": 888.707
    },
    {
      "epoch": 0.9107910032853171,
      "grad_norm": 0.44200921058654785,
      "learning_rate": 0.00010936708860759495,
      "loss": 0.1833,
      "num_input_tokens_seen": 1682032,
      "step": 901,
      "train_runtime": 1894.2738,
      "train_tokens_per_second": 887.956
    },
    {
      "epoch": 0.9118018701036138,
      "grad_norm": 0.44553565979003906,
      "learning_rate": 0.00010926582278481012,
      "loss": 0.2013,
      "num_input_tokens_seen": 1683836,
      "step": 902,
      "train_runtime": 1895.917,
      "train_tokens_per_second": 888.138
    },
    {
      "epoch": 0.9128127369219106,
      "grad_norm": 0.4635987877845764,
      "learning_rate": 0.00010916455696202533,
      "loss": 0.1742,
      "num_input_tokens_seen": 1685886,
      "step": 903,
      "train_runtime": 1897.7649,
      "train_tokens_per_second": 888.353
    },
    {
      "epoch": 0.9138236037402072,
      "grad_norm": 0.49628710746765137,
      "learning_rate": 0.00010906329113924052,
      "loss": 0.2352,
      "num_input_tokens_seen": 1687828,
      "step": 904,
      "train_runtime": 1899.5327,
      "train_tokens_per_second": 888.549
    },
    {
      "epoch": 0.9148344705585039,
      "grad_norm": 0.4151511490345001,
      "learning_rate": 0.0001089620253164557,
      "loss": 0.1543,
      "num_input_tokens_seen": 1689976,
      "step": 905,
      "train_runtime": 1901.4789,
      "train_tokens_per_second": 888.769
    },
    {
      "epoch": 0.9158453373768006,
      "grad_norm": 0.4763267934322357,
      "learning_rate": 0.00010886075949367089,
      "loss": 0.2248,
      "num_input_tokens_seen": 1692238,
      "step": 906,
      "train_runtime": 1903.4353,
      "train_tokens_per_second": 889.044
    },
    {
      "epoch": 0.9168562041950973,
      "grad_norm": 0.5129389762878418,
      "learning_rate": 0.00010875949367088606,
      "loss": 0.263,
      "num_input_tokens_seen": 1693838,
      "step": 907,
      "train_runtime": 1905.0418,
      "train_tokens_per_second": 889.134
    },
    {
      "epoch": 0.9178670710133939,
      "grad_norm": 0.4036412835121155,
      "learning_rate": 0.00010865822784810127,
      "loss": 0.1658,
      "num_input_tokens_seen": 1695652,
      "step": 908,
      "train_runtime": 1906.8261,
      "train_tokens_per_second": 889.254
    },
    {
      "epoch": 0.9188779378316907,
      "grad_norm": 0.4002031683921814,
      "learning_rate": 0.00010855696202531646,
      "loss": 0.1565,
      "num_input_tokens_seen": 1697394,
      "step": 909,
      "train_runtime": 1908.5323,
      "train_tokens_per_second": 889.371
    },
    {
      "epoch": 0.9198888046499873,
      "grad_norm": 0.37694117426872253,
      "learning_rate": 0.00010845569620253164,
      "loss": 0.1316,
      "num_input_tokens_seen": 1699096,
      "step": 910,
      "train_runtime": 1910.1776,
      "train_tokens_per_second": 889.496
    },
    {
      "epoch": 0.9208996714682841,
      "grad_norm": 0.45809051394462585,
      "learning_rate": 0.00010835443037974684,
      "loss": 0.1461,
      "num_input_tokens_seen": 1700584,
      "step": 911,
      "train_runtime": 1911.777,
      "train_tokens_per_second": 889.531
    },
    {
      "epoch": 0.9219105382865808,
      "grad_norm": 0.3088432252407074,
      "learning_rate": 0.00010825316455696203,
      "loss": 0.1236,
      "num_input_tokens_seen": 1702466,
      "step": 912,
      "train_runtime": 1913.5441,
      "train_tokens_per_second": 889.693
    },
    {
      "epoch": 0.9229214051048774,
      "grad_norm": 0.37983253598213196,
      "learning_rate": 0.00010815189873417721,
      "loss": 0.2337,
      "num_input_tokens_seen": 1704424,
      "step": 913,
      "train_runtime": 1915.3946,
      "train_tokens_per_second": 889.855
    },
    {
      "epoch": 0.9239322719231742,
      "grad_norm": 0.4673791825771332,
      "learning_rate": 0.00010805063291139241,
      "loss": 0.2648,
      "num_input_tokens_seen": 1706634,
      "step": 914,
      "train_runtime": 1917.3303,
      "train_tokens_per_second": 890.11
    },
    {
      "epoch": 0.9249431387414708,
      "grad_norm": 0.3637625277042389,
      "learning_rate": 0.00010794936708860759,
      "loss": 0.2364,
      "num_input_tokens_seen": 1708190,
      "step": 915,
      "train_runtime": 1918.9495,
      "train_tokens_per_second": 890.169
    },
    {
      "epoch": 0.9259540055597675,
      "grad_norm": 0.5477350354194641,
      "learning_rate": 0.00010784810126582278,
      "loss": 0.2038,
      "num_input_tokens_seen": 1710078,
      "step": 916,
      "train_runtime": 1920.6526,
      "train_tokens_per_second": 890.363
    },
    {
      "epoch": 0.9269648723780642,
      "grad_norm": 0.37270766496658325,
      "learning_rate": 0.00010774683544303799,
      "loss": 0.3118,
      "num_input_tokens_seen": 1712108,
      "step": 917,
      "train_runtime": 1922.451,
      "train_tokens_per_second": 890.586
    },
    {
      "epoch": 0.9279757391963609,
      "grad_norm": 0.47466373443603516,
      "learning_rate": 0.00010764556962025316,
      "loss": 0.1641,
      "num_input_tokens_seen": 1713890,
      "step": 918,
      "train_runtime": 1924.1644,
      "train_tokens_per_second": 890.719
    },
    {
      "epoch": 0.9289866060146575,
      "grad_norm": 0.34475672245025635,
      "learning_rate": 0.00010754430379746835,
      "loss": 0.1686,
      "num_input_tokens_seen": 1715514,
      "step": 919,
      "train_runtime": 1925.8389,
      "train_tokens_per_second": 890.788
    },
    {
      "epoch": 0.9299974728329543,
      "grad_norm": 0.4040345251560211,
      "learning_rate": 0.00010744303797468356,
      "loss": 0.2292,
      "num_input_tokens_seen": 1717310,
      "step": 920,
      "train_runtime": 1927.4771,
      "train_tokens_per_second": 890.963
    },
    {
      "epoch": 0.931008339651251,
      "grad_norm": 0.31060582399368286,
      "learning_rate": 0.00010734177215189874,
      "loss": 0.1675,
      "num_input_tokens_seen": 1719370,
      "step": 921,
      "train_runtime": 1929.2779,
      "train_tokens_per_second": 891.199
    },
    {
      "epoch": 0.9320192064695476,
      "grad_norm": 0.5087334513664246,
      "learning_rate": 0.00010724050632911393,
      "loss": 0.2097,
      "num_input_tokens_seen": 1720898,
      "step": 922,
      "train_runtime": 1930.8921,
      "train_tokens_per_second": 891.245
    },
    {
      "epoch": 0.9330300732878444,
      "grad_norm": 0.49266955256462097,
      "learning_rate": 0.0001071392405063291,
      "loss": 0.1669,
      "num_input_tokens_seen": 1722320,
      "step": 923,
      "train_runtime": 1932.4963,
      "train_tokens_per_second": 891.241
    },
    {
      "epoch": 0.934040940106141,
      "grad_norm": 0.4892714321613312,
      "learning_rate": 0.00010703797468354431,
      "loss": 0.1801,
      "num_input_tokens_seen": 1724026,
      "step": 924,
      "train_runtime": 1934.1203,
      "train_tokens_per_second": 891.375
    },
    {
      "epoch": 0.9350518069244377,
      "grad_norm": 0.38709861040115356,
      "learning_rate": 0.0001069367088607595,
      "loss": 0.2007,
      "num_input_tokens_seen": 1725858,
      "step": 925,
      "train_runtime": 1935.8307,
      "train_tokens_per_second": 891.534
    },
    {
      "epoch": 0.9360626737427344,
      "grad_norm": 0.35277503728866577,
      "learning_rate": 0.00010683544303797468,
      "loss": 0.1485,
      "num_input_tokens_seen": 1728054,
      "step": 926,
      "train_runtime": 1937.7581,
      "train_tokens_per_second": 891.78
    },
    {
      "epoch": 0.9370735405610311,
      "grad_norm": 0.4426005482673645,
      "learning_rate": 0.00010673417721518988,
      "loss": 0.24,
      "num_input_tokens_seen": 1729916,
      "step": 927,
      "train_runtime": 1939.4776,
      "train_tokens_per_second": 891.949
    },
    {
      "epoch": 0.9380844073793277,
      "grad_norm": 0.500525176525116,
      "learning_rate": 0.00010663291139240507,
      "loss": 0.1672,
      "num_input_tokens_seen": 1731898,
      "step": 928,
      "train_runtime": 1941.2709,
      "train_tokens_per_second": 892.146
    },
    {
      "epoch": 0.9390952741976245,
      "grad_norm": 0.4140866696834564,
      "learning_rate": 0.00010653164556962025,
      "loss": 0.1954,
      "num_input_tokens_seen": 1733716,
      "step": 929,
      "train_runtime": 1942.9892,
      "train_tokens_per_second": 892.293
    },
    {
      "epoch": 0.9401061410159212,
      "grad_norm": 0.47640150785446167,
      "learning_rate": 0.00010643037974683545,
      "loss": 0.1977,
      "num_input_tokens_seen": 1735832,
      "step": 930,
      "train_runtime": 1944.8405,
      "train_tokens_per_second": 892.532
    },
    {
      "epoch": 0.9411170078342178,
      "grad_norm": 0.36944422125816345,
      "learning_rate": 0.00010632911392405063,
      "loss": 0.122,
      "num_input_tokens_seen": 1737384,
      "step": 931,
      "train_runtime": 1948.2939,
      "train_tokens_per_second": 891.746
    },
    {
      "epoch": 0.9421278746525146,
      "grad_norm": 0.5811285972595215,
      "learning_rate": 0.00010622784810126582,
      "loss": 0.2328,
      "num_input_tokens_seen": 1739056,
      "step": 932,
      "train_runtime": 1949.9299,
      "train_tokens_per_second": 891.856
    },
    {
      "epoch": 0.9431387414708112,
      "grad_norm": 0.34662237763404846,
      "learning_rate": 0.00010612658227848103,
      "loss": 0.1625,
      "num_input_tokens_seen": 1740828,
      "step": 933,
      "train_runtime": 1951.7454,
      "train_tokens_per_second": 891.934
    },
    {
      "epoch": 0.9441496082891079,
      "grad_norm": 0.4259108603000641,
      "learning_rate": 0.0001060253164556962,
      "loss": 0.2795,
      "num_input_tokens_seen": 1742910,
      "step": 934,
      "train_runtime": 1953.5277,
      "train_tokens_per_second": 892.186
    },
    {
      "epoch": 0.9451604751074046,
      "grad_norm": 0.48577380180358887,
      "learning_rate": 0.0001059240506329114,
      "loss": 0.2043,
      "num_input_tokens_seen": 1744572,
      "step": 935,
      "train_runtime": 1955.1964,
      "train_tokens_per_second": 892.275
    },
    {
      "epoch": 0.9461713419257013,
      "grad_norm": 0.43631938099861145,
      "learning_rate": 0.0001058227848101266,
      "loss": 0.2168,
      "num_input_tokens_seen": 1746228,
      "step": 936,
      "train_runtime": 1956.8999,
      "train_tokens_per_second": 892.344
    },
    {
      "epoch": 0.9471822087439979,
      "grad_norm": 0.5370936393737793,
      "learning_rate": 0.00010572151898734178,
      "loss": 0.1935,
      "num_input_tokens_seen": 1747712,
      "step": 937,
      "train_runtime": 1958.5506,
      "train_tokens_per_second": 892.35
    },
    {
      "epoch": 0.9481930755622947,
      "grad_norm": 0.42087486386299133,
      "learning_rate": 0.00010562025316455697,
      "loss": 0.1533,
      "num_input_tokens_seen": 1749338,
      "step": 938,
      "train_runtime": 1960.1919,
      "train_tokens_per_second": 892.432
    },
    {
      "epoch": 0.9492039423805914,
      "grad_norm": 0.40185943245887756,
      "learning_rate": 0.00010551898734177214,
      "loss": 0.1454,
      "num_input_tokens_seen": 1751606,
      "step": 939,
      "train_runtime": 1962.0743,
      "train_tokens_per_second": 892.732
    },
    {
      "epoch": 0.950214809198888,
      "grad_norm": 0.48460665345191956,
      "learning_rate": 0.00010541772151898735,
      "loss": 0.1571,
      "num_input_tokens_seen": 1753256,
      "step": 940,
      "train_runtime": 1963.7585,
      "train_tokens_per_second": 892.806
    },
    {
      "epoch": 0.9512256760171848,
      "grad_norm": 0.38726770877838135,
      "learning_rate": 0.00010531645569620254,
      "loss": 0.1855,
      "num_input_tokens_seen": 1755012,
      "step": 941,
      "train_runtime": 1965.4757,
      "train_tokens_per_second": 892.92
    },
    {
      "epoch": 0.9522365428354814,
      "grad_norm": 0.4741378724575043,
      "learning_rate": 0.00010521518987341772,
      "loss": 0.2114,
      "num_input_tokens_seen": 1757092,
      "step": 942,
      "train_runtime": 1967.4445,
      "train_tokens_per_second": 893.083
    },
    {
      "epoch": 0.9532474096537781,
      "grad_norm": 0.4324609637260437,
      "learning_rate": 0.00010511392405063292,
      "loss": 0.1409,
      "num_input_tokens_seen": 1759050,
      "step": 943,
      "train_runtime": 1969.2276,
      "train_tokens_per_second": 893.269
    },
    {
      "epoch": 0.9542582764720748,
      "grad_norm": 0.44481709599494934,
      "learning_rate": 0.00010501265822784811,
      "loss": 0.15,
      "num_input_tokens_seen": 1761294,
      "step": 944,
      "train_runtime": 1971.1068,
      "train_tokens_per_second": 893.556
    },
    {
      "epoch": 0.9552691432903715,
      "grad_norm": 0.5897843241691589,
      "learning_rate": 0.00010491139240506329,
      "loss": 0.2701,
      "num_input_tokens_seen": 1762946,
      "step": 945,
      "train_runtime": 1972.804,
      "train_tokens_per_second": 893.624
    },
    {
      "epoch": 0.9562800101086681,
      "grad_norm": 0.4730886220932007,
      "learning_rate": 0.0001048101265822785,
      "loss": 0.2201,
      "num_input_tokens_seen": 1764748,
      "step": 946,
      "train_runtime": 1974.5699,
      "train_tokens_per_second": 893.738
    },
    {
      "epoch": 0.9572908769269649,
      "grad_norm": 0.40383127331733704,
      "learning_rate": 0.00010470886075949367,
      "loss": 0.1941,
      "num_input_tokens_seen": 1766552,
      "step": 947,
      "train_runtime": 1976.2317,
      "train_tokens_per_second": 893.899
    },
    {
      "epoch": 0.9583017437452616,
      "grad_norm": 0.4525023102760315,
      "learning_rate": 0.00010460759493670886,
      "loss": 0.1174,
      "num_input_tokens_seen": 1768384,
      "step": 948,
      "train_runtime": 1978.0176,
      "train_tokens_per_second": 894.018
    },
    {
      "epoch": 0.9593126105635582,
      "grad_norm": 0.4887596070766449,
      "learning_rate": 0.00010450632911392407,
      "loss": 0.2909,
      "num_input_tokens_seen": 1770632,
      "step": 949,
      "train_runtime": 1979.9106,
      "train_tokens_per_second": 894.299
    },
    {
      "epoch": 0.960323477381855,
      "grad_norm": 0.41555631160736084,
      "learning_rate": 0.00010440506329113924,
      "loss": 0.1795,
      "num_input_tokens_seen": 1772254,
      "step": 950,
      "train_runtime": 1981.5407,
      "train_tokens_per_second": 894.382
    },
    {
      "epoch": 0.9613343442001516,
      "grad_norm": 0.47362515330314636,
      "learning_rate": 0.00010430379746835443,
      "loss": 0.1567,
      "num_input_tokens_seen": 1774112,
      "step": 951,
      "train_runtime": 1983.2649,
      "train_tokens_per_second": 894.541
    },
    {
      "epoch": 0.9623452110184483,
      "grad_norm": 0.31603702902793884,
      "learning_rate": 0.00010420253164556964,
      "loss": 0.2471,
      "num_input_tokens_seen": 1776174,
      "step": 952,
      "train_runtime": 1985.193,
      "train_tokens_per_second": 894.711
    },
    {
      "epoch": 0.963356077836745,
      "grad_norm": 0.5144686102867126,
      "learning_rate": 0.00010410126582278482,
      "loss": 0.1522,
      "num_input_tokens_seen": 1777928,
      "step": 953,
      "train_runtime": 1986.8906,
      "train_tokens_per_second": 894.829
    },
    {
      "epoch": 0.9643669446550417,
      "grad_norm": 0.4585995674133301,
      "learning_rate": 0.00010400000000000001,
      "loss": 0.2274,
      "num_input_tokens_seen": 1779882,
      "step": 954,
      "train_runtime": 1988.6373,
      "train_tokens_per_second": 895.026
    },
    {
      "epoch": 0.9653778114733383,
      "grad_norm": 0.4180617332458496,
      "learning_rate": 0.00010389873417721518,
      "loss": 0.1751,
      "num_input_tokens_seen": 1782042,
      "step": 955,
      "train_runtime": 1990.5086,
      "train_tokens_per_second": 895.27
    },
    {
      "epoch": 0.9663886782916351,
      "grad_norm": 0.4923889636993408,
      "learning_rate": 0.00010379746835443039,
      "loss": 0.2123,
      "num_input_tokens_seen": 1783590,
      "step": 956,
      "train_runtime": 1992.1379,
      "train_tokens_per_second": 895.315
    },
    {
      "epoch": 0.9673995451099318,
      "grad_norm": 0.3163554072380066,
      "learning_rate": 0.00010369620253164558,
      "loss": 0.1705,
      "num_input_tokens_seen": 1785550,
      "step": 957,
      "train_runtime": 1993.9263,
      "train_tokens_per_second": 895.494
    },
    {
      "epoch": 0.9684104119282284,
      "grad_norm": 0.4780999720096588,
      "learning_rate": 0.00010359493670886076,
      "loss": 0.1776,
      "num_input_tokens_seen": 1787542,
      "step": 958,
      "train_runtime": 1995.6758,
      "train_tokens_per_second": 895.708
    },
    {
      "epoch": 0.9694212787465252,
      "grad_norm": 0.40335267782211304,
      "learning_rate": 0.00010349367088607596,
      "loss": 0.2025,
      "num_input_tokens_seen": 1789302,
      "step": 959,
      "train_runtime": 1997.4394,
      "train_tokens_per_second": 895.798
    },
    {
      "epoch": 0.9704321455648218,
      "grad_norm": 0.38821494579315186,
      "learning_rate": 0.00010339240506329115,
      "loss": 0.157,
      "num_input_tokens_seen": 1790948,
      "step": 960,
      "train_runtime": 1999.1173,
      "train_tokens_per_second": 895.869
    },
    {
      "epoch": 0.9714430123831185,
      "grad_norm": 0.4874069094657898,
      "learning_rate": 0.00010329113924050633,
      "loss": 0.1973,
      "num_input_tokens_seen": 1792550,
      "step": 961,
      "train_runtime": 2002.5561,
      "train_tokens_per_second": 895.131
    },
    {
      "epoch": 0.9724538792014152,
      "grad_norm": 0.47805511951446533,
      "learning_rate": 0.00010318987341772153,
      "loss": 0.2067,
      "num_input_tokens_seen": 1794244,
      "step": 962,
      "train_runtime": 2004.2911,
      "train_tokens_per_second": 895.201
    },
    {
      "epoch": 0.9734647460197119,
      "grad_norm": 0.4922966957092285,
      "learning_rate": 0.0001030886075949367,
      "loss": 0.217,
      "num_input_tokens_seen": 1796822,
      "step": 963,
      "train_runtime": 2006.3891,
      "train_tokens_per_second": 895.55
    },
    {
      "epoch": 0.9744756128380085,
      "grad_norm": 0.4031835198402405,
      "learning_rate": 0.0001029873417721519,
      "loss": 0.1446,
      "num_input_tokens_seen": 1798388,
      "step": 964,
      "train_runtime": 2008.0338,
      "train_tokens_per_second": 895.596
    },
    {
      "epoch": 0.9754864796563053,
      "grad_norm": 0.38097310066223145,
      "learning_rate": 0.0001028860759493671,
      "loss": 0.1828,
      "num_input_tokens_seen": 1799948,
      "step": 965,
      "train_runtime": 2009.7162,
      "train_tokens_per_second": 895.623
    },
    {
      "epoch": 0.976497346474602,
      "grad_norm": 0.46160629391670227,
      "learning_rate": 0.00010278481012658227,
      "loss": 0.1712,
      "num_input_tokens_seen": 1801522,
      "step": 966,
      "train_runtime": 2011.3349,
      "train_tokens_per_second": 895.685
    },
    {
      "epoch": 0.9775082132928986,
      "grad_norm": 0.38964974880218506,
      "learning_rate": 0.00010268354430379747,
      "loss": 0.1581,
      "num_input_tokens_seen": 1803924,
      "step": 967,
      "train_runtime": 2013.3514,
      "train_tokens_per_second": 895.981
    },
    {
      "epoch": 0.9785190801111954,
      "grad_norm": 0.5489251613616943,
      "learning_rate": 0.00010258227848101268,
      "loss": 0.1597,
      "num_input_tokens_seen": 1805646,
      "step": 968,
      "train_runtime": 2014.9986,
      "train_tokens_per_second": 896.103
    },
    {
      "epoch": 0.979529946929492,
      "grad_norm": 0.4700818359851837,
      "learning_rate": 0.00010248101265822784,
      "loss": 0.1419,
      "num_input_tokens_seen": 1807420,
      "step": 969,
      "train_runtime": 2016.6913,
      "train_tokens_per_second": 896.23
    },
    {
      "epoch": 0.9805408137477887,
      "grad_norm": 0.41527441143989563,
      "learning_rate": 0.00010237974683544305,
      "loss": 0.1869,
      "num_input_tokens_seen": 1809092,
      "step": 970,
      "train_runtime": 2018.3856,
      "train_tokens_per_second": 896.306
    },
    {
      "epoch": 0.9815516805660854,
      "grad_norm": 0.4509820342063904,
      "learning_rate": 0.00010227848101265822,
      "loss": 0.2115,
      "num_input_tokens_seen": 1810918,
      "step": 971,
      "train_runtime": 2020.1121,
      "train_tokens_per_second": 896.444
    },
    {
      "epoch": 0.9825625473843821,
      "grad_norm": 0.42594748735427856,
      "learning_rate": 0.00010217721518987342,
      "loss": 0.1725,
      "num_input_tokens_seen": 1812822,
      "step": 972,
      "train_runtime": 2021.8401,
      "train_tokens_per_second": 896.62
    },
    {
      "epoch": 0.9835734142026787,
      "grad_norm": 0.3903284966945648,
      "learning_rate": 0.00010207594936708862,
      "loss": 0.2107,
      "num_input_tokens_seen": 1814844,
      "step": 973,
      "train_runtime": 2023.6509,
      "train_tokens_per_second": 896.817
    },
    {
      "epoch": 0.9845842810209755,
      "grad_norm": 0.4419352412223816,
      "learning_rate": 0.0001019746835443038,
      "loss": 0.2286,
      "num_input_tokens_seen": 1816696,
      "step": 974,
      "train_runtime": 2025.4299,
      "train_tokens_per_second": 896.943
    },
    {
      "epoch": 0.9855951478392722,
      "grad_norm": 0.32668042182922363,
      "learning_rate": 0.00010187341772151899,
      "loss": 0.1398,
      "num_input_tokens_seen": 1818574,
      "step": 975,
      "train_runtime": 2027.2656,
      "train_tokens_per_second": 897.058
    },
    {
      "epoch": 0.9866060146575689,
      "grad_norm": 0.46488189697265625,
      "learning_rate": 0.00010177215189873419,
      "loss": 0.1968,
      "num_input_tokens_seen": 1820238,
      "step": 976,
      "train_runtime": 2028.9558,
      "train_tokens_per_second": 897.13
    },
    {
      "epoch": 0.9876168814758656,
      "grad_norm": 0.41473039984703064,
      "learning_rate": 0.00010167088607594937,
      "loss": 0.1952,
      "num_input_tokens_seen": 1822212,
      "step": 977,
      "train_runtime": 2030.7573,
      "train_tokens_per_second": 897.307
    },
    {
      "epoch": 0.9886277482941622,
      "grad_norm": 0.5555244088172913,
      "learning_rate": 0.00010156962025316456,
      "loss": 0.2162,
      "num_input_tokens_seen": 1824152,
      "step": 978,
      "train_runtime": 2032.5897,
      "train_tokens_per_second": 897.452
    },
    {
      "epoch": 0.989638615112459,
      "grad_norm": 0.5394989252090454,
      "learning_rate": 0.00010146835443037974,
      "loss": 0.1803,
      "num_input_tokens_seen": 1825594,
      "step": 979,
      "train_runtime": 2034.1999,
      "train_tokens_per_second": 897.451
    },
    {
      "epoch": 0.9906494819307556,
      "grad_norm": 0.531579852104187,
      "learning_rate": 0.00010136708860759494,
      "loss": 0.2835,
      "num_input_tokens_seen": 1827552,
      "step": 980,
      "train_runtime": 2035.9638,
      "train_tokens_per_second": 897.635
    },
    {
      "epoch": 0.9916603487490523,
      "grad_norm": 0.43535366654396057,
      "learning_rate": 0.00010126582278481013,
      "loss": 0.1601,
      "num_input_tokens_seen": 1829358,
      "step": 981,
      "train_runtime": 2037.686,
      "train_tokens_per_second": 897.762
    },
    {
      "epoch": 0.992671215567349,
      "grad_norm": 0.5188285708427429,
      "learning_rate": 0.00010116455696202531,
      "loss": 0.1547,
      "num_input_tokens_seen": 1830736,
      "step": 982,
      "train_runtime": 2039.2779,
      "train_tokens_per_second": 897.737
    },
    {
      "epoch": 0.9936820823856457,
      "grad_norm": 0.4486296474933624,
      "learning_rate": 0.00010106329113924051,
      "loss": 0.3112,
      "num_input_tokens_seen": 1833394,
      "step": 983,
      "train_runtime": 2041.4343,
      "train_tokens_per_second": 898.091
    },
    {
      "epoch": 0.9946929492039424,
      "grad_norm": 0.4547290802001953,
      "learning_rate": 0.0001009620253164557,
      "loss": 0.2592,
      "num_input_tokens_seen": 1835190,
      "step": 984,
      "train_runtime": 2043.195,
      "train_tokens_per_second": 898.196
    },
    {
      "epoch": 0.9957038160222391,
      "grad_norm": 0.5215474367141724,
      "learning_rate": 0.00010086075949367088,
      "loss": 0.1686,
      "num_input_tokens_seen": 1836778,
      "step": 985,
      "train_runtime": 2044.8837,
      "train_tokens_per_second": 898.231
    },
    {
      "epoch": 0.9967146828405358,
      "grad_norm": 0.5124213695526123,
      "learning_rate": 0.00010075949367088609,
      "loss": 0.1822,
      "num_input_tokens_seen": 1838628,
      "step": 986,
      "train_runtime": 2046.6339,
      "train_tokens_per_second": 898.367
    },
    {
      "epoch": 0.9977255496588324,
      "grad_norm": 0.4229728877544403,
      "learning_rate": 0.00010065822784810126,
      "loss": 0.1431,
      "num_input_tokens_seen": 1840538,
      "step": 987,
      "train_runtime": 2048.4102,
      "train_tokens_per_second": 898.52
    },
    {
      "epoch": 0.9987364164771292,
      "grad_norm": 0.45670852065086365,
      "learning_rate": 0.00010055696202531646,
      "loss": 0.1699,
      "num_input_tokens_seen": 1842408,
      "step": 988,
      "train_runtime": 2050.1812,
      "train_tokens_per_second": 898.656
    },
    {
      "epoch": 0.9997472832954258,
      "grad_norm": 0.5166110992431641,
      "learning_rate": 0.00010045569620253166,
      "loss": 0.2698,
      "num_input_tokens_seen": 1844106,
      "step": 989,
      "train_runtime": 2051.8893,
      "train_tokens_per_second": 898.736
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7430378794670105,
      "learning_rate": 0.00010035443037974684,
      "loss": 0.0709,
      "num_input_tokens_seen": 1844288,
      "step": 990,
      "train_runtime": 2053.7136,
      "train_tokens_per_second": 898.026
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.19176971912384033,
      "eval_runtime": 61.5512,
      "eval_samples_per_second": 14.281,
      "eval_steps_per_second": 7.149,
      "num_input_tokens_seen": 1844288,
      "step": 990
    },
    {
      "epoch": 1.0010108668182967,
      "grad_norm": 0.24361146986484528,
      "learning_rate": 0.00010025316455696203,
      "loss": 0.046,
      "num_input_tokens_seen": 1845740,
      "step": 991,
      "train_runtime": 2119.0572,
      "train_tokens_per_second": 871.019
    },
    {
      "epoch": 1.0020217336365933,
      "grad_norm": 0.3321000635623932,
      "learning_rate": 0.00010015189873417723,
      "loss": 0.094,
      "num_input_tokens_seen": 1847776,
      "step": 992,
      "train_runtime": 2120.9125,
      "train_tokens_per_second": 871.217
    },
    {
      "epoch": 1.0030326004548902,
      "grad_norm": 0.27356022596359253,
      "learning_rate": 0.00010005063291139241,
      "loss": 0.062,
      "num_input_tokens_seen": 1849706,
      "step": 993,
      "train_runtime": 2122.7041,
      "train_tokens_per_second": 871.391
    },
    {
      "epoch": 1.0040434672731868,
      "grad_norm": 0.2503027617931366,
      "learning_rate": 9.99493670886076e-05,
      "loss": 0.1359,
      "num_input_tokens_seen": 1851240,
      "step": 994,
      "train_runtime": 2124.317,
      "train_tokens_per_second": 871.452
    },
    {
      "epoch": 1.0050543340914835,
      "grad_norm": 0.2093895673751831,
      "learning_rate": 9.984810126582279e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 1853212,
      "step": 995,
      "train_runtime": 2126.1284,
      "train_tokens_per_second": 871.637
    },
    {
      "epoch": 1.0060652009097801,
      "grad_norm": 0.2690771520137787,
      "learning_rate": 9.974683544303798e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 1855100,
      "step": 996,
      "train_runtime": 2127.8927,
      "train_tokens_per_second": 871.801
    },
    {
      "epoch": 1.0070760677280768,
      "grad_norm": 0.22197206318378448,
      "learning_rate": 9.964556962025316e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 1856994,
      "step": 997,
      "train_runtime": 2129.6679,
      "train_tokens_per_second": 871.964
    },
    {
      "epoch": 1.0080869345463734,
      "grad_norm": 0.34467798471450806,
      "learning_rate": 9.954430379746836e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 1858826,
      "step": 998,
      "train_runtime": 2131.4066,
      "train_tokens_per_second": 872.112
    },
    {
      "epoch": 1.0090978013646703,
      "grad_norm": 0.20582760870456696,
      "learning_rate": 9.944303797468355e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 1860690,
      "step": 999,
      "train_runtime": 2133.1821,
      "train_tokens_per_second": 872.26
    },
    {
      "epoch": 1.010108668182967,
      "grad_norm": 0.3075924217700958,
      "learning_rate": 9.934177215189873e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 1862246,
      "step": 1000,
      "train_runtime": 2134.804,
      "train_tokens_per_second": 872.326
    },
    {
      "epoch": 1.0111195350012636,
      "grad_norm": 0.2964280843734741,
      "learning_rate": 9.924050632911392e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 1863944,
      "step": 1001,
      "train_runtime": 2136.4715,
      "train_tokens_per_second": 872.44
    },
    {
      "epoch": 1.0121304018195603,
      "grad_norm": 0.3516748547554016,
      "learning_rate": 9.913924050632913e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 1865368,
      "step": 1002,
      "train_runtime": 2138.0769,
      "train_tokens_per_second": 872.451
    },
    {
      "epoch": 1.013141268637857,
      "grad_norm": 0.2860864996910095,
      "learning_rate": 9.90379746835443e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 1867136,
      "step": 1003,
      "train_runtime": 2139.7771,
      "train_tokens_per_second": 872.584
    },
    {
      "epoch": 1.0141521354561536,
      "grad_norm": 0.37862107157707214,
      "learning_rate": 9.89367088607595e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 1869088,
      "step": 1004,
      "train_runtime": 2141.5819,
      "train_tokens_per_second": 872.76
    },
    {
      "epoch": 1.0151630022744504,
      "grad_norm": 0.3571180999279022,
      "learning_rate": 9.883544303797469e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 1871130,
      "step": 1005,
      "train_runtime": 2143.4425,
      "train_tokens_per_second": 872.956
    },
    {
      "epoch": 1.016173869092747,
      "grad_norm": 0.26994094252586365,
      "learning_rate": 9.873417721518988e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 1872784,
      "step": 1006,
      "train_runtime": 2145.1338,
      "train_tokens_per_second": 873.038
    },
    {
      "epoch": 1.0171847359110437,
      "grad_norm": 0.2408159077167511,
      "learning_rate": 9.863291139240507e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 1874410,
      "step": 1007,
      "train_runtime": 2146.7525,
      "train_tokens_per_second": 873.137
    },
    {
      "epoch": 1.0181956027293404,
      "grad_norm": 0.3475908637046814,
      "learning_rate": 9.853164556962026e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 1876104,
      "step": 1008,
      "train_runtime": 2148.4439,
      "train_tokens_per_second": 873.239
    },
    {
      "epoch": 1.019206469547637,
      "grad_norm": 0.40700891613960266,
      "learning_rate": 9.843037974683545e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 1878062,
      "step": 1009,
      "train_runtime": 2150.2693,
      "train_tokens_per_second": 873.408
    },
    {
      "epoch": 1.0202173363659337,
      "grad_norm": 0.3501214385032654,
      "learning_rate": 9.832911392405064e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 1879806,
      "step": 1010,
      "train_runtime": 2151.9607,
      "train_tokens_per_second": 873.532
    },
    {
      "epoch": 1.0212282031842306,
      "grad_norm": 0.40707865357398987,
      "learning_rate": 9.822784810126583e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 1881208,
      "step": 1011,
      "train_runtime": 2153.5599,
      "train_tokens_per_second": 873.534
    },
    {
      "epoch": 1.0222390700025272,
      "grad_norm": 0.34942948818206787,
      "learning_rate": 9.812658227848102e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 1883406,
      "step": 1012,
      "train_runtime": 2155.4902,
      "train_tokens_per_second": 873.772
    },
    {
      "epoch": 1.0232499368208239,
      "grad_norm": 0.3891485035419464,
      "learning_rate": 9.80253164556962e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 1885370,
      "step": 1013,
      "train_runtime": 2157.2709,
      "train_tokens_per_second": 873.961
    },
    {
      "epoch": 1.0242608036391205,
      "grad_norm": 0.5246986150741577,
      "learning_rate": 9.79240506329114e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 1886950,
      "step": 1014,
      "train_runtime": 2158.9591,
      "train_tokens_per_second": 874.009
    },
    {
      "epoch": 1.0252716704574172,
      "grad_norm": 0.5961692929267883,
      "learning_rate": 9.78227848101266e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 1888814,
      "step": 1015,
      "train_runtime": 2160.6831,
      "train_tokens_per_second": 874.174
    },
    {
      "epoch": 1.0262825372757138,
      "grad_norm": 0.4271422326564789,
      "learning_rate": 9.772151898734177e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 1890906,
      "step": 1016,
      "train_runtime": 2162.5297,
      "train_tokens_per_second": 874.395
    },
    {
      "epoch": 1.0272934040940107,
      "grad_norm": 0.4589087665081024,
      "learning_rate": 9.762025316455696e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 1892428,
      "step": 1017,
      "train_runtime": 2164.1205,
      "train_tokens_per_second": 874.456
    },
    {
      "epoch": 1.0283042709123074,
      "grad_norm": 0.24373966455459595,
      "learning_rate": 9.751898734177217e-05,
      "loss": 0.104,
      "num_input_tokens_seen": 1893974,
      "step": 1018,
      "train_runtime": 2165.709,
      "train_tokens_per_second": 874.528
    },
    {
      "epoch": 1.029315137730604,
      "grad_norm": 0.45167598128318787,
      "learning_rate": 9.741772151898734e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 1896382,
      "step": 1019,
      "train_runtime": 2167.6871,
      "train_tokens_per_second": 874.841
    },
    {
      "epoch": 1.0303260045489007,
      "grad_norm": 0.27746036648750305,
      "learning_rate": 9.731645569620254e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 1898498,
      "step": 1020,
      "train_runtime": 2169.5488,
      "train_tokens_per_second": 875.066
    },
    {
      "epoch": 1.0313368713671973,
      "grad_norm": 0.39703670144081116,
      "learning_rate": 9.721518987341773e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 1900310,
      "step": 1021,
      "train_runtime": 2173.0011,
      "train_tokens_per_second": 874.509
    },
    {
      "epoch": 1.032347738185494,
      "grad_norm": 0.3907689154148102,
      "learning_rate": 9.711392405063292e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 1902106,
      "step": 1022,
      "train_runtime": 2174.6974,
      "train_tokens_per_second": 874.653
    },
    {
      "epoch": 1.0333586050037908,
      "grad_norm": 0.3423595428466797,
      "learning_rate": 9.701265822784811e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 1903878,
      "step": 1023,
      "train_runtime": 2176.3963,
      "train_tokens_per_second": 874.785
    },
    {
      "epoch": 1.0343694718220875,
      "grad_norm": 0.42263948917388916,
      "learning_rate": 9.69113924050633e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 1905900,
      "step": 1024,
      "train_runtime": 2178.2753,
      "train_tokens_per_second": 874.958
    },
    {
      "epoch": 1.0353803386403841,
      "grad_norm": 0.27596285939216614,
      "learning_rate": 9.681012658227848e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 1907698,
      "step": 1025,
      "train_runtime": 2179.9902,
      "train_tokens_per_second": 875.095
    },
    {
      "epoch": 1.0363912054586808,
      "grad_norm": 0.3455818295478821,
      "learning_rate": 9.670886075949368e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 1910000,
      "step": 1026,
      "train_runtime": 2181.9243,
      "train_tokens_per_second": 875.374
    },
    {
      "epoch": 1.0374020722769774,
      "grad_norm": 0.3340929448604584,
      "learning_rate": 9.660759493670887e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 1912230,
      "step": 1027,
      "train_runtime": 2183.7967,
      "train_tokens_per_second": 875.645
    },
    {
      "epoch": 1.038412939095274,
      "grad_norm": 0.40362825989723206,
      "learning_rate": 9.650632911392405e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 1914156,
      "step": 1028,
      "train_runtime": 2185.6328,
      "train_tokens_per_second": 875.79
    },
    {
      "epoch": 1.039423805913571,
      "grad_norm": 0.2972779870033264,
      "learning_rate": 9.640506329113924e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 1916246,
      "step": 1029,
      "train_runtime": 2187.4457,
      "train_tokens_per_second": 876.02
    },
    {
      "epoch": 1.0404346727318676,
      "grad_norm": 0.4286695122718811,
      "learning_rate": 9.630379746835444e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 1918102,
      "step": 1030,
      "train_runtime": 2189.1716,
      "train_tokens_per_second": 876.177
    },
    {
      "epoch": 1.0414455395501643,
      "grad_norm": 0.4273298382759094,
      "learning_rate": 9.620253164556962e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 1919908,
      "step": 1031,
      "train_runtime": 2190.8848,
      "train_tokens_per_second": 876.316
    },
    {
      "epoch": 1.042456406368461,
      "grad_norm": 0.36308878660202026,
      "learning_rate": 9.610126582278481e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 1921780,
      "step": 1032,
      "train_runtime": 2192.6132,
      "train_tokens_per_second": 876.479
    },
    {
      "epoch": 1.0434672731867576,
      "grad_norm": 0.33895182609558105,
      "learning_rate": 9.6e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 1923268,
      "step": 1033,
      "train_runtime": 2194.2234,
      "train_tokens_per_second": 876.514
    },
    {
      "epoch": 1.0444781400050542,
      "grad_norm": 0.2617810368537903,
      "learning_rate": 9.58987341772152e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 1925508,
      "step": 1034,
      "train_runtime": 2196.1433,
      "train_tokens_per_second": 876.768
    },
    {
      "epoch": 1.045489006823351,
      "grad_norm": 0.2915429472923279,
      "learning_rate": 9.579746835443038e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 1927592,
      "step": 1035,
      "train_runtime": 2198.0043,
      "train_tokens_per_second": 876.974
    },
    {
      "epoch": 1.0464998736416478,
      "grad_norm": 0.30873245000839233,
      "learning_rate": 9.569620253164557e-05,
      "loss": 0.0949,
      "num_input_tokens_seen": 1929746,
      "step": 1036,
      "train_runtime": 2199.9238,
      "train_tokens_per_second": 877.188
    },
    {
      "epoch": 1.0475107404599444,
      "grad_norm": 0.359919935464859,
      "learning_rate": 9.559493670886077e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 1931610,
      "step": 1037,
      "train_runtime": 2201.6888,
      "train_tokens_per_second": 877.331
    },
    {
      "epoch": 1.048521607278241,
      "grad_norm": 0.2546060383319855,
      "learning_rate": 9.549367088607596e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 1933750,
      "step": 1038,
      "train_runtime": 2203.5409,
      "train_tokens_per_second": 877.565
    },
    {
      "epoch": 1.0495324740965377,
      "grad_norm": 0.3829679787158966,
      "learning_rate": 9.539240506329115e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 1935630,
      "step": 1039,
      "train_runtime": 2205.3721,
      "train_tokens_per_second": 877.689
    },
    {
      "epoch": 1.0505433409148344,
      "grad_norm": 0.4177955985069275,
      "learning_rate": 9.529113924050634e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 1937598,
      "step": 1040,
      "train_runtime": 2207.165,
      "train_tokens_per_second": 877.867
    },
    {
      "epoch": 1.0515542077331312,
      "grad_norm": 0.45883163809776306,
      "learning_rate": 9.518987341772152e-05,
      "loss": 0.1335,
      "num_input_tokens_seen": 1939094,
      "step": 1041,
      "train_runtime": 2208.789,
      "train_tokens_per_second": 877.899
    },
    {
      "epoch": 1.052565074551428,
      "grad_norm": 0.4552958905696869,
      "learning_rate": 9.508860759493672e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 1941070,
      "step": 1042,
      "train_runtime": 2210.5737,
      "train_tokens_per_second": 878.084
    },
    {
      "epoch": 1.0535759413697245,
      "grad_norm": 0.18470849096775055,
      "learning_rate": 9.498734177215191e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 1942698,
      "step": 1043,
      "train_runtime": 2212.2018,
      "train_tokens_per_second": 878.174
    },
    {
      "epoch": 1.0545868081880212,
      "grad_norm": 0.4260298013687134,
      "learning_rate": 9.488607594936709e-05,
      "loss": 0.1243,
      "num_input_tokens_seen": 1944222,
      "step": 1044,
      "train_runtime": 2213.8126,
      "train_tokens_per_second": 878.223
    },
    {
      "epoch": 1.0555976750063178,
      "grad_norm": 0.3968278169631958,
      "learning_rate": 9.478481012658228e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 1946298,
      "step": 1045,
      "train_runtime": 2215.6159,
      "train_tokens_per_second": 878.446
    },
    {
      "epoch": 1.0566085418246145,
      "grad_norm": 0.2895992696285248,
      "learning_rate": 9.468354430379748e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 1947932,
      "step": 1046,
      "train_runtime": 2217.3417,
      "train_tokens_per_second": 878.499
    },
    {
      "epoch": 1.0576194086429114,
      "grad_norm": 0.4029696583747864,
      "learning_rate": 9.458227848101266e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 1950306,
      "step": 1047,
      "train_runtime": 2219.2727,
      "train_tokens_per_second": 878.804
    },
    {
      "epoch": 1.058630275461208,
      "grad_norm": 0.3831733167171478,
      "learning_rate": 9.448101265822785e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 1951976,
      "step": 1048,
      "train_runtime": 2220.9575,
      "train_tokens_per_second": 878.889
    },
    {
      "epoch": 1.0596411422795047,
      "grad_norm": 0.3044146001338959,
      "learning_rate": 9.437974683544304e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 1953982,
      "step": 1049,
      "train_runtime": 2222.8203,
      "train_tokens_per_second": 879.055
    },
    {
      "epoch": 1.0606520090978013,
      "grad_norm": 0.36417829990386963,
      "learning_rate": 9.427848101265823e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 1956104,
      "step": 1050,
      "train_runtime": 2224.6377,
      "train_tokens_per_second": 879.291
    },
    {
      "epoch": 1.061662875916098,
      "grad_norm": 0.2965594530105591,
      "learning_rate": 9.417721518987342e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 1957770,
      "step": 1051,
      "train_runtime": 2228.4921,
      "train_tokens_per_second": 878.518
    },
    {
      "epoch": 1.0626737427343946,
      "grad_norm": 0.3719140589237213,
      "learning_rate": 9.407594936708861e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 1959422,
      "step": 1052,
      "train_runtime": 2230.1841,
      "train_tokens_per_second": 878.592
    },
    {
      "epoch": 1.0636846095526915,
      "grad_norm": 0.3414093852043152,
      "learning_rate": 9.397468354430379e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 1961164,
      "step": 1053,
      "train_runtime": 2231.9,
      "train_tokens_per_second": 878.697
    },
    {
      "epoch": 1.0646954763709882,
      "grad_norm": 0.2993382513523102,
      "learning_rate": 9.3873417721519e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 1962830,
      "step": 1054,
      "train_runtime": 2233.5124,
      "train_tokens_per_second": 878.809
    },
    {
      "epoch": 1.0657063431892848,
      "grad_norm": 0.3396962881088257,
      "learning_rate": 9.377215189873419e-05,
      "loss": 0.1145,
      "num_input_tokens_seen": 1965386,
      "step": 1055,
      "train_runtime": 2235.5603,
      "train_tokens_per_second": 879.147
    },
    {
      "epoch": 1.0667172100075815,
      "grad_norm": 0.3782525658607483,
      "learning_rate": 9.367088607594936e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 1967074,
      "step": 1056,
      "train_runtime": 2237.2015,
      "train_tokens_per_second": 879.257
    },
    {
      "epoch": 1.0677280768258781,
      "grad_norm": 0.5178813934326172,
      "learning_rate": 9.356962025316456e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 1969174,
      "step": 1057,
      "train_runtime": 2239.1319,
      "train_tokens_per_second": 879.436
    },
    {
      "epoch": 1.068738943644175,
      "grad_norm": 0.31324502825737,
      "learning_rate": 9.346835443037976e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 1970990,
      "step": 1058,
      "train_runtime": 2240.8356,
      "train_tokens_per_second": 879.578
    },
    {
      "epoch": 1.0697498104624716,
      "grad_norm": 0.26674196124076843,
      "learning_rate": 9.336708860759494e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 1972654,
      "step": 1059,
      "train_runtime": 2242.4661,
      "train_tokens_per_second": 879.681
    },
    {
      "epoch": 1.0707606772807683,
      "grad_norm": 0.3870038092136383,
      "learning_rate": 9.326582278481013e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 1974310,
      "step": 1060,
      "train_runtime": 2244.0988,
      "train_tokens_per_second": 879.779
    },
    {
      "epoch": 1.071771544099065,
      "grad_norm": 0.27661511301994324,
      "learning_rate": 9.316455696202532e-05,
      "loss": 0.1257,
      "num_input_tokens_seen": 1976672,
      "step": 1061,
      "train_runtime": 2246.1201,
      "train_tokens_per_second": 880.038
    },
    {
      "epoch": 1.0727824109173616,
      "grad_norm": 0.45711833238601685,
      "learning_rate": 9.306329113924051e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 1978514,
      "step": 1062,
      "train_runtime": 2247.8339,
      "train_tokens_per_second": 880.187
    },
    {
      "epoch": 1.0737932777356582,
      "grad_norm": 0.3385152816772461,
      "learning_rate": 9.29620253164557e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 1979866,
      "step": 1063,
      "train_runtime": 2249.4216,
      "train_tokens_per_second": 880.167
    },
    {
      "epoch": 1.074804144553955,
      "grad_norm": 0.5051232576370239,
      "learning_rate": 9.286075949367089e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 1981886,
      "step": 1064,
      "train_runtime": 2251.2888,
      "train_tokens_per_second": 880.334
    },
    {
      "epoch": 1.0758150113722518,
      "grad_norm": 0.4586871266365051,
      "learning_rate": 9.275949367088608e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 1983562,
      "step": 1065,
      "train_runtime": 2252.9276,
      "train_tokens_per_second": 880.438
    },
    {
      "epoch": 1.0768258781905484,
      "grad_norm": 0.2923198938369751,
      "learning_rate": 9.265822784810127e-05,
      "loss": 0.1598,
      "num_input_tokens_seen": 1985662,
      "step": 1066,
      "train_runtime": 2254.7797,
      "train_tokens_per_second": 880.646
    },
    {
      "epoch": 1.077836745008845,
      "grad_norm": 0.297481507062912,
      "learning_rate": 9.255696202531646e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 1987774,
      "step": 1067,
      "train_runtime": 2256.5919,
      "train_tokens_per_second": 880.874
    },
    {
      "epoch": 1.0788476118271417,
      "grad_norm": 0.2412688434123993,
      "learning_rate": 9.245569620253165e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 1989582,
      "step": 1068,
      "train_runtime": 2258.3519,
      "train_tokens_per_second": 880.988
    },
    {
      "epoch": 1.0798584786454384,
      "grad_norm": 0.31675273180007935,
      "learning_rate": 9.235443037974683e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 1991786,
      "step": 1069,
      "train_runtime": 2260.2833,
      "train_tokens_per_second": 881.211
    },
    {
      "epoch": 1.0808693454637353,
      "grad_norm": 0.3629102408885956,
      "learning_rate": 9.225316455696204e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 1994004,
      "step": 1070,
      "train_runtime": 2262.153,
      "train_tokens_per_second": 881.463
    },
    {
      "epoch": 1.081880212282032,
      "grad_norm": 0.45810672640800476,
      "learning_rate": 9.215189873417723e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 1995720,
      "step": 1071,
      "train_runtime": 2263.8917,
      "train_tokens_per_second": 881.544
    },
    {
      "epoch": 1.0828910791003286,
      "grad_norm": 0.38692861795425415,
      "learning_rate": 9.20506329113924e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 1997650,
      "step": 1072,
      "train_runtime": 2265.6152,
      "train_tokens_per_second": 881.725
    },
    {
      "epoch": 1.0839019459186252,
      "grad_norm": 0.34584349393844604,
      "learning_rate": 9.19493670886076e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 1999524,
      "step": 1073,
      "train_runtime": 2267.2896,
      "train_tokens_per_second": 881.901
    },
    {
      "epoch": 1.0849128127369219,
      "grad_norm": 0.3601796627044678,
      "learning_rate": 9.18481012658228e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 2001770,
      "step": 1074,
      "train_runtime": 2269.2289,
      "train_tokens_per_second": 882.137
    },
    {
      "epoch": 1.0859236795552185,
      "grad_norm": 0.32827287912368774,
      "learning_rate": 9.174683544303798e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 2003678,
      "step": 1075,
      "train_runtime": 2270.9486,
      "train_tokens_per_second": 882.309
    },
    {
      "epoch": 1.0869345463735154,
      "grad_norm": 0.3239614963531494,
      "learning_rate": 9.164556962025317e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 2005346,
      "step": 1076,
      "train_runtime": 2272.5704,
      "train_tokens_per_second": 882.413
    },
    {
      "epoch": 1.087945413191812,
      "grad_norm": 0.4321143627166748,
      "learning_rate": 9.154430379746836e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 2006840,
      "step": 1077,
      "train_runtime": 2274.1903,
      "train_tokens_per_second": 882.442
    },
    {
      "epoch": 1.0889562800101087,
      "grad_norm": 0.2403334379196167,
      "learning_rate": 9.144303797468355e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 2008588,
      "step": 1078,
      "train_runtime": 2275.966,
      "train_tokens_per_second": 882.521
    },
    {
      "epoch": 1.0899671468284053,
      "grad_norm": 0.3006885051727295,
      "learning_rate": 9.134177215189874e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 2010390,
      "step": 1079,
      "train_runtime": 2277.6646,
      "train_tokens_per_second": 882.654
    },
    {
      "epoch": 1.090978013646702,
      "grad_norm": 0.2884374260902405,
      "learning_rate": 9.124050632911393e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 2012638,
      "step": 1080,
      "train_runtime": 2279.6622,
      "train_tokens_per_second": 882.867
    },
    {
      "epoch": 1.0919888804649986,
      "grad_norm": 0.3214394152164459,
      "learning_rate": 9.113924050632912e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 2014266,
      "step": 1081,
      "train_runtime": 2283.3275,
      "train_tokens_per_second": 882.163
    },
    {
      "epoch": 1.0929997472832955,
      "grad_norm": 0.39283737540245056,
      "learning_rate": 9.103797468354431e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 2016072,
      "step": 1082,
      "train_runtime": 2285.0962,
      "train_tokens_per_second": 882.27
    },
    {
      "epoch": 1.0940106141015922,
      "grad_norm": 0.3431110382080078,
      "learning_rate": 9.09367088607595e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 2017770,
      "step": 1083,
      "train_runtime": 2286.7899,
      "train_tokens_per_second": 882.359
    },
    {
      "epoch": 1.0950214809198888,
      "grad_norm": 0.32653939723968506,
      "learning_rate": 9.083544303797468e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 2019542,
      "step": 1084,
      "train_runtime": 2288.5158,
      "train_tokens_per_second": 882.468
    },
    {
      "epoch": 1.0960323477381855,
      "grad_norm": 0.242202028632164,
      "learning_rate": 9.073417721518987e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 2021506,
      "step": 1085,
      "train_runtime": 2290.3281,
      "train_tokens_per_second": 882.627
    },
    {
      "epoch": 1.0970432145564821,
      "grad_norm": 0.3730691969394684,
      "learning_rate": 9.063291139240506e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 2023396,
      "step": 1086,
      "train_runtime": 2292.152,
      "train_tokens_per_second": 882.749
    },
    {
      "epoch": 1.0980540813747788,
      "grad_norm": 0.39965662360191345,
      "learning_rate": 9.053164556962025e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 2025186,
      "step": 1087,
      "train_runtime": 2293.9077,
      "train_tokens_per_second": 882.854
    },
    {
      "epoch": 1.0990649481930757,
      "grad_norm": 0.34458547830581665,
      "learning_rate": 9.043037974683544e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 2026682,
      "step": 1088,
      "train_runtime": 2295.5735,
      "train_tokens_per_second": 882.865
    },
    {
      "epoch": 1.1000758150113723,
      "grad_norm": 0.34177741408348083,
      "learning_rate": 9.032911392405064e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 2028208,
      "step": 1089,
      "train_runtime": 2297.2198,
      "train_tokens_per_second": 882.897
    },
    {
      "epoch": 1.101086681829669,
      "grad_norm": 0.48805323243141174,
      "learning_rate": 9.022784810126583e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 2029754,
      "step": 1090,
      "train_runtime": 2298.8366,
      "train_tokens_per_second": 882.948
    },
    {
      "epoch": 1.1020975486479656,
      "grad_norm": 0.288225919008255,
      "learning_rate": 9.012658227848102e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 2031872,
      "step": 1091,
      "train_runtime": 2300.7141,
      "train_tokens_per_second": 883.148
    },
    {
      "epoch": 1.1031084154662623,
      "grad_norm": 0.42553961277008057,
      "learning_rate": 9.002531645569621e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 2033980,
      "step": 1092,
      "train_runtime": 2302.5273,
      "train_tokens_per_second": 883.368
    },
    {
      "epoch": 1.104119282284559,
      "grad_norm": 0.5542956590652466,
      "learning_rate": 8.99240506329114e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 2036178,
      "step": 1093,
      "train_runtime": 2304.394,
      "train_tokens_per_second": 883.607
    },
    {
      "epoch": 1.1051301491028558,
      "grad_norm": 0.41452354192733765,
      "learning_rate": 8.982278481012658e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 2037792,
      "step": 1094,
      "train_runtime": 2306.0108,
      "train_tokens_per_second": 883.687
    },
    {
      "epoch": 1.1061410159211524,
      "grad_norm": 0.3179354667663574,
      "learning_rate": 8.972151898734178e-05,
      "loss": 0.1416,
      "num_input_tokens_seen": 2040220,
      "step": 1095,
      "train_runtime": 2308.0192,
      "train_tokens_per_second": 883.97
    },
    {
      "epoch": 1.107151882739449,
      "grad_norm": 0.3286201059818268,
      "learning_rate": 8.962025316455697e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 2041982,
      "step": 1096,
      "train_runtime": 2309.7185,
      "train_tokens_per_second": 884.083
    },
    {
      "epoch": 1.1081627495577457,
      "grad_norm": 0.421648234128952,
      "learning_rate": 8.951898734177215e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 2044038,
      "step": 1097,
      "train_runtime": 2311.455,
      "train_tokens_per_second": 884.308
    },
    {
      "epoch": 1.1091736163760424,
      "grad_norm": 0.35226207971572876,
      "learning_rate": 8.941772151898734e-05,
      "loss": 0.1058,
      "num_input_tokens_seen": 2045950,
      "step": 1098,
      "train_runtime": 2313.294,
      "train_tokens_per_second": 884.431
    },
    {
      "epoch": 1.110184483194339,
      "grad_norm": 0.20200014114379883,
      "learning_rate": 8.931645569620254e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 2047962,
      "step": 1099,
      "train_runtime": 2315.1192,
      "train_tokens_per_second": 884.603
    },
    {
      "epoch": 1.111195350012636,
      "grad_norm": 0.4560726583003998,
      "learning_rate": 8.921518987341772e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 2049682,
      "step": 1100,
      "train_runtime": 2316.8105,
      "train_tokens_per_second": 884.7
    },
    {
      "epoch": 1.1122062168309326,
      "grad_norm": 0.3359977602958679,
      "learning_rate": 8.911392405063291e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 2051370,
      "step": 1101,
      "train_runtime": 2318.4953,
      "train_tokens_per_second": 884.785
    },
    {
      "epoch": 1.1132170836492292,
      "grad_norm": 0.5087964534759521,
      "learning_rate": 8.90126582278481e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 2053164,
      "step": 1102,
      "train_runtime": 2320.2096,
      "train_tokens_per_second": 884.905
    },
    {
      "epoch": 1.1142279504675259,
      "grad_norm": 0.37344732880592346,
      "learning_rate": 8.89113924050633e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 2055164,
      "step": 1103,
      "train_runtime": 2322.0528,
      "train_tokens_per_second": 885.063
    },
    {
      "epoch": 1.1152388172858225,
      "grad_norm": 0.3142015337944031,
      "learning_rate": 8.881012658227848e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 2057456,
      "step": 1104,
      "train_runtime": 2323.9863,
      "train_tokens_per_second": 885.313
    },
    {
      "epoch": 1.1162496841041194,
      "grad_norm": 0.3405570685863495,
      "learning_rate": 8.870886075949368e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 2059120,
      "step": 1105,
      "train_runtime": 2325.5819,
      "train_tokens_per_second": 885.421
    },
    {
      "epoch": 1.117260550922416,
      "grad_norm": 0.31558433175086975,
      "learning_rate": 8.860759493670887e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 2060882,
      "step": 1106,
      "train_runtime": 2327.2823,
      "train_tokens_per_second": 885.532
    },
    {
      "epoch": 1.1182714177407127,
      "grad_norm": 0.42414557933807373,
      "learning_rate": 8.850632911392406e-05,
      "loss": 0.1039,
      "num_input_tokens_seen": 2063240,
      "step": 1107,
      "train_runtime": 2329.3205,
      "train_tokens_per_second": 885.769
    },
    {
      "epoch": 1.1192822845590094,
      "grad_norm": 0.6100692749023438,
      "learning_rate": 8.840506329113925e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 2064608,
      "step": 1108,
      "train_runtime": 2330.9285,
      "train_tokens_per_second": 885.745
    },
    {
      "epoch": 1.120293151377306,
      "grad_norm": 0.44799551367759705,
      "learning_rate": 8.830379746835444e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 2066824,
      "step": 1109,
      "train_runtime": 2332.7952,
      "train_tokens_per_second": 885.986
    },
    {
      "epoch": 1.1213040181956027,
      "grad_norm": 0.34168317914009094,
      "learning_rate": 8.820253164556962e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 2068742,
      "step": 1110,
      "train_runtime": 2334.5112,
      "train_tokens_per_second": 886.156
    },
    {
      "epoch": 1.1223148850138993,
      "grad_norm": 0.39773309230804443,
      "learning_rate": 8.810126582278482e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 2071100,
      "step": 1111,
      "train_runtime": 2338.6237,
      "train_tokens_per_second": 885.606
    },
    {
      "epoch": 1.1233257518321962,
      "grad_norm": 0.26959148049354553,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 2072880,
      "step": 1112,
      "train_runtime": 2340.3205,
      "train_tokens_per_second": 885.725
    },
    {
      "epoch": 1.1243366186504928,
      "grad_norm": 0.3094308376312256,
      "learning_rate": 8.789873417721519e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 2075158,
      "step": 1113,
      "train_runtime": 2342.2121,
      "train_tokens_per_second": 885.982
    },
    {
      "epoch": 1.1253474854687895,
      "grad_norm": 0.4985480010509491,
      "learning_rate": 8.779746835443038e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 2077388,
      "step": 1114,
      "train_runtime": 2344.0949,
      "train_tokens_per_second": 886.222
    },
    {
      "epoch": 1.1263583522870861,
      "grad_norm": 0.3150419592857361,
      "learning_rate": 8.769620253164558e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 2079292,
      "step": 1115,
      "train_runtime": 2345.8065,
      "train_tokens_per_second": 886.387
    },
    {
      "epoch": 1.1273692191053828,
      "grad_norm": 0.4649672508239746,
      "learning_rate": 8.759493670886076e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 2080902,
      "step": 1116,
      "train_runtime": 2347.4431,
      "train_tokens_per_second": 886.455
    },
    {
      "epoch": 1.1283800859236797,
      "grad_norm": 0.4875417649745941,
      "learning_rate": 8.749367088607595e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 2082774,
      "step": 1117,
      "train_runtime": 2349.1682,
      "train_tokens_per_second": 886.601
    },
    {
      "epoch": 1.1293909527419763,
      "grad_norm": 0.2514392137527466,
      "learning_rate": 8.739240506329114e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 2084720,
      "step": 1118,
      "train_runtime": 2351.0173,
      "train_tokens_per_second": 886.731
    },
    {
      "epoch": 1.130401819560273,
      "grad_norm": 0.19609114527702332,
      "learning_rate": 8.729113924050633e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 2086904,
      "step": 1119,
      "train_runtime": 2352.9594,
      "train_tokens_per_second": 886.927
    },
    {
      "epoch": 1.1314126863785696,
      "grad_norm": 0.29445865750312805,
      "learning_rate": 8.718987341772152e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 2088958,
      "step": 1120,
      "train_runtime": 2354.8254,
      "train_tokens_per_second": 887.097
    },
    {
      "epoch": 1.1324235531968663,
      "grad_norm": 0.4594338834285736,
      "learning_rate": 8.708860759493672e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 2090674,
      "step": 1121,
      "train_runtime": 2356.5126,
      "train_tokens_per_second": 887.19
    },
    {
      "epoch": 1.133434420015163,
      "grad_norm": 0.390930712223053,
      "learning_rate": 8.698734177215189e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 2092630,
      "step": 1122,
      "train_runtime": 2358.2898,
      "train_tokens_per_second": 887.351
    },
    {
      "epoch": 1.1344452868334596,
      "grad_norm": 0.3874446451663971,
      "learning_rate": 8.68860759493671e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 2094284,
      "step": 1123,
      "train_runtime": 2359.9757,
      "train_tokens_per_second": 887.418
    },
    {
      "epoch": 1.1354561536517565,
      "grad_norm": 0.360215425491333,
      "learning_rate": 8.678481012658229e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 2096298,
      "step": 1124,
      "train_runtime": 2361.8277,
      "train_tokens_per_second": 887.575
    },
    {
      "epoch": 1.136467020470053,
      "grad_norm": 0.4048649072647095,
      "learning_rate": 8.668354430379747e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 2098198,
      "step": 1125,
      "train_runtime": 2363.6,
      "train_tokens_per_second": 887.713
    },
    {
      "epoch": 1.1374778872883498,
      "grad_norm": 0.3208785951137543,
      "learning_rate": 8.658227848101266e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 2099910,
      "step": 1126,
      "train_runtime": 2365.2767,
      "train_tokens_per_second": 887.807
    },
    {
      "epoch": 1.1384887541066464,
      "grad_norm": 0.43860435485839844,
      "learning_rate": 8.648101265822786e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 2101548,
      "step": 1127,
      "train_runtime": 2366.9564,
      "train_tokens_per_second": 887.869
    },
    {
      "epoch": 1.139499620924943,
      "grad_norm": 0.30121687054634094,
      "learning_rate": 8.637974683544304e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 2104028,
      "step": 1128,
      "train_runtime": 2369.0309,
      "train_tokens_per_second": 888.139
    },
    {
      "epoch": 1.14051048774324,
      "grad_norm": 0.3415144681930542,
      "learning_rate": 8.627848101265823e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 2105922,
      "step": 1129,
      "train_runtime": 2370.8187,
      "train_tokens_per_second": 888.268
    },
    {
      "epoch": 1.1415213545615366,
      "grad_norm": 0.2572648227214813,
      "learning_rate": 8.617721518987342e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 2107714,
      "step": 1130,
      "train_runtime": 2372.5233,
      "train_tokens_per_second": 888.385
    },
    {
      "epoch": 1.1425322213798332,
      "grad_norm": 0.2720210552215576,
      "learning_rate": 8.607594936708861e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 2110432,
      "step": 1131,
      "train_runtime": 2374.6963,
      "train_tokens_per_second": 888.717
    },
    {
      "epoch": 1.1435430881981299,
      "grad_norm": 0.3224233090877533,
      "learning_rate": 8.59746835443038e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 2112168,
      "step": 1132,
      "train_runtime": 2376.4259,
      "train_tokens_per_second": 888.8
    },
    {
      "epoch": 1.1445539550164265,
      "grad_norm": 0.5241628289222717,
      "learning_rate": 8.587341772151899e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 2113676,
      "step": 1133,
      "train_runtime": 2378.0408,
      "train_tokens_per_second": 888.831
    },
    {
      "epoch": 1.1455648218347232,
      "grad_norm": 0.27714425325393677,
      "learning_rate": 8.577215189873418e-05,
      "loss": 0.1047,
      "num_input_tokens_seen": 2115176,
      "step": 1134,
      "train_runtime": 2379.645,
      "train_tokens_per_second": 888.862
    },
    {
      "epoch": 1.1465756886530198,
      "grad_norm": 0.32457640767097473,
      "learning_rate": 8.567088607594937e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 2116832,
      "step": 1135,
      "train_runtime": 2381.276,
      "train_tokens_per_second": 888.949
    },
    {
      "epoch": 1.1475865554713167,
      "grad_norm": 0.4045936167240143,
      "learning_rate": 8.556962025316456e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 2118662,
      "step": 1136,
      "train_runtime": 2383.0636,
      "train_tokens_per_second": 889.05
    },
    {
      "epoch": 1.1485974222896134,
      "grad_norm": 0.3420049846172333,
      "learning_rate": 8.546835443037976e-05,
      "loss": 0.0969,
      "num_input_tokens_seen": 2121236,
      "step": 1137,
      "train_runtime": 2385.1986,
      "train_tokens_per_second": 889.333
    },
    {
      "epoch": 1.14960828910791,
      "grad_norm": 0.3446069657802582,
      "learning_rate": 8.536708860759493e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 2123430,
      "step": 1138,
      "train_runtime": 2387.1744,
      "train_tokens_per_second": 889.516
    },
    {
      "epoch": 1.1506191559262067,
      "grad_norm": 0.32399219274520874,
      "learning_rate": 8.526582278481014e-05,
      "loss": 0.0283,
      "num_input_tokens_seen": 2124798,
      "step": 1139,
      "train_runtime": 2388.7569,
      "train_tokens_per_second": 889.499
    },
    {
      "epoch": 1.1516300227445033,
      "grad_norm": 0.4422810673713684,
      "learning_rate": 8.516455696202533e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 2126354,
      "step": 1140,
      "train_runtime": 2390.4271,
      "train_tokens_per_second": 889.529
    },
    {
      "epoch": 1.1526408895628002,
      "grad_norm": 0.43380090594291687,
      "learning_rate": 8.50632911392405e-05,
      "loss": 0.1153,
      "num_input_tokens_seen": 2128266,
      "step": 1141,
      "train_runtime": 2393.949,
      "train_tokens_per_second": 889.019
    },
    {
      "epoch": 1.1536517563810968,
      "grad_norm": 0.6191344857215881,
      "learning_rate": 8.49620253164557e-05,
      "loss": 0.1236,
      "num_input_tokens_seen": 2129752,
      "step": 1142,
      "train_runtime": 2395.5717,
      "train_tokens_per_second": 889.037
    },
    {
      "epoch": 1.1546626231993935,
      "grad_norm": 0.40897995233535767,
      "learning_rate": 8.48607594936709e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 2131820,
      "step": 1143,
      "train_runtime": 2397.4345,
      "train_tokens_per_second": 889.209
    },
    {
      "epoch": 1.1556734900176902,
      "grad_norm": 0.28095927834510803,
      "learning_rate": 8.475949367088608e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 2133928,
      "step": 1144,
      "train_runtime": 2399.2461,
      "train_tokens_per_second": 889.416
    },
    {
      "epoch": 1.1566843568359868,
      "grad_norm": 0.2282506823539734,
      "learning_rate": 8.465822784810127e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 2136234,
      "step": 1145,
      "train_runtime": 2401.1357,
      "train_tokens_per_second": 889.677
    },
    {
      "epoch": 1.1576952236542835,
      "grad_norm": 0.30796393752098083,
      "learning_rate": 8.455696202531646e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 2138160,
      "step": 1146,
      "train_runtime": 2402.8561,
      "train_tokens_per_second": 889.841
    },
    {
      "epoch": 1.1587060904725803,
      "grad_norm": 0.3619937300682068,
      "learning_rate": 8.445569620253165e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 2139622,
      "step": 1147,
      "train_runtime": 2404.4774,
      "train_tokens_per_second": 889.849
    },
    {
      "epoch": 1.159716957290877,
      "grad_norm": 0.41102486848831177,
      "learning_rate": 8.435443037974684e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 2141228,
      "step": 1148,
      "train_runtime": 2406.1442,
      "train_tokens_per_second": 889.9
    },
    {
      "epoch": 1.1607278241091736,
      "grad_norm": 0.5577710270881653,
      "learning_rate": 8.425316455696203e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 2143202,
      "step": 1149,
      "train_runtime": 2407.9249,
      "train_tokens_per_second": 890.062
    },
    {
      "epoch": 1.1617386909274703,
      "grad_norm": 0.3255091905593872,
      "learning_rate": 8.415189873417721e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 2145368,
      "step": 1150,
      "train_runtime": 2409.8477,
      "train_tokens_per_second": 890.25
    },
    {
      "epoch": 1.162749557745767,
      "grad_norm": 0.3634073734283447,
      "learning_rate": 8.405063291139241e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 2147748,
      "step": 1151,
      "train_runtime": 2411.8587,
      "train_tokens_per_second": 890.495
    },
    {
      "epoch": 1.1637604245640636,
      "grad_norm": 0.3885149359703064,
      "learning_rate": 8.39493670886076e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 2149888,
      "step": 1152,
      "train_runtime": 2413.7537,
      "train_tokens_per_second": 890.682
    },
    {
      "epoch": 1.1647712913823605,
      "grad_norm": 0.3698866665363312,
      "learning_rate": 8.384810126582278e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 2151634,
      "step": 1153,
      "train_runtime": 2415.4029,
      "train_tokens_per_second": 890.797
    },
    {
      "epoch": 1.1657821582006571,
      "grad_norm": 0.2277451604604721,
      "learning_rate": 8.374683544303797e-05,
      "loss": 0.0245,
      "num_input_tokens_seen": 2153826,
      "step": 1154,
      "train_runtime": 2417.339,
      "train_tokens_per_second": 890.99
    },
    {
      "epoch": 1.1667930250189538,
      "grad_norm": 0.26030606031417847,
      "learning_rate": 8.364556962025318e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 2155448,
      "step": 1155,
      "train_runtime": 2419.0155,
      "train_tokens_per_second": 891.043
    },
    {
      "epoch": 1.1678038918372504,
      "grad_norm": 0.33531326055526733,
      "learning_rate": 8.354430379746835e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 2157312,
      "step": 1156,
      "train_runtime": 2420.7965,
      "train_tokens_per_second": 891.158
    },
    {
      "epoch": 1.168814758655547,
      "grad_norm": 0.35256895422935486,
      "learning_rate": 8.344303797468354e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 2158738,
      "step": 1157,
      "train_runtime": 2422.3912,
      "train_tokens_per_second": 891.16
    },
    {
      "epoch": 1.1698256254738437,
      "grad_norm": 0.32339709997177124,
      "learning_rate": 8.334177215189874e-05,
      "loss": 0.077,
      "num_input_tokens_seen": 2160564,
      "step": 1158,
      "train_runtime": 2424.1473,
      "train_tokens_per_second": 891.268
    },
    {
      "epoch": 1.1708364922921406,
      "grad_norm": 0.5041233897209167,
      "learning_rate": 8.324050632911393e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 2162618,
      "step": 1159,
      "train_runtime": 2426.01,
      "train_tokens_per_second": 891.43
    },
    {
      "epoch": 1.1718473591104372,
      "grad_norm": 0.6854803562164307,
      "learning_rate": 8.313924050632912e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 2164408,
      "step": 1160,
      "train_runtime": 2427.7162,
      "train_tokens_per_second": 891.541
    },
    {
      "epoch": 1.172858225928734,
      "grad_norm": 0.48390182852745056,
      "learning_rate": 8.303797468354431e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 2166420,
      "step": 1161,
      "train_runtime": 2429.5733,
      "train_tokens_per_second": 891.687
    },
    {
      "epoch": 1.1738690927470306,
      "grad_norm": 0.27975279092788696,
      "learning_rate": 8.29367088607595e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 2168606,
      "step": 1162,
      "train_runtime": 2431.4379,
      "train_tokens_per_second": 891.903
    },
    {
      "epoch": 1.1748799595653272,
      "grad_norm": 0.3532312214374542,
      "learning_rate": 8.283544303797469e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 2170476,
      "step": 1163,
      "train_runtime": 2433.221,
      "train_tokens_per_second": 892.018
    },
    {
      "epoch": 1.175890826383624,
      "grad_norm": 0.2860186696052551,
      "learning_rate": 8.273417721518988e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 2172142,
      "step": 1164,
      "train_runtime": 2434.8494,
      "train_tokens_per_second": 892.105
    },
    {
      "epoch": 1.1769016932019207,
      "grad_norm": 0.23471611738204956,
      "learning_rate": 8.263291139240507e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 2173994,
      "step": 1165,
      "train_runtime": 2436.6149,
      "train_tokens_per_second": 892.219
    },
    {
      "epoch": 1.1779125600202174,
      "grad_norm": 0.2529040575027466,
      "learning_rate": 8.253164556962025e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 2176036,
      "step": 1166,
      "train_runtime": 2438.4839,
      "train_tokens_per_second": 892.373
    },
    {
      "epoch": 1.178923426838514,
      "grad_norm": 0.5520626902580261,
      "learning_rate": 8.243037974683545e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 2178188,
      "step": 1167,
      "train_runtime": 2440.3515,
      "train_tokens_per_second": 892.571
    },
    {
      "epoch": 1.1799342936568107,
      "grad_norm": 0.26322516798973083,
      "learning_rate": 8.232911392405064e-05,
      "loss": 0.1911,
      "num_input_tokens_seen": 2180116,
      "step": 1168,
      "train_runtime": 2442.1151,
      "train_tokens_per_second": 892.716
    },
    {
      "epoch": 1.1809451604751073,
      "grad_norm": 0.30632102489471436,
      "learning_rate": 8.222784810126582e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 2182022,
      "step": 1169,
      "train_runtime": 2443.9252,
      "train_tokens_per_second": 892.835
    },
    {
      "epoch": 1.181956027293404,
      "grad_norm": 0.37983787059783936,
      "learning_rate": 8.212658227848101e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 2183786,
      "step": 1170,
      "train_runtime": 2445.614,
      "train_tokens_per_second": 892.94
    },
    {
      "epoch": 1.1829668941117009,
      "grad_norm": 0.27352458238601685,
      "learning_rate": 8.202531645569622e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 2185608,
      "step": 1171,
      "train_runtime": 2449.5647,
      "train_tokens_per_second": 892.243
    },
    {
      "epoch": 1.1839777609299975,
      "grad_norm": 0.3310551643371582,
      "learning_rate": 8.19240506329114e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 2187262,
      "step": 1172,
      "train_runtime": 2451.1776,
      "train_tokens_per_second": 892.331
    },
    {
      "epoch": 1.1849886277482942,
      "grad_norm": 0.2715113162994385,
      "learning_rate": 8.182278481012658e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 2189264,
      "step": 1173,
      "train_runtime": 2452.9596,
      "train_tokens_per_second": 892.499
    },
    {
      "epoch": 1.1859994945665908,
      "grad_norm": 0.24654069542884827,
      "learning_rate": 8.172151898734178e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 2191248,
      "step": 1174,
      "train_runtime": 2454.751,
      "train_tokens_per_second": 892.656
    },
    {
      "epoch": 1.1870103613848875,
      "grad_norm": 0.25723373889923096,
      "learning_rate": 8.162025316455697e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 2193162,
      "step": 1175,
      "train_runtime": 2456.5376,
      "train_tokens_per_second": 892.786
    },
    {
      "epoch": 1.1880212282031843,
      "grad_norm": 0.25851204991340637,
      "learning_rate": 8.151898734177216e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 2195312,
      "step": 1176,
      "train_runtime": 2458.4245,
      "train_tokens_per_second": 892.975
    },
    {
      "epoch": 1.189032095021481,
      "grad_norm": 0.3143210709095001,
      "learning_rate": 8.141772151898735e-05,
      "loss": 0.1391,
      "num_input_tokens_seen": 2196974,
      "step": 1177,
      "train_runtime": 2460.0596,
      "train_tokens_per_second": 893.057
    },
    {
      "epoch": 1.1900429618397776,
      "grad_norm": 0.32049065828323364,
      "learning_rate": 8.131645569620253e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 2198798,
      "step": 1178,
      "train_runtime": 2461.7765,
      "train_tokens_per_second": 893.175
    },
    {
      "epoch": 1.1910538286580743,
      "grad_norm": 0.4614190459251404,
      "learning_rate": 8.121518987341773e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 2200684,
      "step": 1179,
      "train_runtime": 2463.535,
      "train_tokens_per_second": 893.303
    },
    {
      "epoch": 1.192064695476371,
      "grad_norm": 0.3254811763763428,
      "learning_rate": 8.111392405063292e-05,
      "loss": 0.2184,
      "num_input_tokens_seen": 2202666,
      "step": 1180,
      "train_runtime": 2465.3133,
      "train_tokens_per_second": 893.463
    },
    {
      "epoch": 1.1930755622946676,
      "grad_norm": 0.3342306315898895,
      "learning_rate": 8.10126582278481e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 2204414,
      "step": 1181,
      "train_runtime": 2467.1121,
      "train_tokens_per_second": 893.52
    },
    {
      "epoch": 1.1940864291129643,
      "grad_norm": 0.43980565667152405,
      "learning_rate": 8.091139240506329e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 2206550,
      "step": 1182,
      "train_runtime": 2468.9852,
      "train_tokens_per_second": 893.707
    },
    {
      "epoch": 1.1950972959312611,
      "grad_norm": 0.35847440361976624,
      "learning_rate": 8.08101265822785e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 2208522,
      "step": 1183,
      "train_runtime": 2470.7873,
      "train_tokens_per_second": 893.854
    },
    {
      "epoch": 1.1961081627495578,
      "grad_norm": 0.3575953543186188,
      "learning_rate": 8.070886075949367e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 2209880,
      "step": 1184,
      "train_runtime": 2472.393,
      "train_tokens_per_second": 893.822
    },
    {
      "epoch": 1.1971190295678544,
      "grad_norm": 0.25793489813804626,
      "learning_rate": 8.060759493670886e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 2211972,
      "step": 1185,
      "train_runtime": 2474.2572,
      "train_tokens_per_second": 893.994
    },
    {
      "epoch": 1.198129896386151,
      "grad_norm": 0.4657045900821686,
      "learning_rate": 8.050632911392405e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 2213722,
      "step": 1186,
      "train_runtime": 2475.9573,
      "train_tokens_per_second": 894.087
    },
    {
      "epoch": 1.1991407632044477,
      "grad_norm": 0.466621071100235,
      "learning_rate": 8.040506329113924e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 2215788,
      "step": 1187,
      "train_runtime": 2477.8066,
      "train_tokens_per_second": 894.254
    },
    {
      "epoch": 1.2001516300227446,
      "grad_norm": 0.4542536735534668,
      "learning_rate": 8.030379746835443e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 2217242,
      "step": 1188,
      "train_runtime": 2479.4571,
      "train_tokens_per_second": 894.245
    },
    {
      "epoch": 1.2001516300227446,
      "eval_loss": 0.20719842612743378,
      "eval_runtime": 61.1791,
      "eval_samples_per_second": 14.368,
      "eval_steps_per_second": 7.192,
      "num_input_tokens_seen": 2217242,
      "step": 1188
    },
    {
      "epoch": 1.2011624968410413,
      "grad_norm": 0.34237897396087646,
      "learning_rate": 8.020253164556962e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 2218976,
      "step": 1189,
      "train_runtime": 2542.3915,
      "train_tokens_per_second": 872.791
    },
    {
      "epoch": 1.202173363659338,
      "grad_norm": 0.31419843435287476,
      "learning_rate": 8.010126582278482e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 2221286,
      "step": 1190,
      "train_runtime": 2544.3813,
      "train_tokens_per_second": 873.016
    },
    {
      "epoch": 1.2031842304776346,
      "grad_norm": 0.2560994625091553,
      "learning_rate": 8e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 2223160,
      "step": 1191,
      "train_runtime": 2546.0924,
      "train_tokens_per_second": 873.165
    },
    {
      "epoch": 1.2041950972959312,
      "grad_norm": 0.3249039351940155,
      "learning_rate": 7.98987341772152e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 2224882,
      "step": 1192,
      "train_runtime": 2547.8416,
      "train_tokens_per_second": 873.242
    },
    {
      "epoch": 1.2052059641142279,
      "grad_norm": 0.4017309248447418,
      "learning_rate": 7.979746835443039e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 2226830,
      "step": 1193,
      "train_runtime": 2549.6224,
      "train_tokens_per_second": 873.396
    },
    {
      "epoch": 1.2062168309325245,
      "grad_norm": 0.3083512783050537,
      "learning_rate": 7.969620253164557e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 2229160,
      "step": 1194,
      "train_runtime": 2551.5595,
      "train_tokens_per_second": 873.646
    },
    {
      "epoch": 1.2072276977508214,
      "grad_norm": 0.3623747229576111,
      "learning_rate": 7.959493670886077e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 2230942,
      "step": 1195,
      "train_runtime": 2553.3159,
      "train_tokens_per_second": 873.743
    },
    {
      "epoch": 1.208238564569118,
      "grad_norm": 0.3974745571613312,
      "learning_rate": 7.949367088607596e-05,
      "loss": 0.0992,
      "num_input_tokens_seen": 2232456,
      "step": 1196,
      "train_runtime": 2554.8992,
      "train_tokens_per_second": 873.794
    },
    {
      "epoch": 1.2092494313874147,
      "grad_norm": 0.3379092216491699,
      "learning_rate": 7.939240506329114e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 2234064,
      "step": 1197,
      "train_runtime": 2556.5189,
      "train_tokens_per_second": 873.87
    },
    {
      "epoch": 1.2102602982057113,
      "grad_norm": 0.44574904441833496,
      "learning_rate": 7.929113924050633e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 2235656,
      "step": 1198,
      "train_runtime": 2558.1361,
      "train_tokens_per_second": 873.939
    },
    {
      "epoch": 1.211271165024008,
      "grad_norm": 0.4556899070739746,
      "learning_rate": 7.918987341772153e-05,
      "loss": 0.1008,
      "num_input_tokens_seen": 2237148,
      "step": 1199,
      "train_runtime": 2559.8067,
      "train_tokens_per_second": 873.952
    },
    {
      "epoch": 1.2122820318423049,
      "grad_norm": 0.3602530360221863,
      "learning_rate": 7.908860759493671e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 2238558,
      "step": 1200,
      "train_runtime": 2561.4404,
      "train_tokens_per_second": 873.945
    },
    {
      "epoch": 1.2132928986606015,
      "grad_norm": 0.3926827311515808,
      "learning_rate": 7.89873417721519e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 2240102,
      "step": 1201,
      "train_runtime": 2565.3309,
      "train_tokens_per_second": 873.221
    },
    {
      "epoch": 1.2143037654788982,
      "grad_norm": 0.36526402831077576,
      "learning_rate": 7.888607594936709e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 2241606,
      "step": 1202,
      "train_runtime": 2566.9898,
      "train_tokens_per_second": 873.243
    },
    {
      "epoch": 1.2153146322971948,
      "grad_norm": 0.26066580414772034,
      "learning_rate": 7.878481012658228e-05,
      "loss": 0.0252,
      "num_input_tokens_seen": 2243026,
      "step": 1203,
      "train_runtime": 2568.5626,
      "train_tokens_per_second": 873.261
    },
    {
      "epoch": 1.2163254991154915,
      "grad_norm": 0.32624346017837524,
      "learning_rate": 7.868354430379747e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 2245108,
      "step": 1204,
      "train_runtime": 2570.4056,
      "train_tokens_per_second": 873.445
    },
    {
      "epoch": 1.2173363659337881,
      "grad_norm": 0.35179993510246277,
      "learning_rate": 7.858227848101266e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 2246944,
      "step": 1205,
      "train_runtime": 2572.1135,
      "train_tokens_per_second": 873.579
    },
    {
      "epoch": 1.218347232752085,
      "grad_norm": 0.4054023027420044,
      "learning_rate": 7.848101265822784e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 2248626,
      "step": 1206,
      "train_runtime": 2573.8102,
      "train_tokens_per_second": 873.656
    },
    {
      "epoch": 1.2193580995703817,
      "grad_norm": 0.28881245851516724,
      "learning_rate": 7.837974683544305e-05,
      "loss": 0.0969,
      "num_input_tokens_seen": 2251106,
      "step": 1207,
      "train_runtime": 2575.8394,
      "train_tokens_per_second": 873.931
    },
    {
      "epoch": 1.2203689663886783,
      "grad_norm": 0.3547137677669525,
      "learning_rate": 7.827848101265824e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 2252336,
      "step": 1208,
      "train_runtime": 2577.3751,
      "train_tokens_per_second": 873.888
    },
    {
      "epoch": 1.221379833206975,
      "grad_norm": 0.4218279719352722,
      "learning_rate": 7.817721518987341e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 2254426,
      "step": 1209,
      "train_runtime": 2579.1698,
      "train_tokens_per_second": 874.09
    },
    {
      "epoch": 1.2223907000252716,
      "grad_norm": 0.36659857630729675,
      "learning_rate": 7.80759493670886e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 2256016,
      "step": 1210,
      "train_runtime": 2580.7789,
      "train_tokens_per_second": 874.161
    },
    {
      "epoch": 1.2234015668435683,
      "grad_norm": 0.34027570486068726,
      "learning_rate": 7.797468354430381e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 2257654,
      "step": 1211,
      "train_runtime": 2582.3909,
      "train_tokens_per_second": 874.25
    },
    {
      "epoch": 1.2244124336618651,
      "grad_norm": 0.3136456608772278,
      "learning_rate": 7.787341772151899e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 2259316,
      "step": 1212,
      "train_runtime": 2584.0769,
      "train_tokens_per_second": 874.322
    },
    {
      "epoch": 1.2254233004801618,
      "grad_norm": 0.23020558059215546,
      "learning_rate": 7.777215189873418e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 2261760,
      "step": 1213,
      "train_runtime": 2586.1467,
      "train_tokens_per_second": 874.568
    },
    {
      "epoch": 1.2264341672984584,
      "grad_norm": 0.5002903342247009,
      "learning_rate": 7.767088607594937e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 2263620,
      "step": 1214,
      "train_runtime": 2587.9609,
      "train_tokens_per_second": 874.673
    },
    {
      "epoch": 1.227445034116755,
      "grad_norm": 0.26744261384010315,
      "learning_rate": 7.756962025316456e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 2266624,
      "step": 1215,
      "train_runtime": 2590.3504,
      "train_tokens_per_second": 875.026
    },
    {
      "epoch": 1.2284559009350517,
      "grad_norm": 0.2974698841571808,
      "learning_rate": 7.746835443037975e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 2268118,
      "step": 1216,
      "train_runtime": 2591.9528,
      "train_tokens_per_second": 875.061
    },
    {
      "epoch": 1.2294667677533484,
      "grad_norm": 0.3523530066013336,
      "learning_rate": 7.736708860759494e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 2270112,
      "step": 1217,
      "train_runtime": 2593.7499,
      "train_tokens_per_second": 875.224
    },
    {
      "epoch": 1.2304776345716453,
      "grad_norm": 0.306835412979126,
      "learning_rate": 7.726582278481013e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 2272270,
      "step": 1218,
      "train_runtime": 2595.5679,
      "train_tokens_per_second": 875.442
    },
    {
      "epoch": 1.231488501389942,
      "grad_norm": 0.25264307856559753,
      "learning_rate": 7.716455696202532e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 2274252,
      "step": 1219,
      "train_runtime": 2597.3688,
      "train_tokens_per_second": 875.598
    },
    {
      "epoch": 1.2324993682082386,
      "grad_norm": 0.39503711462020874,
      "learning_rate": 7.706329113924051e-05,
      "loss": 0.1344,
      "num_input_tokens_seen": 2276608,
      "step": 1220,
      "train_runtime": 2599.3439,
      "train_tokens_per_second": 875.839
    },
    {
      "epoch": 1.2335102350265352,
      "grad_norm": 0.35884588956832886,
      "learning_rate": 7.69620253164557e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 2278464,
      "step": 1221,
      "train_runtime": 2601.0513,
      "train_tokens_per_second": 875.978
    },
    {
      "epoch": 1.2345211018448319,
      "grad_norm": 0.2843872904777527,
      "learning_rate": 7.686075949367088e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 2280334,
      "step": 1222,
      "train_runtime": 2602.7639,
      "train_tokens_per_second": 876.12
    },
    {
      "epoch": 1.2355319686631288,
      "grad_norm": 0.40163278579711914,
      "learning_rate": 7.675949367088609e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 2282350,
      "step": 1223,
      "train_runtime": 2604.5568,
      "train_tokens_per_second": 876.291
    },
    {
      "epoch": 1.2365428354814254,
      "grad_norm": 0.3127462863922119,
      "learning_rate": 7.665822784810128e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 2284232,
      "step": 1224,
      "train_runtime": 2606.2777,
      "train_tokens_per_second": 876.435
    },
    {
      "epoch": 1.237553702299722,
      "grad_norm": 0.3897799551486969,
      "learning_rate": 7.655696202531645e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 2286378,
      "step": 1225,
      "train_runtime": 2608.1654,
      "train_tokens_per_second": 876.623
    },
    {
      "epoch": 1.2385645691180187,
      "grad_norm": 0.4516952931880951,
      "learning_rate": 7.645569620253165e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 2288376,
      "step": 1226,
      "train_runtime": 2609.9431,
      "train_tokens_per_second": 876.792
    },
    {
      "epoch": 1.2395754359363154,
      "grad_norm": 0.41795146465301514,
      "learning_rate": 7.635443037974685e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 2290008,
      "step": 1227,
      "train_runtime": 2611.6834,
      "train_tokens_per_second": 876.832
    },
    {
      "epoch": 1.240586302754612,
      "grad_norm": 0.26265859603881836,
      "learning_rate": 7.625316455696203e-05,
      "loss": 0.1215,
      "num_input_tokens_seen": 2291396,
      "step": 1228,
      "train_runtime": 2613.2831,
      "train_tokens_per_second": 876.827
    },
    {
      "epoch": 1.2415971695729087,
      "grad_norm": 0.2844543159008026,
      "learning_rate": 7.615189873417722e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 2293300,
      "step": 1229,
      "train_runtime": 2615.0049,
      "train_tokens_per_second": 876.977
    },
    {
      "epoch": 1.2426080363912055,
      "grad_norm": 0.2673298120498657,
      "learning_rate": 7.605063291139241e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 2295184,
      "step": 1230,
      "train_runtime": 2616.7145,
      "train_tokens_per_second": 877.124
    },
    {
      "epoch": 1.2436189032095022,
      "grad_norm": 0.4001712203025818,
      "learning_rate": 7.59493670886076e-05,
      "loss": 0.0542,
      "num_input_tokens_seen": 2296654,
      "step": 1231,
      "train_runtime": 2620.9099,
      "train_tokens_per_second": 876.281
    },
    {
      "epoch": 1.2446297700277988,
      "grad_norm": 0.44643983244895935,
      "learning_rate": 7.584810126582279e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 2299028,
      "step": 1232,
      "train_runtime": 2622.8712,
      "train_tokens_per_second": 876.531
    },
    {
      "epoch": 1.2456406368460955,
      "grad_norm": 0.3430020809173584,
      "learning_rate": 7.574683544303798e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 2300760,
      "step": 1233,
      "train_runtime": 2624.5658,
      "train_tokens_per_second": 876.625
    },
    {
      "epoch": 1.2466515036643921,
      "grad_norm": 0.45810046792030334,
      "learning_rate": 7.564556962025316e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 2302192,
      "step": 1234,
      "train_runtime": 2626.2183,
      "train_tokens_per_second": 876.619
    },
    {
      "epoch": 1.247662370482689,
      "grad_norm": 0.40292829275131226,
      "learning_rate": 7.554430379746836e-05,
      "loss": 0.1584,
      "num_input_tokens_seen": 2303964,
      "step": 1235,
      "train_runtime": 2627.9219,
      "train_tokens_per_second": 876.725
    },
    {
      "epoch": 1.2486732373009857,
      "grad_norm": 0.2423546016216278,
      "learning_rate": 7.544303797468355e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 2306384,
      "step": 1236,
      "train_runtime": 2629.9422,
      "train_tokens_per_second": 876.971
    },
    {
      "epoch": 1.2496841041192823,
      "grad_norm": 0.231626957654953,
      "learning_rate": 7.534177215189873e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 2308302,
      "step": 1237,
      "train_runtime": 2631.7215,
      "train_tokens_per_second": 877.107
    },
    {
      "epoch": 1.250694970937579,
      "grad_norm": 0.2823111116886139,
      "learning_rate": 7.524050632911392e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 2310652,
      "step": 1238,
      "train_runtime": 2633.6657,
      "train_tokens_per_second": 877.352
    },
    {
      "epoch": 1.2517058377558756,
      "grad_norm": 0.28391867876052856,
      "learning_rate": 7.513924050632913e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 2312742,
      "step": 1239,
      "train_runtime": 2635.5242,
      "train_tokens_per_second": 877.526
    },
    {
      "epoch": 1.2527167045741723,
      "grad_norm": 0.290244996547699,
      "learning_rate": 7.50379746835443e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 2314768,
      "step": 1240,
      "train_runtime": 2637.3763,
      "train_tokens_per_second": 877.678
    },
    {
      "epoch": 1.253727571392469,
      "grad_norm": 0.2558806240558624,
      "learning_rate": 7.49367088607595e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 2316864,
      "step": 1241,
      "train_runtime": 2639.2359,
      "train_tokens_per_second": 877.854
    },
    {
      "epoch": 1.2547384382107658,
      "grad_norm": 0.3218919634819031,
      "learning_rate": 7.483544303797469e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 2319096,
      "step": 1242,
      "train_runtime": 2641.1826,
      "train_tokens_per_second": 878.052
    },
    {
      "epoch": 1.2557493050290625,
      "grad_norm": 0.24845680594444275,
      "learning_rate": 7.473417721518988e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 2321292,
      "step": 1243,
      "train_runtime": 2643.0508,
      "train_tokens_per_second": 878.262
    },
    {
      "epoch": 1.256760171847359,
      "grad_norm": 0.2059890478849411,
      "learning_rate": 7.463291139240507e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 2323460,
      "step": 1244,
      "train_runtime": 2644.9776,
      "train_tokens_per_second": 878.442
    },
    {
      "epoch": 1.2577710386656558,
      "grad_norm": 0.6027005314826965,
      "learning_rate": 7.453164556962026e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 2325092,
      "step": 1245,
      "train_runtime": 2646.5949,
      "train_tokens_per_second": 878.522
    },
    {
      "epoch": 1.2587819054839524,
      "grad_norm": 0.40883567929267883,
      "learning_rate": 7.443037974683545e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 2327080,
      "step": 1246,
      "train_runtime": 2648.3789,
      "train_tokens_per_second": 878.681
    },
    {
      "epoch": 1.2597927723022493,
      "grad_norm": 0.5405418276786804,
      "learning_rate": 7.432911392405064e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 2329076,
      "step": 1247,
      "train_runtime": 2650.1605,
      "train_tokens_per_second": 878.843
    },
    {
      "epoch": 1.260803639120546,
      "grad_norm": 0.49589258432388306,
      "learning_rate": 7.422784810126583e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 2330756,
      "step": 1248,
      "train_runtime": 2651.8634,
      "train_tokens_per_second": 878.913
    },
    {
      "epoch": 1.2618145059388426,
      "grad_norm": 0.5561320781707764,
      "learning_rate": 7.412658227848102e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 2332434,
      "step": 1249,
      "train_runtime": 2653.5588,
      "train_tokens_per_second": 878.983
    },
    {
      "epoch": 1.2628253727571392,
      "grad_norm": 0.2283097803592682,
      "learning_rate": 7.40253164556962e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 2334348,
      "step": 1250,
      "train_runtime": 2655.3257,
      "train_tokens_per_second": 879.119
    },
    {
      "epoch": 1.263836239575436,
      "grad_norm": 0.5350798964500427,
      "learning_rate": 7.39240506329114e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 2335872,
      "step": 1251,
      "train_runtime": 2656.9785,
      "train_tokens_per_second": 879.146
    },
    {
      "epoch": 1.2648471063937325,
      "grad_norm": 0.4096823036670685,
      "learning_rate": 7.38227848101266e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 2337458,
      "step": 1252,
      "train_runtime": 2658.604,
      "train_tokens_per_second": 879.205
    },
    {
      "epoch": 1.2658579732120292,
      "grad_norm": 0.3206482231616974,
      "learning_rate": 7.372151898734177e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 2339236,
      "step": 1253,
      "train_runtime": 2660.2984,
      "train_tokens_per_second": 879.313
    },
    {
      "epoch": 1.266868840030326,
      "grad_norm": 0.2557467222213745,
      "learning_rate": 7.362025316455696e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 2340906,
      "step": 1254,
      "train_runtime": 2661.9943,
      "train_tokens_per_second": 879.381
    },
    {
      "epoch": 1.2678797068486227,
      "grad_norm": 0.35144689679145813,
      "learning_rate": 7.351898734177217e-05,
      "loss": 0.1125,
      "num_input_tokens_seen": 2342640,
      "step": 1255,
      "train_runtime": 2663.6248,
      "train_tokens_per_second": 879.493
    },
    {
      "epoch": 1.2688905736669194,
      "grad_norm": 0.28803375363349915,
      "learning_rate": 7.341772151898734e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 2344838,
      "step": 1256,
      "train_runtime": 2665.5011,
      "train_tokens_per_second": 879.699
    },
    {
      "epoch": 1.269901440485216,
      "grad_norm": 0.4108582139015198,
      "learning_rate": 7.331645569620253e-05,
      "loss": 0.086,
      "num_input_tokens_seen": 2346846,
      "step": 1257,
      "train_runtime": 2667.3419,
      "train_tokens_per_second": 879.844
    },
    {
      "epoch": 1.270912307303513,
      "grad_norm": 0.23005342483520508,
      "learning_rate": 7.321518987341773e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 2348908,
      "step": 1258,
      "train_runtime": 2669.2325,
      "train_tokens_per_second": 879.994
    },
    {
      "epoch": 1.2719231741218096,
      "grad_norm": 0.26932668685913086,
      "learning_rate": 7.311392405063292e-05,
      "loss": 0.0154,
      "num_input_tokens_seen": 2350608,
      "step": 1259,
      "train_runtime": 2670.9113,
      "train_tokens_per_second": 880.077
    },
    {
      "epoch": 1.2729340409401062,
      "grad_norm": 0.23961332440376282,
      "learning_rate": 7.301265822784811e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 2352324,
      "step": 1260,
      "train_runtime": 2672.6533,
      "train_tokens_per_second": 880.146
    },
    {
      "epoch": 1.2739449077584029,
      "grad_norm": 0.28488975763320923,
      "learning_rate": 7.29113924050633e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 2353934,
      "step": 1261,
      "train_runtime": 2676.1921,
      "train_tokens_per_second": 879.583
    },
    {
      "epoch": 1.2749557745766995,
      "grad_norm": 0.3717709183692932,
      "learning_rate": 7.281012658227848e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 2356002,
      "step": 1262,
      "train_runtime": 2678.0484,
      "train_tokens_per_second": 879.746
    },
    {
      "epoch": 1.2759666413949962,
      "grad_norm": 0.47229495644569397,
      "learning_rate": 7.270886075949367e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 2357928,
      "step": 1263,
      "train_runtime": 2679.8879,
      "train_tokens_per_second": 879.861
    },
    {
      "epoch": 1.2769775082132928,
      "grad_norm": 0.32415246963500977,
      "learning_rate": 7.260759493670887e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 2360130,
      "step": 1264,
      "train_runtime": 2681.829,
      "train_tokens_per_second": 880.045
    },
    {
      "epoch": 1.2779883750315895,
      "grad_norm": 0.30456259846687317,
      "learning_rate": 7.250632911392405e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 2361726,
      "step": 1265,
      "train_runtime": 2683.5246,
      "train_tokens_per_second": 880.084
    },
    {
      "epoch": 1.2789992418498863,
      "grad_norm": 0.4040810465812683,
      "learning_rate": 7.240506329113924e-05,
      "loss": 0.1235,
      "num_input_tokens_seen": 2363846,
      "step": 1266,
      "train_runtime": 2685.4004,
      "train_tokens_per_second": 880.258
    },
    {
      "epoch": 1.280010108668183,
      "grad_norm": 0.33810141682624817,
      "learning_rate": 7.230379746835443e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 2366592,
      "step": 1267,
      "train_runtime": 2687.5555,
      "train_tokens_per_second": 880.574
    },
    {
      "epoch": 1.2810209754864796,
      "grad_norm": 0.3703427314758301,
      "learning_rate": 7.220253164556962e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 2368360,
      "step": 1268,
      "train_runtime": 2689.2081,
      "train_tokens_per_second": 880.69
    },
    {
      "epoch": 1.2820318423047763,
      "grad_norm": 0.2576245069503784,
      "learning_rate": 7.210126582278481e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 2370460,
      "step": 1269,
      "train_runtime": 2691.1227,
      "train_tokens_per_second": 880.844
    },
    {
      "epoch": 1.2830427091230732,
      "grad_norm": 0.2916818857192993,
      "learning_rate": 7.2e-05,
      "loss": 0.0788,
      "num_input_tokens_seen": 2372182,
      "step": 1270,
      "train_runtime": 2692.8773,
      "train_tokens_per_second": 880.91
    },
    {
      "epoch": 1.2840535759413698,
      "grad_norm": 0.39087337255477905,
      "learning_rate": 7.189873417721519e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 2373984,
      "step": 1271,
      "train_runtime": 2694.5948,
      "train_tokens_per_second": 881.017
    },
    {
      "epoch": 1.2850644427596665,
      "grad_norm": 0.4398673176765442,
      "learning_rate": 7.179746835443038e-05,
      "loss": 0.1272,
      "num_input_tokens_seen": 2375920,
      "step": 1272,
      "train_runtime": 2696.3816,
      "train_tokens_per_second": 881.151
    },
    {
      "epoch": 1.2860753095779631,
      "grad_norm": 0.3793856203556061,
      "learning_rate": 7.169620253164557e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 2378138,
      "step": 1273,
      "train_runtime": 2698.2551,
      "train_tokens_per_second": 881.361
    },
    {
      "epoch": 1.2870861763962598,
      "grad_norm": 0.2971085011959076,
      "learning_rate": 7.159493670886077e-05,
      "loss": 0.0252,
      "num_input_tokens_seen": 2379852,
      "step": 1274,
      "train_runtime": 2699.8921,
      "train_tokens_per_second": 881.462
    },
    {
      "epoch": 1.2880970432145564,
      "grad_norm": 0.48829659819602966,
      "learning_rate": 7.149367088607594e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 2381660,
      "step": 1275,
      "train_runtime": 2701.6418,
      "train_tokens_per_second": 881.56
    },
    {
      "epoch": 1.289107910032853,
      "grad_norm": 0.37237298488616943,
      "learning_rate": 7.139240506329115e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 2383732,
      "step": 1276,
      "train_runtime": 2703.5435,
      "train_tokens_per_second": 881.707
    },
    {
      "epoch": 1.2901187768511497,
      "grad_norm": 0.33351263403892517,
      "learning_rate": 7.129113924050634e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 2385694,
      "step": 1277,
      "train_runtime": 2705.3429,
      "train_tokens_per_second": 881.845
    },
    {
      "epoch": 1.2911296436694466,
      "grad_norm": 0.3546872138977051,
      "learning_rate": 7.118987341772152e-05,
      "loss": 0.1491,
      "num_input_tokens_seen": 2387894,
      "step": 1278,
      "train_runtime": 2707.2325,
      "train_tokens_per_second": 882.042
    },
    {
      "epoch": 1.2921405104877433,
      "grad_norm": 0.34062474966049194,
      "learning_rate": 7.10886075949367e-05,
      "loss": 0.0945,
      "num_input_tokens_seen": 2390248,
      "step": 1279,
      "train_runtime": 2709.2367,
      "train_tokens_per_second": 882.259
    },
    {
      "epoch": 1.29315137730604,
      "grad_norm": 0.37184667587280273,
      "learning_rate": 7.098734177215191e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 2391690,
      "step": 1280,
      "train_runtime": 2710.8289,
      "train_tokens_per_second": 882.273
    },
    {
      "epoch": 1.2941622441243366,
      "grad_norm": 0.4290751814842224,
      "learning_rate": 7.088607594936709e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 2393278,
      "step": 1281,
      "train_runtime": 2712.516,
      "train_tokens_per_second": 882.309
    },
    {
      "epoch": 1.2951731109426334,
      "grad_norm": 0.29007312655448914,
      "learning_rate": 7.078481012658228e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 2396084,
      "step": 1282,
      "train_runtime": 2714.8145,
      "train_tokens_per_second": 882.596
    },
    {
      "epoch": 1.29618397776093,
      "grad_norm": 0.23827768862247467,
      "learning_rate": 7.068354430379747e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 2398130,
      "step": 1283,
      "train_runtime": 2716.6148,
      "train_tokens_per_second": 882.764
    },
    {
      "epoch": 1.2971948445792267,
      "grad_norm": 0.28175145387649536,
      "learning_rate": 7.058227848101266e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 2399778,
      "step": 1284,
      "train_runtime": 2718.2537,
      "train_tokens_per_second": 882.838
    },
    {
      "epoch": 1.2982057113975234,
      "grad_norm": 0.44550999999046326,
      "learning_rate": 7.048101265822785e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 2401530,
      "step": 1285,
      "train_runtime": 2719.9545,
      "train_tokens_per_second": 882.93
    },
    {
      "epoch": 1.29921657821582,
      "grad_norm": 0.3413219749927521,
      "learning_rate": 7.037974683544304e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 2403522,
      "step": 1286,
      "train_runtime": 2721.7392,
      "train_tokens_per_second": 883.083
    },
    {
      "epoch": 1.3002274450341167,
      "grad_norm": 0.36198854446411133,
      "learning_rate": 7.027848101265823e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 2405756,
      "step": 1287,
      "train_runtime": 2723.6802,
      "train_tokens_per_second": 883.274
    },
    {
      "epoch": 1.3012383118524133,
      "grad_norm": 0.3648776710033417,
      "learning_rate": 7.017721518987342e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 2407866,
      "step": 1288,
      "train_runtime": 2725.5304,
      "train_tokens_per_second": 883.449
    },
    {
      "epoch": 1.3022491786707102,
      "grad_norm": 0.21727511286735535,
      "learning_rate": 7.007594936708861e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 2409742,
      "step": 1289,
      "train_runtime": 2727.2362,
      "train_tokens_per_second": 883.584
    },
    {
      "epoch": 1.3032600454890069,
      "grad_norm": 0.37237033247947693,
      "learning_rate": 6.99746835443038e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 2411498,
      "step": 1290,
      "train_runtime": 2728.9538,
      "train_tokens_per_second": 883.671
    },
    {
      "epoch": 1.3042709123073035,
      "grad_norm": 0.35302239656448364,
      "learning_rate": 6.987341772151898e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 2413020,
      "step": 1291,
      "train_runtime": 2732.5694,
      "train_tokens_per_second": 883.059
    },
    {
      "epoch": 1.3052817791256002,
      "grad_norm": 0.45146867632865906,
      "learning_rate": 6.977215189873419e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 2414744,
      "step": 1292,
      "train_runtime": 2734.3149,
      "train_tokens_per_second": 883.126
    },
    {
      "epoch": 1.3062926459438968,
      "grad_norm": 0.30090656876564026,
      "learning_rate": 6.967088607594938e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 2416852,
      "step": 1293,
      "train_runtime": 2736.1742,
      "train_tokens_per_second": 883.296
    },
    {
      "epoch": 1.3073035127621937,
      "grad_norm": 0.33946192264556885,
      "learning_rate": 6.956962025316455e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 2418370,
      "step": 1294,
      "train_runtime": 2737.8548,
      "train_tokens_per_second": 883.308
    },
    {
      "epoch": 1.3083143795804903,
      "grad_norm": 0.44936761260032654,
      "learning_rate": 6.946835443037975e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 2420094,
      "step": 1295,
      "train_runtime": 2739.5672,
      "train_tokens_per_second": 883.386
    },
    {
      "epoch": 1.309325246398787,
      "grad_norm": 0.37103450298309326,
      "learning_rate": 6.936708860759494e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 2422344,
      "step": 1296,
      "train_runtime": 2741.5021,
      "train_tokens_per_second": 883.583
    },
    {
      "epoch": 1.3103361132170837,
      "grad_norm": 0.27331382036209106,
      "learning_rate": 6.926582278481013e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 2424196,
      "step": 1297,
      "train_runtime": 2743.1713,
      "train_tokens_per_second": 883.72
    },
    {
      "epoch": 1.3113469800353803,
      "grad_norm": 0.46463069319725037,
      "learning_rate": 6.916455696202532e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 2425842,
      "step": 1298,
      "train_runtime": 2744.9191,
      "train_tokens_per_second": 883.757
    },
    {
      "epoch": 1.312357846853677,
      "grad_norm": 0.41775521636009216,
      "learning_rate": 6.906329113924051e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 2427566,
      "step": 1299,
      "train_runtime": 2746.8158,
      "train_tokens_per_second": 883.775
    },
    {
      "epoch": 1.3133687136719736,
      "grad_norm": 0.3045608699321747,
      "learning_rate": 6.89620253164557e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 2429296,
      "step": 1300,
      "train_runtime": 2748.5091,
      "train_tokens_per_second": 883.86
    },
    {
      "epoch": 1.3143795804902705,
      "grad_norm": 0.3907722532749176,
      "learning_rate": 6.886075949367089e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 2431194,
      "step": 1301,
      "train_runtime": 2750.3507,
      "train_tokens_per_second": 883.958
    },
    {
      "epoch": 1.3153904473085671,
      "grad_norm": 0.3585439622402191,
      "learning_rate": 6.875949367088608e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 2433072,
      "step": 1302,
      "train_runtime": 2752.0697,
      "train_tokens_per_second": 884.088
    },
    {
      "epoch": 1.3164013141268638,
      "grad_norm": 0.513984739780426,
      "learning_rate": 6.865822784810126e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 2435092,
      "step": 1303,
      "train_runtime": 2753.8707,
      "train_tokens_per_second": 884.243
    },
    {
      "epoch": 1.3174121809451604,
      "grad_norm": 0.4001423418521881,
      "learning_rate": 6.855696202531646e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 2437404,
      "step": 1304,
      "train_runtime": 2755.9159,
      "train_tokens_per_second": 884.426
    },
    {
      "epoch": 1.318423047763457,
      "grad_norm": 0.3177363872528076,
      "learning_rate": 6.845569620253165e-05,
      "loss": 0.1289,
      "num_input_tokens_seen": 2439236,
      "step": 1305,
      "train_runtime": 2757.6363,
      "train_tokens_per_second": 884.539
    },
    {
      "epoch": 1.319433914581754,
      "grad_norm": 0.32245272397994995,
      "learning_rate": 6.835443037974683e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 2441176,
      "step": 1306,
      "train_runtime": 2759.4412,
      "train_tokens_per_second": 884.663
    },
    {
      "epoch": 1.3204447814000506,
      "grad_norm": 0.3399496078491211,
      "learning_rate": 6.825316455696202e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 2442982,
      "step": 1307,
      "train_runtime": 2761.1924,
      "train_tokens_per_second": 884.756
    },
    {
      "epoch": 1.3214556482183473,
      "grad_norm": 0.3486602008342743,
      "learning_rate": 6.815189873417723e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 2444502,
      "step": 1308,
      "train_runtime": 2762.797,
      "train_tokens_per_second": 884.792
    },
    {
      "epoch": 1.322466515036644,
      "grad_norm": 0.28970780968666077,
      "learning_rate": 6.80506329113924e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 2446644,
      "step": 1309,
      "train_runtime": 2764.653,
      "train_tokens_per_second": 884.973
    },
    {
      "epoch": 1.3234773818549406,
      "grad_norm": 0.3342728912830353,
      "learning_rate": 6.79493670886076e-05,
      "loss": 0.1295,
      "num_input_tokens_seen": 2448848,
      "step": 1310,
      "train_runtime": 2766.6034,
      "train_tokens_per_second": 885.146
    },
    {
      "epoch": 1.3244882486732372,
      "grad_norm": 0.5038242340087891,
      "learning_rate": 6.784810126582279e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 2450452,
      "step": 1311,
      "train_runtime": 2768.2202,
      "train_tokens_per_second": 885.208
    },
    {
      "epoch": 1.3254991154915339,
      "grad_norm": 0.44264596700668335,
      "learning_rate": 6.774683544303798e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 2452540,
      "step": 1312,
      "train_runtime": 2769.9719,
      "train_tokens_per_second": 885.402
    },
    {
      "epoch": 1.3265099823098307,
      "grad_norm": 0.337912380695343,
      "learning_rate": 6.764556962025317e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 2454236,
      "step": 1313,
      "train_runtime": 2771.6721,
      "train_tokens_per_second": 885.471
    },
    {
      "epoch": 1.3275208491281274,
      "grad_norm": 0.2829691469669342,
      "learning_rate": 6.754430379746836e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 2456178,
      "step": 1314,
      "train_runtime": 2773.5032,
      "train_tokens_per_second": 885.587
    },
    {
      "epoch": 1.328531715946424,
      "grad_norm": 0.18844102323055267,
      "learning_rate": 6.744303797468355e-05,
      "loss": 0.0127,
      "num_input_tokens_seen": 2457798,
      "step": 1315,
      "train_runtime": 2775.1247,
      "train_tokens_per_second": 885.653
    },
    {
      "epoch": 1.3295425827647207,
      "grad_norm": 0.3522207736968994,
      "learning_rate": 6.734177215189874e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 2459950,
      "step": 1316,
      "train_runtime": 2776.9924,
      "train_tokens_per_second": 885.832
    },
    {
      "epoch": 1.3305534495830174,
      "grad_norm": 0.36738115549087524,
      "learning_rate": 6.724050632911393e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 2461440,
      "step": 1317,
      "train_runtime": 2778.6313,
      "train_tokens_per_second": 885.846
    },
    {
      "epoch": 1.3315643164013142,
      "grad_norm": 0.408785879611969,
      "learning_rate": 6.713924050632912e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 2463362,
      "step": 1318,
      "train_runtime": 2780.431,
      "train_tokens_per_second": 885.964
    },
    {
      "epoch": 1.3325751832196109,
      "grad_norm": 0.3836084306240082,
      "learning_rate": 6.70379746835443e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 2465416,
      "step": 1319,
      "train_runtime": 2782.2881,
      "train_tokens_per_second": 886.111
    },
    {
      "epoch": 1.3335860500379075,
      "grad_norm": 0.4433416724205017,
      "learning_rate": 6.69367088607595e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 2467378,
      "step": 1320,
      "train_runtime": 2784.1379,
      "train_tokens_per_second": 886.227
    },
    {
      "epoch": 1.3345969168562042,
      "grad_norm": 0.29245951771736145,
      "learning_rate": 6.68354430379747e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 2469082,
      "step": 1321,
      "train_runtime": 2787.5549,
      "train_tokens_per_second": 885.752
    },
    {
      "epoch": 1.3356077836745008,
      "grad_norm": 0.356962651014328,
      "learning_rate": 6.673417721518987e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 2470532,
      "step": 1322,
      "train_runtime": 2789.1885,
      "train_tokens_per_second": 885.753
    },
    {
      "epoch": 1.3366186504927975,
      "grad_norm": 0.38898488879203796,
      "learning_rate": 6.663291139240506e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 2471952,
      "step": 1323,
      "train_runtime": 2790.7695,
      "train_tokens_per_second": 885.76
    },
    {
      "epoch": 1.3376295173110941,
      "grad_norm": 0.33297595381736755,
      "learning_rate": 6.653164556962027e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 2473502,
      "step": 1324,
      "train_runtime": 2792.3898,
      "train_tokens_per_second": 885.801
    },
    {
      "epoch": 1.338640384129391,
      "grad_norm": 0.3344525694847107,
      "learning_rate": 6.643037974683544e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 2475404,
      "step": 1325,
      "train_runtime": 2794.0516,
      "train_tokens_per_second": 885.955
    },
    {
      "epoch": 1.3396512509476877,
      "grad_norm": 0.25987255573272705,
      "learning_rate": 6.632911392405063e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 2477228,
      "step": 1326,
      "train_runtime": 2795.7635,
      "train_tokens_per_second": 886.065
    },
    {
      "epoch": 1.3406621177659843,
      "grad_norm": 0.36128169298171997,
      "learning_rate": 6.622784810126583e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 2479076,
      "step": 1327,
      "train_runtime": 2797.5404,
      "train_tokens_per_second": 886.163
    },
    {
      "epoch": 1.341672984584281,
      "grad_norm": 0.3184932768344879,
      "learning_rate": 6.612658227848102e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 2480822,
      "step": 1328,
      "train_runtime": 2799.2869,
      "train_tokens_per_second": 886.234
    },
    {
      "epoch": 1.3426838514025778,
      "grad_norm": 0.33535587787628174,
      "learning_rate": 6.602531645569621e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 2482556,
      "step": 1329,
      "train_runtime": 2801.035,
      "train_tokens_per_second": 886.3
    },
    {
      "epoch": 1.3436947182208745,
      "grad_norm": 0.5460361838340759,
      "learning_rate": 6.59240506329114e-05,
      "loss": 0.1411,
      "num_input_tokens_seen": 2484660,
      "step": 1330,
      "train_runtime": 2802.9043,
      "train_tokens_per_second": 886.459
    },
    {
      "epoch": 1.3447055850391711,
      "grad_norm": 0.40229368209838867,
      "learning_rate": 6.582278481012658e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 2486850,
      "step": 1331,
      "train_runtime": 2804.8425,
      "train_tokens_per_second": 886.627
    },
    {
      "epoch": 1.3457164518574678,
      "grad_norm": 0.5042855143547058,
      "learning_rate": 6.572151898734178e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 2488602,
      "step": 1332,
      "train_runtime": 2806.5521,
      "train_tokens_per_second": 886.711
    },
    {
      "epoch": 1.3467273186757645,
      "grad_norm": 0.5470770597457886,
      "learning_rate": 6.562025316455697e-05,
      "loss": 0.1217,
      "num_input_tokens_seen": 2490866,
      "step": 1333,
      "train_runtime": 2808.483,
      "train_tokens_per_second": 886.908
    },
    {
      "epoch": 1.347738185494061,
      "grad_norm": 0.42767834663391113,
      "learning_rate": 6.551898734177215e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 2492602,
      "step": 1334,
      "train_runtime": 2810.1877,
      "train_tokens_per_second": 886.988
    },
    {
      "epoch": 1.3487490523123578,
      "grad_norm": 0.40733814239501953,
      "learning_rate": 6.541772151898734e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 2494526,
      "step": 1335,
      "train_runtime": 2811.9748,
      "train_tokens_per_second": 887.108
    },
    {
      "epoch": 1.3497599191306544,
      "grad_norm": 0.41067591309547424,
      "learning_rate": 6.531645569620254e-05,
      "loss": 0.1522,
      "num_input_tokens_seen": 2496016,
      "step": 1336,
      "train_runtime": 2813.6241,
      "train_tokens_per_second": 887.118
    },
    {
      "epoch": 1.3507707859489513,
      "grad_norm": 0.26213085651397705,
      "learning_rate": 6.521518987341772e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 2497684,
      "step": 1337,
      "train_runtime": 2815.2457,
      "train_tokens_per_second": 887.199
    },
    {
      "epoch": 1.351781652767248,
      "grad_norm": 0.4205595850944519,
      "learning_rate": 6.511392405063291e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 2499736,
      "step": 1338,
      "train_runtime": 2817.045,
      "train_tokens_per_second": 887.361
    },
    {
      "epoch": 1.3527925195855446,
      "grad_norm": 0.3436855971813202,
      "learning_rate": 6.50126582278481e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 2501926,
      "step": 1339,
      "train_runtime": 2818.916,
      "train_tokens_per_second": 887.549
    },
    {
      "epoch": 1.3538033864038412,
      "grad_norm": 0.31082549691200256,
      "learning_rate": 6.491139240506329e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 2503974,
      "step": 1340,
      "train_runtime": 2820.7687,
      "train_tokens_per_second": 887.692
    },
    {
      "epoch": 1.354814253222138,
      "grad_norm": 0.3663766086101532,
      "learning_rate": 6.481012658227848e-05,
      "loss": 0.1151,
      "num_input_tokens_seen": 2506114,
      "step": 1341,
      "train_runtime": 2822.5852,
      "train_tokens_per_second": 887.879
    },
    {
      "epoch": 1.3558251200404348,
      "grad_norm": 0.341423362493515,
      "learning_rate": 6.470886075949367e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 2507914,
      "step": 1342,
      "train_runtime": 2824.3415,
      "train_tokens_per_second": 887.964
    },
    {
      "epoch": 1.3568359868587314,
      "grad_norm": 0.4148929715156555,
      "learning_rate": 6.460759493670887e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 2509662,
      "step": 1343,
      "train_runtime": 2826.1124,
      "train_tokens_per_second": 888.026
    },
    {
      "epoch": 1.357846853677028,
      "grad_norm": 0.35411393642425537,
      "learning_rate": 6.450632911392406e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 2511950,
      "step": 1344,
      "train_runtime": 2828.1134,
      "train_tokens_per_second": 888.207
    },
    {
      "epoch": 1.3588577204953247,
      "grad_norm": 0.3845215141773224,
      "learning_rate": 6.440506329113925e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 2513682,
      "step": 1345,
      "train_runtime": 2829.7547,
      "train_tokens_per_second": 888.304
    },
    {
      "epoch": 1.3598685873136214,
      "grad_norm": 0.30208733677864075,
      "learning_rate": 6.430379746835444e-05,
      "loss": 0.2562,
      "num_input_tokens_seen": 2515712,
      "step": 1346,
      "train_runtime": 2831.6435,
      "train_tokens_per_second": 888.428
    },
    {
      "epoch": 1.360879454131918,
      "grad_norm": 0.28983309864997864,
      "learning_rate": 6.420253164556962e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 2517720,
      "step": 1347,
      "train_runtime": 2833.4729,
      "train_tokens_per_second": 888.563
    },
    {
      "epoch": 1.3618903209502147,
      "grad_norm": 0.2847015857696533,
      "learning_rate": 6.410126582278482e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 2519868,
      "step": 1348,
      "train_runtime": 2835.3992,
      "train_tokens_per_second": 888.717
    },
    {
      "epoch": 1.3629011877685115,
      "grad_norm": 0.2689174711704254,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 2521818,
      "step": 1349,
      "train_runtime": 2837.0755,
      "train_tokens_per_second": 888.879
    },
    {
      "epoch": 1.3639120545868082,
      "grad_norm": 0.41363146901130676,
      "learning_rate": 6.389873417721519e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 2523804,
      "step": 1350,
      "train_runtime": 2838.8603,
      "train_tokens_per_second": 889.02
    },
    {
      "epoch": 1.3649229214051048,
      "grad_norm": 0.29461362957954407,
      "learning_rate": 6.379746835443038e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 2525558,
      "step": 1351,
      "train_runtime": 2842.5738,
      "train_tokens_per_second": 888.476
    },
    {
      "epoch": 1.3659337882234015,
      "grad_norm": 0.3976391553878784,
      "learning_rate": 6.369620253164558e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 2527482,
      "step": 1352,
      "train_runtime": 2844.296,
      "train_tokens_per_second": 888.614
    },
    {
      "epoch": 1.3669446550416984,
      "grad_norm": 0.4617343544960022,
      "learning_rate": 6.359493670886076e-05,
      "loss": 0.1341,
      "num_input_tokens_seen": 2529434,
      "step": 1353,
      "train_runtime": 2846.0612,
      "train_tokens_per_second": 888.749
    },
    {
      "epoch": 1.367955521859995,
      "grad_norm": 0.2891669273376465,
      "learning_rate": 6.349367088607595e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 2531414,
      "step": 1354,
      "train_runtime": 2847.8729,
      "train_tokens_per_second": 888.879
    },
    {
      "epoch": 1.3689663886782917,
      "grad_norm": 0.16765184700489044,
      "learning_rate": 6.339240506329114e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 2532808,
      "step": 1355,
      "train_runtime": 2849.4694,
      "train_tokens_per_second": 888.87
    },
    {
      "epoch": 1.3699772554965883,
      "grad_norm": 0.4383516311645508,
      "learning_rate": 6.329113924050633e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 2534488,
      "step": 1356,
      "train_runtime": 2851.1636,
      "train_tokens_per_second": 888.931
    },
    {
      "epoch": 1.370988122314885,
      "grad_norm": 0.3356960117816925,
      "learning_rate": 6.318987341772152e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 2536656,
      "step": 1357,
      "train_runtime": 2853.1036,
      "train_tokens_per_second": 889.087
    },
    {
      "epoch": 1.3719989891331816,
      "grad_norm": 0.24229836463928223,
      "learning_rate": 6.308860759493671e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 2538702,
      "step": 1358,
      "train_runtime": 2854.9724,
      "train_tokens_per_second": 889.221
    },
    {
      "epoch": 1.3730098559514783,
      "grad_norm": 0.41483476758003235,
      "learning_rate": 6.298734177215189e-05,
      "loss": 0.1443,
      "num_input_tokens_seen": 2540198,
      "step": 1359,
      "train_runtime": 2856.5761,
      "train_tokens_per_second": 889.246
    },
    {
      "epoch": 1.3740207227697752,
      "grad_norm": 0.36345934867858887,
      "learning_rate": 6.28860759493671e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 2542196,
      "step": 1360,
      "train_runtime": 2858.4441,
      "train_tokens_per_second": 889.364
    },
    {
      "epoch": 1.3750315895880718,
      "grad_norm": 0.4974648058414459,
      "learning_rate": 6.278481012658229e-05,
      "loss": 0.1213,
      "num_input_tokens_seen": 2543920,
      "step": 1361,
      "train_runtime": 2860.1455,
      "train_tokens_per_second": 889.437
    },
    {
      "epoch": 1.3760424564063685,
      "grad_norm": 0.3402421474456787,
      "learning_rate": 6.268354430379746e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 2545776,
      "step": 1362,
      "train_runtime": 2861.9062,
      "train_tokens_per_second": 889.539
    },
    {
      "epoch": 1.3770533232246651,
      "grad_norm": 0.20843306183815002,
      "learning_rate": 6.258227848101266e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 2547750,
      "step": 1363,
      "train_runtime": 2863.7086,
      "train_tokens_per_second": 889.668
    },
    {
      "epoch": 1.3780641900429618,
      "grad_norm": 0.36995962262153625,
      "learning_rate": 6.248101265822786e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 2549424,
      "step": 1364,
      "train_runtime": 2865.4682,
      "train_tokens_per_second": 889.706
    },
    {
      "epoch": 1.3790750568612586,
      "grad_norm": 0.34429794549942017,
      "learning_rate": 6.237974683544304e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 2550926,
      "step": 1365,
      "train_runtime": 2867.1842,
      "train_tokens_per_second": 889.697
    },
    {
      "epoch": 1.3800859236795553,
      "grad_norm": 0.27338913083076477,
      "learning_rate": 6.227848101265823e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 2552352,
      "step": 1366,
      "train_runtime": 2868.7893,
      "train_tokens_per_second": 889.697
    },
    {
      "epoch": 1.381096790497852,
      "grad_norm": 0.41747793555259705,
      "learning_rate": 6.217721518987342e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 2554060,
      "step": 1367,
      "train_runtime": 2870.5414,
      "train_tokens_per_second": 889.749
    },
    {
      "epoch": 1.3821076573161486,
      "grad_norm": 0.3080999255180359,
      "learning_rate": 6.207594936708861e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 2556302,
      "step": 1368,
      "train_runtime": 2872.4817,
      "train_tokens_per_second": 889.928
    },
    {
      "epoch": 1.3831185241344452,
      "grad_norm": 0.27430495619773865,
      "learning_rate": 6.19746835443038e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 2558024,
      "step": 1369,
      "train_runtime": 2874.1655,
      "train_tokens_per_second": 890.006
    },
    {
      "epoch": 1.384129390952742,
      "grad_norm": 0.21422269940376282,
      "learning_rate": 6.187341772151899e-05,
      "loss": 0.1405,
      "num_input_tokens_seen": 2560244,
      "step": 1370,
      "train_runtime": 2876.0679,
      "train_tokens_per_second": 890.189
    },
    {
      "epoch": 1.3851402577710386,
      "grad_norm": 0.295729398727417,
      "learning_rate": 6.177215189873418e-05,
      "loss": 0.015,
      "num_input_tokens_seen": 2562024,
      "step": 1371,
      "train_runtime": 2877.7989,
      "train_tokens_per_second": 890.272
    },
    {
      "epoch": 1.3861511245893354,
      "grad_norm": 0.4139345586299896,
      "learning_rate": 6.167088607594937e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 2563776,
      "step": 1372,
      "train_runtime": 2879.422,
      "train_tokens_per_second": 890.379
    },
    {
      "epoch": 1.387161991407632,
      "grad_norm": 0.2090672254562378,
      "learning_rate": 6.156962025316456e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 2565618,
      "step": 1373,
      "train_runtime": 2881.1338,
      "train_tokens_per_second": 890.489
    },
    {
      "epoch": 1.3881728582259287,
      "grad_norm": 0.3067643344402313,
      "learning_rate": 6.146835443037975e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 2567786,
      "step": 1374,
      "train_runtime": 2883.0016,
      "train_tokens_per_second": 890.664
    },
    {
      "epoch": 1.3891837250442254,
      "grad_norm": 0.3311604857444763,
      "learning_rate": 6.136708860759493e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 2569640,
      "step": 1375,
      "train_runtime": 2884.7164,
      "train_tokens_per_second": 890.777
    },
    {
      "epoch": 1.390194591862522,
      "grad_norm": 0.424031138420105,
      "learning_rate": 6.126582278481014e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 2571758,
      "step": 1376,
      "train_runtime": 2886.5299,
      "train_tokens_per_second": 890.951
    },
    {
      "epoch": 1.391205458680819,
      "grad_norm": 0.5352917909622192,
      "learning_rate": 6.116455696202533e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 2573584,
      "step": 1377,
      "train_runtime": 2888.2479,
      "train_tokens_per_second": 891.054
    },
    {
      "epoch": 1.3922163254991156,
      "grad_norm": 0.2344108521938324,
      "learning_rate": 6.10632911392405e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 2576038,
      "step": 1378,
      "train_runtime": 2890.3425,
      "train_tokens_per_second": 891.257
    },
    {
      "epoch": 1.3932271923174122,
      "grad_norm": 0.33207178115844727,
      "learning_rate": 6.0962025316455695e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 2578072,
      "step": 1379,
      "train_runtime": 2892.1479,
      "train_tokens_per_second": 891.404
    },
    {
      "epoch": 1.3942380591357089,
      "grad_norm": 0.2377287745475769,
      "learning_rate": 6.086075949367089e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 2579774,
      "step": 1380,
      "train_runtime": 2893.905,
      "train_tokens_per_second": 891.451
    },
    {
      "epoch": 1.3952489259540055,
      "grad_norm": 0.24418236315250397,
      "learning_rate": 6.0759493670886084e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 2581642,
      "step": 1381,
      "train_runtime": 2897.7031,
      "train_tokens_per_second": 890.927
    },
    {
      "epoch": 1.3962597927723022,
      "grad_norm": 0.22728894650936127,
      "learning_rate": 6.065822784810127e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 2583334,
      "step": 1382,
      "train_runtime": 2899.3988,
      "train_tokens_per_second": 890.99
    },
    {
      "epoch": 1.3972706595905988,
      "grad_norm": 0.33506718277931213,
      "learning_rate": 6.055696202531645e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 2585250,
      "step": 1383,
      "train_runtime": 2901.198,
      "train_tokens_per_second": 891.097
    },
    {
      "epoch": 1.3982815264088957,
      "grad_norm": 0.3591923415660858,
      "learning_rate": 6.0455696202531656e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 2587364,
      "step": 1384,
      "train_runtime": 2903.069,
      "train_tokens_per_second": 891.251
    },
    {
      "epoch": 1.3992923932271923,
      "grad_norm": 0.2582985758781433,
      "learning_rate": 6.035443037974684e-05,
      "loss": 0.1784,
      "num_input_tokens_seen": 2588714,
      "step": 1385,
      "train_runtime": 2904.6736,
      "train_tokens_per_second": 891.224
    },
    {
      "epoch": 1.400303260045489,
      "grad_norm": 0.3208470046520233,
      "learning_rate": 6.0253164556962024e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 2590470,
      "step": 1386,
      "train_runtime": 2906.3788,
      "train_tokens_per_second": 891.305
    },
    {
      "epoch": 1.400303260045489,
      "eval_loss": 0.21137197315692902,
      "eval_runtime": 61.4915,
      "eval_samples_per_second": 14.295,
      "eval_steps_per_second": 7.155,
      "num_input_tokens_seen": 2590470,
      "step": 1386
    },
    {
      "epoch": 1.4013141268637856,
      "grad_norm": 0.28233349323272705,
      "learning_rate": 6.0151898734177215e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 2592686,
      "step": 1387,
      "train_runtime": 2969.748,
      "train_tokens_per_second": 873.032
    },
    {
      "epoch": 1.4023249936820825,
      "grad_norm": 0.3834882080554962,
      "learning_rate": 6.005063291139241e-05,
      "loss": 0.1386,
      "num_input_tokens_seen": 2594374,
      "step": 1388,
      "train_runtime": 2971.3719,
      "train_tokens_per_second": 873.123
    },
    {
      "epoch": 1.4033358605003792,
      "grad_norm": 0.3518659472465515,
      "learning_rate": 5.99493670886076e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 2596580,
      "step": 1389,
      "train_runtime": 2973.2451,
      "train_tokens_per_second": 873.315
    },
    {
      "epoch": 1.4043467273186758,
      "grad_norm": 0.37290677428245544,
      "learning_rate": 5.984810126582279e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 2598032,
      "step": 1390,
      "train_runtime": 2974.8336,
      "train_tokens_per_second": 873.337
    },
    {
      "epoch": 1.4053575941369725,
      "grad_norm": 0.4034007489681244,
      "learning_rate": 5.974683544303797e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 2599522,
      "step": 1391,
      "train_runtime": 2976.4546,
      "train_tokens_per_second": 873.362
    },
    {
      "epoch": 1.4063684609552691,
      "grad_norm": 0.30299344658851624,
      "learning_rate": 5.964556962025317e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 2601542,
      "step": 1392,
      "train_runtime": 2978.236,
      "train_tokens_per_second": 873.518
    },
    {
      "epoch": 1.4073793277735658,
      "grad_norm": 0.3894141614437103,
      "learning_rate": 5.954430379746836e-05,
      "loss": 0.0768,
      "num_input_tokens_seen": 2603504,
      "step": 1393,
      "train_runtime": 2980.0739,
      "train_tokens_per_second": 873.637
    },
    {
      "epoch": 1.4083901945918624,
      "grad_norm": 0.2617016136646271,
      "learning_rate": 5.9443037974683544e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 2605752,
      "step": 1394,
      "train_runtime": 2981.9398,
      "train_tokens_per_second": 873.845
    },
    {
      "epoch": 1.409401061410159,
      "grad_norm": 0.24910065531730652,
      "learning_rate": 5.9341772151898735e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 2607290,
      "step": 1395,
      "train_runtime": 2983.5491,
      "train_tokens_per_second": 873.889
    },
    {
      "epoch": 1.410411928228456,
      "grad_norm": 0.3577074408531189,
      "learning_rate": 5.924050632911393e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 2609448,
      "step": 1396,
      "train_runtime": 2985.4559,
      "train_tokens_per_second": 874.053
    },
    {
      "epoch": 1.4114227950467526,
      "grad_norm": 0.46218645572662354,
      "learning_rate": 5.913924050632912e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 2611600,
      "step": 1397,
      "train_runtime": 2987.3305,
      "train_tokens_per_second": 874.225
    },
    {
      "epoch": 1.4124336618650493,
      "grad_norm": 0.5550116300582886,
      "learning_rate": 5.903797468354431e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 2612936,
      "step": 1398,
      "train_runtime": 2988.9389,
      "train_tokens_per_second": 874.202
    },
    {
      "epoch": 1.413444528683346,
      "grad_norm": 0.5216653943061829,
      "learning_rate": 5.893670886075949e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 2614638,
      "step": 1399,
      "train_runtime": 2990.6319,
      "train_tokens_per_second": 874.276
    },
    {
      "epoch": 1.4144553955016428,
      "grad_norm": 0.36964768171310425,
      "learning_rate": 5.883544303797469e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 2616204,
      "step": 1400,
      "train_runtime": 2992.2461,
      "train_tokens_per_second": 874.328
    },
    {
      "epoch": 1.4154662623199394,
      "grad_norm": 0.38824814558029175,
      "learning_rate": 5.873417721518988e-05,
      "loss": 0.1205,
      "num_input_tokens_seen": 2618418,
      "step": 1401,
      "train_runtime": 2994.2495,
      "train_tokens_per_second": 874.482
    },
    {
      "epoch": 1.416477129138236,
      "grad_norm": 0.4073863923549652,
      "learning_rate": 5.8632911392405064e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 2620624,
      "step": 1402,
      "train_runtime": 2996.1851,
      "train_tokens_per_second": 874.654
    },
    {
      "epoch": 1.4174879959565327,
      "grad_norm": 0.5225967168807983,
      "learning_rate": 5.8531645569620255e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 2622554,
      "step": 1403,
      "train_runtime": 2997.9627,
      "train_tokens_per_second": 874.779
    },
    {
      "epoch": 1.4184988627748294,
      "grad_norm": 0.35591739416122437,
      "learning_rate": 5.843037974683545e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 2625194,
      "step": 1404,
      "train_runtime": 3000.1237,
      "train_tokens_per_second": 875.029
    },
    {
      "epoch": 1.419509729593126,
      "grad_norm": 0.3223467767238617,
      "learning_rate": 5.832911392405064e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 2627128,
      "step": 1405,
      "train_runtime": 3001.8466,
      "train_tokens_per_second": 875.171
    },
    {
      "epoch": 1.4205205964114227,
      "grad_norm": 0.36426904797554016,
      "learning_rate": 5.822784810126583e-05,
      "loss": 0.2153,
      "num_input_tokens_seen": 2629208,
      "step": 1406,
      "train_runtime": 3003.729,
      "train_tokens_per_second": 875.315
    },
    {
      "epoch": 1.4215314632297193,
      "grad_norm": 0.39519429206848145,
      "learning_rate": 5.812658227848101e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 2631256,
      "step": 1407,
      "train_runtime": 3005.5097,
      "train_tokens_per_second": 875.477
    },
    {
      "epoch": 1.4225423300480162,
      "grad_norm": 0.47091037034988403,
      "learning_rate": 5.802531645569621e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 2633626,
      "step": 1408,
      "train_runtime": 3007.5159,
      "train_tokens_per_second": 875.681
    },
    {
      "epoch": 1.4235531968663129,
      "grad_norm": 0.3183493912220001,
      "learning_rate": 5.79240506329114e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 2635146,
      "step": 1409,
      "train_runtime": 3009.2026,
      "train_tokens_per_second": 875.696
    },
    {
      "epoch": 1.4245640636846095,
      "grad_norm": 0.4309399724006653,
      "learning_rate": 5.7822784810126584e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 2636952,
      "step": 1410,
      "train_runtime": 3010.919,
      "train_tokens_per_second": 875.796
    },
    {
      "epoch": 1.4255749305029062,
      "grad_norm": 0.2628016173839569,
      "learning_rate": 5.772151898734177e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 2639144,
      "step": 1411,
      "train_runtime": 3014.8839,
      "train_tokens_per_second": 875.372
    },
    {
      "epoch": 1.426585797321203,
      "grad_norm": 0.2927537262439728,
      "learning_rate": 5.762025316455697e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 2641920,
      "step": 1412,
      "train_runtime": 3017.1754,
      "train_tokens_per_second": 875.627
    },
    {
      "epoch": 1.4275966641394997,
      "grad_norm": 0.40520450472831726,
      "learning_rate": 5.751898734177216e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 2643916,
      "step": 1413,
      "train_runtime": 3019.0157,
      "train_tokens_per_second": 875.754
    },
    {
      "epoch": 1.4286075309577964,
      "grad_norm": 0.43820634484291077,
      "learning_rate": 5.741772151898734e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 2645654,
      "step": 1414,
      "train_runtime": 3020.7261,
      "train_tokens_per_second": 875.834
    },
    {
      "epoch": 1.429618397776093,
      "grad_norm": 0.33604276180267334,
      "learning_rate": 5.731645569620253e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 2647666,
      "step": 1415,
      "train_runtime": 3022.5572,
      "train_tokens_per_second": 875.969
    },
    {
      "epoch": 1.4306292645943897,
      "grad_norm": 0.3733352720737457,
      "learning_rate": 5.721518987341773e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 2649700,
      "step": 1416,
      "train_runtime": 3024.4378,
      "train_tokens_per_second": 876.097
    },
    {
      "epoch": 1.4316401314126863,
      "grad_norm": 0.1657431572675705,
      "learning_rate": 5.7113924050632913e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 2651426,
      "step": 1417,
      "train_runtime": 3026.1256,
      "train_tokens_per_second": 876.178
    },
    {
      "epoch": 1.432650998230983,
      "grad_norm": 0.37970614433288574,
      "learning_rate": 5.7012658227848104e-05,
      "loss": 0.104,
      "num_input_tokens_seen": 2653400,
      "step": 1418,
      "train_runtime": 3027.9692,
      "train_tokens_per_second": 876.297
    },
    {
      "epoch": 1.4336618650492798,
      "grad_norm": 0.28047189116477966,
      "learning_rate": 5.691139240506329e-05,
      "loss": 0.1058,
      "num_input_tokens_seen": 2655334,
      "step": 1419,
      "train_runtime": 3029.8187,
      "train_tokens_per_second": 876.4
    },
    {
      "epoch": 1.4346727318675765,
      "grad_norm": 0.37715446949005127,
      "learning_rate": 5.6810126582278486e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 2657226,
      "step": 1420,
      "train_runtime": 3031.5956,
      "train_tokens_per_second": 876.511
    },
    {
      "epoch": 1.4356835986858731,
      "grad_norm": 0.258142352104187,
      "learning_rate": 5.670886075949368e-05,
      "loss": 0.0281,
      "num_input_tokens_seen": 2659126,
      "step": 1421,
      "train_runtime": 3033.3202,
      "train_tokens_per_second": 876.639
    },
    {
      "epoch": 1.4366944655041698,
      "grad_norm": 0.254810631275177,
      "learning_rate": 5.660759493670886e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 2660518,
      "step": 1422,
      "train_runtime": 3034.9171,
      "train_tokens_per_second": 876.636
    },
    {
      "epoch": 1.4377053323224664,
      "grad_norm": 0.1798725575208664,
      "learning_rate": 5.650632911392405e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 2663040,
      "step": 1423,
      "train_runtime": 3036.9671,
      "train_tokens_per_second": 876.875
    },
    {
      "epoch": 1.4387161991407633,
      "grad_norm": 0.25288665294647217,
      "learning_rate": 5.640506329113925e-05,
      "loss": 0.1269,
      "num_input_tokens_seen": 2664688,
      "step": 1424,
      "train_runtime": 3038.6365,
      "train_tokens_per_second": 876.935
    },
    {
      "epoch": 1.43972706595906,
      "grad_norm": 0.47461217641830444,
      "learning_rate": 5.6303797468354433e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 2666324,
      "step": 1425,
      "train_runtime": 3040.2615,
      "train_tokens_per_second": 877.005
    },
    {
      "epoch": 1.4407379327773566,
      "grad_norm": 0.31943273544311523,
      "learning_rate": 5.6202531645569624e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 2667832,
      "step": 1426,
      "train_runtime": 3041.8692,
      "train_tokens_per_second": 877.037
    },
    {
      "epoch": 1.4417487995956533,
      "grad_norm": 0.3467959463596344,
      "learning_rate": 5.610126582278481e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 2670182,
      "step": 1427,
      "train_runtime": 3043.8189,
      "train_tokens_per_second": 877.247
    },
    {
      "epoch": 1.44275966641395,
      "grad_norm": 0.2863708734512329,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 2672256,
      "step": 1428,
      "train_runtime": 3045.6698,
      "train_tokens_per_second": 877.395
    },
    {
      "epoch": 1.4437705332322466,
      "grad_norm": 0.2715647518634796,
      "learning_rate": 5.58987341772152e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 2674658,
      "step": 1429,
      "train_runtime": 3047.7358,
      "train_tokens_per_second": 877.589
    },
    {
      "epoch": 1.4447814000505432,
      "grad_norm": 0.4483523666858673,
      "learning_rate": 5.579746835443038e-05,
      "loss": 0.1065,
      "num_input_tokens_seen": 2676262,
      "step": 1430,
      "train_runtime": 3049.4143,
      "train_tokens_per_second": 877.631
    },
    {
      "epoch": 1.44579226686884,
      "grad_norm": 0.37149494886398315,
      "learning_rate": 5.569620253164557e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 2677980,
      "step": 1431,
      "train_runtime": 3051.129,
      "train_tokens_per_second": 877.701
    },
    {
      "epoch": 1.4468031336871368,
      "grad_norm": 0.4825318455696106,
      "learning_rate": 5.559493670886077e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 2679978,
      "step": 1432,
      "train_runtime": 3052.9173,
      "train_tokens_per_second": 877.842
    },
    {
      "epoch": 1.4478140005054334,
      "grad_norm": 0.3357631266117096,
      "learning_rate": 5.549367088607595e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 2681664,
      "step": 1433,
      "train_runtime": 3054.6065,
      "train_tokens_per_second": 877.908
    },
    {
      "epoch": 1.44882486732373,
      "grad_norm": 0.4023161232471466,
      "learning_rate": 5.5392405063291144e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 2683520,
      "step": 1434,
      "train_runtime": 3056.4155,
      "train_tokens_per_second": 877.996
    },
    {
      "epoch": 1.4498357341420267,
      "grad_norm": 0.3820077180862427,
      "learning_rate": 5.529113924050633e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 2685230,
      "step": 1435,
      "train_runtime": 3058.1037,
      "train_tokens_per_second": 878.07
    },
    {
      "epoch": 1.4508466009603236,
      "grad_norm": 0.34831374883651733,
      "learning_rate": 5.5189873417721526e-05,
      "loss": 0.2111,
      "num_input_tokens_seen": 2687424,
      "step": 1436,
      "train_runtime": 3060.0303,
      "train_tokens_per_second": 878.234
    },
    {
      "epoch": 1.4518574677786202,
      "grad_norm": 0.32861289381980896,
      "learning_rate": 5.508860759493672e-05,
      "loss": 0.2198,
      "num_input_tokens_seen": 2689360,
      "step": 1437,
      "train_runtime": 3061.8241,
      "train_tokens_per_second": 878.352
    },
    {
      "epoch": 1.4528683345969169,
      "grad_norm": 0.23808349668979645,
      "learning_rate": 5.49873417721519e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 2691382,
      "step": 1438,
      "train_runtime": 3063.6001,
      "train_tokens_per_second": 878.503
    },
    {
      "epoch": 1.4538792014152135,
      "grad_norm": 0.14567086100578308,
      "learning_rate": 5.488607594936709e-05,
      "loss": 0.0224,
      "num_input_tokens_seen": 2693508,
      "step": 1439,
      "train_runtime": 3065.5261,
      "train_tokens_per_second": 878.645
    },
    {
      "epoch": 1.4548900682335102,
      "grad_norm": 0.22057169675827026,
      "learning_rate": 5.478481012658229e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 2695340,
      "step": 1440,
      "train_runtime": 3067.2633,
      "train_tokens_per_second": 878.744
    },
    {
      "epoch": 1.4559009350518068,
      "grad_norm": 0.3443472981452942,
      "learning_rate": 5.468354430379747e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 2696972,
      "step": 1441,
      "train_runtime": 3070.6772,
      "train_tokens_per_second": 878.299
    },
    {
      "epoch": 1.4569118018701035,
      "grad_norm": 0.23556926846504211,
      "learning_rate": 5.4582278481012664e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 2699172,
      "step": 1442,
      "train_runtime": 3072.6709,
      "train_tokens_per_second": 878.445
    },
    {
      "epoch": 1.4579226686884004,
      "grad_norm": 0.36331993341445923,
      "learning_rate": 5.448101265822785e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 2700880,
      "step": 1443,
      "train_runtime": 3074.3674,
      "train_tokens_per_second": 878.516
    },
    {
      "epoch": 1.458933535506697,
      "grad_norm": 0.4046872556209564,
      "learning_rate": 5.437974683544303e-05,
      "loss": 0.0983,
      "num_input_tokens_seen": 2702438,
      "step": 1444,
      "train_runtime": 3076.0295,
      "train_tokens_per_second": 878.547
    },
    {
      "epoch": 1.4599444023249937,
      "grad_norm": 0.2128485143184662,
      "learning_rate": 5.427848101265823e-05,
      "loss": 0.0173,
      "num_input_tokens_seen": 2704200,
      "step": 1445,
      "train_runtime": 3077.6838,
      "train_tokens_per_second": 878.648
    },
    {
      "epoch": 1.4609552691432903,
      "grad_norm": 0.38131779432296753,
      "learning_rate": 5.417721518987342e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 2706008,
      "step": 1446,
      "train_runtime": 3079.4984,
      "train_tokens_per_second": 878.717
    },
    {
      "epoch": 1.4619661359615872,
      "grad_norm": 0.35823047161102295,
      "learning_rate": 5.4075949367088605e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 2707540,
      "step": 1447,
      "train_runtime": 3081.1191,
      "train_tokens_per_second": 878.752
    },
    {
      "epoch": 1.4629770027798839,
      "grad_norm": 0.2395031452178955,
      "learning_rate": 5.3974683544303796e-05,
      "loss": 0.1408,
      "num_input_tokens_seen": 2709564,
      "step": 1448,
      "train_runtime": 3082.9246,
      "train_tokens_per_second": 878.894
    },
    {
      "epoch": 1.4639878695981805,
      "grad_norm": 0.40758129954338074,
      "learning_rate": 5.387341772151899e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 2710994,
      "step": 1449,
      "train_runtime": 3084.5386,
      "train_tokens_per_second": 878.898
    },
    {
      "epoch": 1.4649987364164772,
      "grad_norm": 0.31139180064201355,
      "learning_rate": 5.377215189873418e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 2712410,
      "step": 1450,
      "train_runtime": 3086.1274,
      "train_tokens_per_second": 878.904
    },
    {
      "epoch": 1.4660096032347738,
      "grad_norm": 0.2450462430715561,
      "learning_rate": 5.367088607594937e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 2714996,
      "step": 1451,
      "train_runtime": 3088.2612,
      "train_tokens_per_second": 879.134
    },
    {
      "epoch": 1.4670204700530705,
      "grad_norm": 0.40233591198921204,
      "learning_rate": 5.356962025316455e-05,
      "loss": 0.0536,
      "num_input_tokens_seen": 2716772,
      "step": 1452,
      "train_runtime": 3089.9649,
      "train_tokens_per_second": 879.224
    },
    {
      "epoch": 1.468031336871367,
      "grad_norm": 0.24486950039863586,
      "learning_rate": 5.346835443037975e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 2718616,
      "step": 1453,
      "train_runtime": 3091.6601,
      "train_tokens_per_second": 879.339
    },
    {
      "epoch": 1.4690422036896638,
      "grad_norm": 0.42018240690231323,
      "learning_rate": 5.336708860759494e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 2720192,
      "step": 1454,
      "train_runtime": 3093.2788,
      "train_tokens_per_second": 879.388
    },
    {
      "epoch": 1.4700530705079606,
      "grad_norm": 0.198269322514534,
      "learning_rate": 5.3265822784810125e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 2722282,
      "step": 1455,
      "train_runtime": 3095.1375,
      "train_tokens_per_second": 879.535
    },
    {
      "epoch": 1.4710639373262573,
      "grad_norm": 0.20242540538311005,
      "learning_rate": 5.3164556962025316e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 2724290,
      "step": 1456,
      "train_runtime": 3096.984,
      "train_tokens_per_second": 879.659
    },
    {
      "epoch": 1.472074804144554,
      "grad_norm": 0.3628647029399872,
      "learning_rate": 5.306329113924051e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 2726146,
      "step": 1457,
      "train_runtime": 3098.6931,
      "train_tokens_per_second": 879.773
    },
    {
      "epoch": 1.4730856709628506,
      "grad_norm": 0.43536707758903503,
      "learning_rate": 5.29620253164557e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 2727922,
      "step": 1458,
      "train_runtime": 3100.4723,
      "train_tokens_per_second": 879.841
    },
    {
      "epoch": 1.4740965377811475,
      "grad_norm": 0.3844117522239685,
      "learning_rate": 5.286075949367089e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 2729780,
      "step": 1459,
      "train_runtime": 3102.2896,
      "train_tokens_per_second": 879.924
    },
    {
      "epoch": 1.4751074045994441,
      "grad_norm": 0.27764007449150085,
      "learning_rate": 5.275949367088607e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 2731426,
      "step": 1460,
      "train_runtime": 3103.9148,
      "train_tokens_per_second": 879.994
    },
    {
      "epoch": 1.4761182714177408,
      "grad_norm": 0.3226267993450165,
      "learning_rate": 5.265822784810127e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 2733218,
      "step": 1461,
      "train_runtime": 3105.6197,
      "train_tokens_per_second": 880.088
    },
    {
      "epoch": 1.4771291382360374,
      "grad_norm": 0.4703887403011322,
      "learning_rate": 5.255696202531646e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 2734970,
      "step": 1462,
      "train_runtime": 3107.3154,
      "train_tokens_per_second": 880.171
    },
    {
      "epoch": 1.478140005054334,
      "grad_norm": 0.37779703736305237,
      "learning_rate": 5.2455696202531645e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 2737162,
      "step": 1463,
      "train_runtime": 3109.3016,
      "train_tokens_per_second": 880.314
    },
    {
      "epoch": 1.4791508718726307,
      "grad_norm": 0.27329879999160767,
      "learning_rate": 5.2354430379746836e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 2739624,
      "step": 1464,
      "train_runtime": 3111.3646,
      "train_tokens_per_second": 880.522
    },
    {
      "epoch": 1.4801617386909274,
      "grad_norm": 0.3650558888912201,
      "learning_rate": 5.225316455696203e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 2741206,
      "step": 1465,
      "train_runtime": 3112.956,
      "train_tokens_per_second": 880.58
    },
    {
      "epoch": 1.481172605509224,
      "grad_norm": 0.42867252230644226,
      "learning_rate": 5.215189873417722e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 2743158,
      "step": 1466,
      "train_runtime": 3114.7956,
      "train_tokens_per_second": 880.686
    },
    {
      "epoch": 1.482183472327521,
      "grad_norm": 0.36611321568489075,
      "learning_rate": 5.205063291139241e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 2744750,
      "step": 1467,
      "train_runtime": 3116.4765,
      "train_tokens_per_second": 880.722
    },
    {
      "epoch": 1.4831943391458176,
      "grad_norm": 0.30326777696609497,
      "learning_rate": 5.194936708860759e-05,
      "loss": 0.1159,
      "num_input_tokens_seen": 2746582,
      "step": 1468,
      "train_runtime": 3118.2017,
      "train_tokens_per_second": 880.822
    },
    {
      "epoch": 1.4842052059641142,
      "grad_norm": 0.29910722374916077,
      "learning_rate": 5.184810126582279e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 2748374,
      "step": 1469,
      "train_runtime": 3119.9215,
      "train_tokens_per_second": 880.911
    },
    {
      "epoch": 1.4852160727824109,
      "grad_norm": 0.2985762059688568,
      "learning_rate": 5.174683544303798e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 2750720,
      "step": 1470,
      "train_runtime": 3121.8744,
      "train_tokens_per_second": 881.112
    },
    {
      "epoch": 1.4862269396007077,
      "grad_norm": 0.25838208198547363,
      "learning_rate": 5.1645569620253165e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 2752434,
      "step": 1471,
      "train_runtime": 3125.6142,
      "train_tokens_per_second": 880.606
    },
    {
      "epoch": 1.4872378064190044,
      "grad_norm": 0.3674798309803009,
      "learning_rate": 5.154430379746835e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 2754166,
      "step": 1472,
      "train_runtime": 3127.2442,
      "train_tokens_per_second": 880.701
    },
    {
      "epoch": 1.488248673237301,
      "grad_norm": 0.22060970962047577,
      "learning_rate": 5.144303797468355e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 2755320,
      "step": 1473,
      "train_runtime": 3128.7706,
      "train_tokens_per_second": 880.64
    },
    {
      "epoch": 1.4892595400555977,
      "grad_norm": 0.27728044986724854,
      "learning_rate": 5.134177215189874e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 2757252,
      "step": 1474,
      "train_runtime": 3130.5213,
      "train_tokens_per_second": 880.764
    },
    {
      "epoch": 1.4902704068738943,
      "grad_norm": 0.25170692801475525,
      "learning_rate": 5.124050632911392e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 2759246,
      "step": 1475,
      "train_runtime": 3132.362,
      "train_tokens_per_second": 880.883
    },
    {
      "epoch": 1.491281273692191,
      "grad_norm": 0.31603336334228516,
      "learning_rate": 5.113924050632911e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 2761248,
      "step": 1476,
      "train_runtime": 3134.2084,
      "train_tokens_per_second": 881.003
    },
    {
      "epoch": 1.4922921405104876,
      "grad_norm": 0.3758082985877991,
      "learning_rate": 5.103797468354431e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 2762948,
      "step": 1477,
      "train_runtime": 3135.9247,
      "train_tokens_per_second": 881.063
    },
    {
      "epoch": 1.4933030073287845,
      "grad_norm": 0.369672030210495,
      "learning_rate": 5.0936708860759494e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 2764634,
      "step": 1478,
      "train_runtime": 3137.63,
      "train_tokens_per_second": 881.122
    },
    {
      "epoch": 1.4943138741470812,
      "grad_norm": 0.20284321904182434,
      "learning_rate": 5.0835443037974685e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 2766290,
      "step": 1479,
      "train_runtime": 3139.3137,
      "train_tokens_per_second": 881.177
    },
    {
      "epoch": 1.4953247409653778,
      "grad_norm": 0.29138898849487305,
      "learning_rate": 5.073417721518987e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 2768234,
      "step": 1480,
      "train_runtime": 3141.0871,
      "train_tokens_per_second": 881.298
    },
    {
      "epoch": 1.4963356077836745,
      "grad_norm": 0.4115141034126282,
      "learning_rate": 5.0632911392405066e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 2770010,
      "step": 1481,
      "train_runtime": 3142.785,
      "train_tokens_per_second": 881.387
    },
    {
      "epoch": 1.4973464746019711,
      "grad_norm": 0.34994015097618103,
      "learning_rate": 5.053164556962026e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 2771664,
      "step": 1482,
      "train_runtime": 3144.4863,
      "train_tokens_per_second": 881.436
    },
    {
      "epoch": 1.498357341420268,
      "grad_norm": 0.43992385268211365,
      "learning_rate": 5.043037974683544e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 2773310,
      "step": 1483,
      "train_runtime": 3146.1792,
      "train_tokens_per_second": 881.485
    },
    {
      "epoch": 1.4993682082385646,
      "grad_norm": 0.4387452006340027,
      "learning_rate": 5.032911392405063e-05,
      "loss": 0.0745,
      "num_input_tokens_seen": 2775210,
      "step": 1484,
      "train_runtime": 3148.0333,
      "train_tokens_per_second": 881.569
    },
    {
      "epoch": 1.5003790750568613,
      "grad_norm": 0.33656683564186096,
      "learning_rate": 5.022784810126583e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 2777192,
      "step": 1485,
      "train_runtime": 3149.8982,
      "train_tokens_per_second": 881.677
    },
    {
      "epoch": 1.501389941875158,
      "grad_norm": 0.5562018752098083,
      "learning_rate": 5.0126582278481014e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 2778950,
      "step": 1486,
      "train_runtime": 3151.6367,
      "train_tokens_per_second": 881.748
    },
    {
      "epoch": 1.5024008086934546,
      "grad_norm": 0.36017367243766785,
      "learning_rate": 5.0025316455696205e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 2780576,
      "step": 1487,
      "train_runtime": 3153.3209,
      "train_tokens_per_second": 881.793
    },
    {
      "epoch": 1.5034116755117513,
      "grad_norm": 0.3708136975765228,
      "learning_rate": 4.9924050632911396e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 2782116,
      "step": 1488,
      "train_runtime": 3154.9402,
      "train_tokens_per_second": 881.828
    },
    {
      "epoch": 1.504422542330048,
      "grad_norm": 0.42238691449165344,
      "learning_rate": 4.982278481012658e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 2784170,
      "step": 1489,
      "train_runtime": 3156.7509,
      "train_tokens_per_second": 881.973
    },
    {
      "epoch": 1.5054334091483446,
      "grad_norm": 0.3960866630077362,
      "learning_rate": 4.972151898734178e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 2786032,
      "step": 1490,
      "train_runtime": 3158.5452,
      "train_tokens_per_second": 882.062
    },
    {
      "epoch": 1.5064442759666414,
      "grad_norm": 0.3032436966896057,
      "learning_rate": 4.962025316455696e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 2788644,
      "step": 1491,
      "train_runtime": 3160.6442,
      "train_tokens_per_second": 882.302
    },
    {
      "epoch": 1.507455142784938,
      "grad_norm": 0.39984366297721863,
      "learning_rate": 4.951898734177215e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 2790430,
      "step": 1492,
      "train_runtime": 3162.3949,
      "train_tokens_per_second": 882.379
    },
    {
      "epoch": 1.5084660096032347,
      "grad_norm": 0.4460970461368561,
      "learning_rate": 4.941772151898734e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 2792424,
      "step": 1493,
      "train_runtime": 3164.2405,
      "train_tokens_per_second": 882.494
    },
    {
      "epoch": 1.5094768764215316,
      "grad_norm": 0.3381402790546417,
      "learning_rate": 4.9316455696202534e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 2794332,
      "step": 1494,
      "train_runtime": 3166.0287,
      "train_tokens_per_second": 882.598
    },
    {
      "epoch": 1.5104877432398283,
      "grad_norm": 0.36955997347831726,
      "learning_rate": 4.9215189873417725e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 2795796,
      "step": 1495,
      "train_runtime": 3167.6879,
      "train_tokens_per_second": 882.598
    },
    {
      "epoch": 1.511498610058125,
      "grad_norm": 0.407127320766449,
      "learning_rate": 4.9113924050632915e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 2798410,
      "step": 1496,
      "train_runtime": 3169.8453,
      "train_tokens_per_second": 882.822
    },
    {
      "epoch": 1.5125094768764216,
      "grad_norm": 0.38967975974082947,
      "learning_rate": 4.90126582278481e-05,
      "loss": 0.1244,
      "num_input_tokens_seen": 2800614,
      "step": 1497,
      "train_runtime": 3171.7991,
      "train_tokens_per_second": 882.973
    },
    {
      "epoch": 1.5135203436947182,
      "grad_norm": 0.28805011510849,
      "learning_rate": 4.89113924050633e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 2802866,
      "step": 1498,
      "train_runtime": 3173.806,
      "train_tokens_per_second": 883.125
    },
    {
      "epoch": 1.5145312105130149,
      "grad_norm": 0.4989212453365326,
      "learning_rate": 4.881012658227848e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 2804408,
      "step": 1499,
      "train_runtime": 3175.4213,
      "train_tokens_per_second": 883.161
    },
    {
      "epoch": 1.5155420773313115,
      "grad_norm": 0.255536824464798,
      "learning_rate": 4.870886075949367e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 2806676,
      "step": 1500,
      "train_runtime": 3177.43,
      "train_tokens_per_second": 883.316
    },
    {
      "epoch": 1.5165529441496082,
      "grad_norm": 0.3267623782157898,
      "learning_rate": 4.860759493670886e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 2808414,
      "step": 1501,
      "train_runtime": 3181.2601,
      "train_tokens_per_second": 882.799
    },
    {
      "epoch": 1.5175638109679048,
      "grad_norm": 0.30580955743789673,
      "learning_rate": 4.8506329113924054e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 2810224,
      "step": 1502,
      "train_runtime": 3183.084,
      "train_tokens_per_second": 882.862
    },
    {
      "epoch": 1.5185746777862017,
      "grad_norm": 0.43657007813453674,
      "learning_rate": 4.840506329113924e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 2812152,
      "step": 1503,
      "train_runtime": 3184.8541,
      "train_tokens_per_second": 882.977
    },
    {
      "epoch": 1.5195855446044984,
      "grad_norm": 0.2714919149875641,
      "learning_rate": 4.8303797468354435e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 2814074,
      "step": 1504,
      "train_runtime": 3186.6372,
      "train_tokens_per_second": 883.086
    },
    {
      "epoch": 1.520596411422795,
      "grad_norm": 0.46852630376815796,
      "learning_rate": 4.820253164556962e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 2816038,
      "step": 1505,
      "train_runtime": 3188.4322,
      "train_tokens_per_second": 883.205
    },
    {
      "epoch": 1.5216072782410919,
      "grad_norm": 0.40275445580482483,
      "learning_rate": 4.810126582278481e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 2817524,
      "step": 1506,
      "train_runtime": 3190.0526,
      "train_tokens_per_second": 883.222
    },
    {
      "epoch": 1.5226181450593885,
      "grad_norm": 0.44762805104255676,
      "learning_rate": 4.8e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 2819094,
      "step": 1507,
      "train_runtime": 3191.6739,
      "train_tokens_per_second": 883.265
    },
    {
      "epoch": 1.5236290118776852,
      "grad_norm": 0.2706007659435272,
      "learning_rate": 4.789873417721519e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 2820686,
      "step": 1508,
      "train_runtime": 3193.3418,
      "train_tokens_per_second": 883.302
    },
    {
      "epoch": 1.5246398786959818,
      "grad_norm": 0.4310125708580017,
      "learning_rate": 4.779746835443038e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 2822080,
      "step": 1509,
      "train_runtime": 3194.9949,
      "train_tokens_per_second": 883.282
    },
    {
      "epoch": 1.5256507455142785,
      "grad_norm": 0.3630162179470062,
      "learning_rate": 4.7696202531645574e-05,
      "loss": 0.2405,
      "num_input_tokens_seen": 2823804,
      "step": 1510,
      "train_runtime": 3196.6974,
      "train_tokens_per_second": 883.35
    },
    {
      "epoch": 1.5266616123325751,
      "grad_norm": 0.34434300661087036,
      "learning_rate": 4.759493670886076e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 2825408,
      "step": 1511,
      "train_runtime": 3198.4159,
      "train_tokens_per_second": 883.377
    },
    {
      "epoch": 1.5276724791508718,
      "grad_norm": 0.48471885919570923,
      "learning_rate": 4.7493670886075955e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 2827282,
      "step": 1512,
      "train_runtime": 3200.2488,
      "train_tokens_per_second": 883.457
    },
    {
      "epoch": 1.5286833459691684,
      "grad_norm": 0.44771021604537964,
      "learning_rate": 4.739240506329114e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 2829082,
      "step": 1513,
      "train_runtime": 3201.9645,
      "train_tokens_per_second": 883.546
    },
    {
      "epoch": 1.529694212787465,
      "grad_norm": 0.3776264488697052,
      "learning_rate": 4.729113924050633e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 2831086,
      "step": 1514,
      "train_runtime": 3203.8063,
      "train_tokens_per_second": 883.663
    },
    {
      "epoch": 1.530705079605762,
      "grad_norm": 0.33710846304893494,
      "learning_rate": 4.718987341772152e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 2832904,
      "step": 1515,
      "train_runtime": 3205.507,
      "train_tokens_per_second": 883.762
    },
    {
      "epoch": 1.5317159464240586,
      "grad_norm": 0.353338360786438,
      "learning_rate": 4.708860759493671e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 2834864,
      "step": 1516,
      "train_runtime": 3207.3208,
      "train_tokens_per_second": 883.873
    },
    {
      "epoch": 1.5327268132423553,
      "grad_norm": 0.46391379833221436,
      "learning_rate": 4.6987341772151896e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 2836794,
      "step": 1517,
      "train_runtime": 3209.0364,
      "train_tokens_per_second": 884.002
    },
    {
      "epoch": 1.5337376800606521,
      "grad_norm": 0.2619151473045349,
      "learning_rate": 4.6886075949367094e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 2838746,
      "step": 1518,
      "train_runtime": 3210.8295,
      "train_tokens_per_second": 884.116
    },
    {
      "epoch": 1.5347485468789488,
      "grad_norm": 0.46926215291023254,
      "learning_rate": 4.678481012658228e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 2840782,
      "step": 1519,
      "train_runtime": 3212.67,
      "train_tokens_per_second": 884.243
    },
    {
      "epoch": 1.5357594136972454,
      "grad_norm": 0.2542581856250763,
      "learning_rate": 4.668354430379747e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 2842928,
      "step": 1520,
      "train_runtime": 3214.5995,
      "train_tokens_per_second": 884.38
    },
    {
      "epoch": 1.536770280515542,
      "grad_norm": 0.3810141086578369,
      "learning_rate": 4.658227848101266e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 2844540,
      "step": 1521,
      "train_runtime": 3216.2804,
      "train_tokens_per_second": 884.419
    },
    {
      "epoch": 1.5377811473338387,
      "grad_norm": 0.2847379446029663,
      "learning_rate": 4.648101265822785e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 2846628,
      "step": 1522,
      "train_runtime": 3218.1191,
      "train_tokens_per_second": 884.563
    },
    {
      "epoch": 1.5387920141521354,
      "grad_norm": 0.33421334624290466,
      "learning_rate": 4.637974683544304e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 2848616,
      "step": 1523,
      "train_runtime": 3219.9061,
      "train_tokens_per_second": 884.689
    },
    {
      "epoch": 1.539802880970432,
      "grad_norm": 0.45817071199417114,
      "learning_rate": 4.627848101265823e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 2850186,
      "step": 1524,
      "train_runtime": 3221.577,
      "train_tokens_per_second": 884.718
    },
    {
      "epoch": 1.5408137477887287,
      "grad_norm": 0.28081759810447693,
      "learning_rate": 4.6177215189873416e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 2851894,
      "step": 1525,
      "train_runtime": 3223.2834,
      "train_tokens_per_second": 884.779
    },
    {
      "epoch": 1.5418246146070256,
      "grad_norm": 0.3415515124797821,
      "learning_rate": 4.6075949367088614e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 2853984,
      "step": 1526,
      "train_runtime": 3225.1453,
      "train_tokens_per_second": 884.916
    },
    {
      "epoch": 1.5428354814253222,
      "grad_norm": 0.3951292634010315,
      "learning_rate": 4.59746835443038e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 2855646,
      "step": 1527,
      "train_runtime": 3226.7713,
      "train_tokens_per_second": 884.986
    },
    {
      "epoch": 1.5438463482436189,
      "grad_norm": 0.3425312042236328,
      "learning_rate": 4.587341772151899e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 2857392,
      "step": 1528,
      "train_runtime": 3228.4822,
      "train_tokens_per_second": 885.057
    },
    {
      "epoch": 1.5448572150619158,
      "grad_norm": 0.2674088478088379,
      "learning_rate": 4.577215189873418e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 2859654,
      "step": 1529,
      "train_runtime": 3230.4704,
      "train_tokens_per_second": 885.213
    },
    {
      "epoch": 1.5458680818802124,
      "grad_norm": 0.364071786403656,
      "learning_rate": 4.567088607594937e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 2861464,
      "step": 1530,
      "train_runtime": 3232.1115,
      "train_tokens_per_second": 885.323
    },
    {
      "epoch": 1.546878948698509,
      "grad_norm": 0.39632412791252136,
      "learning_rate": 4.556962025316456e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 2863206,
      "step": 1531,
      "train_runtime": 3235.5505,
      "train_tokens_per_second": 884.921
    },
    {
      "epoch": 1.5478898155168057,
      "grad_norm": 0.4037443697452545,
      "learning_rate": 4.546835443037975e-05,
      "loss": 0.0994,
      "num_input_tokens_seen": 2865276,
      "step": 1532,
      "train_runtime": 3237.3464,
      "train_tokens_per_second": 885.069
    },
    {
      "epoch": 1.5489006823351024,
      "grad_norm": 0.26075279712677,
      "learning_rate": 4.5367088607594936e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 2867388,
      "step": 1533,
      "train_runtime": 3239.207,
      "train_tokens_per_second": 885.213
    },
    {
      "epoch": 1.549911549153399,
      "grad_norm": 0.37886476516723633,
      "learning_rate": 4.526582278481013e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 2869110,
      "step": 1534,
      "train_runtime": 3240.9013,
      "train_tokens_per_second": 885.281
    },
    {
      "epoch": 1.5509224159716957,
      "grad_norm": 0.3209759294986725,
      "learning_rate": 4.516455696202532e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 2870932,
      "step": 1535,
      "train_runtime": 3242.6459,
      "train_tokens_per_second": 885.367
    },
    {
      "epoch": 1.5519332827899923,
      "grad_norm": 0.27610719203948975,
      "learning_rate": 4.506329113924051e-05,
      "loss": 0.1039,
      "num_input_tokens_seen": 2872076,
      "step": 1536,
      "train_runtime": 3244.1883,
      "train_tokens_per_second": 885.299
    },
    {
      "epoch": 1.552944149608289,
      "grad_norm": 0.38578173518180847,
      "learning_rate": 4.49620253164557e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 2874312,
      "step": 1537,
      "train_runtime": 3246.0648,
      "train_tokens_per_second": 885.476
    },
    {
      "epoch": 1.5539550164265858,
      "grad_norm": 0.1943003535270691,
      "learning_rate": 4.486075949367089e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 2876204,
      "step": 1538,
      "train_runtime": 3247.7282,
      "train_tokens_per_second": 885.605
    },
    {
      "epoch": 1.5549658832448825,
      "grad_norm": 0.32248440384864807,
      "learning_rate": 4.4759493670886074e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 2878178,
      "step": 1539,
      "train_runtime": 3249.456,
      "train_tokens_per_second": 885.741
    },
    {
      "epoch": 1.5559767500631791,
      "grad_norm": 0.30921298265457153,
      "learning_rate": 4.465822784810127e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 2880524,
      "step": 1540,
      "train_runtime": 3251.5319,
      "train_tokens_per_second": 885.898
    },
    {
      "epoch": 1.556987616881476,
      "grad_norm": 0.3740369975566864,
      "learning_rate": 4.4556962025316456e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 2882240,
      "step": 1541,
      "train_runtime": 3253.2256,
      "train_tokens_per_second": 885.964
    },
    {
      "epoch": 1.5579984836997727,
      "grad_norm": 0.4145367741584778,
      "learning_rate": 4.445569620253165e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 2883906,
      "step": 1542,
      "train_runtime": 3254.927,
      "train_tokens_per_second": 886.012
    },
    {
      "epoch": 1.5590093505180693,
      "grad_norm": 0.30224737524986267,
      "learning_rate": 4.435443037974684e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 2885680,
      "step": 1543,
      "train_runtime": 3256.6356,
      "train_tokens_per_second": 886.092
    },
    {
      "epoch": 1.560020217336366,
      "grad_norm": 0.3344022333621979,
      "learning_rate": 4.425316455696203e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 2887422,
      "step": 1544,
      "train_runtime": 3258.2614,
      "train_tokens_per_second": 886.185
    },
    {
      "epoch": 1.5610310841546626,
      "grad_norm": 0.3814561069011688,
      "learning_rate": 4.415189873417722e-05,
      "loss": 0.1226,
      "num_input_tokens_seen": 2889730,
      "step": 1545,
      "train_runtime": 3260.1931,
      "train_tokens_per_second": 886.368
    },
    {
      "epoch": 1.5620419509729593,
      "grad_norm": 0.37799033522605896,
      "learning_rate": 4.405063291139241e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 2891488,
      "step": 1546,
      "train_runtime": 3261.8884,
      "train_tokens_per_second": 886.446
    },
    {
      "epoch": 1.563052817791256,
      "grad_norm": 0.3428725004196167,
      "learning_rate": 4.3949367088607594e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 2893978,
      "step": 1547,
      "train_runtime": 3263.9428,
      "train_tokens_per_second": 886.651
    },
    {
      "epoch": 1.5640636846095526,
      "grad_norm": 0.3052919805049896,
      "learning_rate": 4.384810126582279e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 2895564,
      "step": 1548,
      "train_runtime": 3265.6051,
      "train_tokens_per_second": 886.685
    },
    {
      "epoch": 1.5650745514278492,
      "grad_norm": 0.30939236283302307,
      "learning_rate": 4.3746835443037976e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 2897434,
      "step": 1549,
      "train_runtime": 3267.3342,
      "train_tokens_per_second": 886.788
    },
    {
      "epoch": 1.566085418246146,
      "grad_norm": 0.3609532117843628,
      "learning_rate": 4.364556962025317e-05,
      "loss": 0.1215,
      "num_input_tokens_seen": 2899540,
      "step": 1550,
      "train_runtime": 3269.2306,
      "train_tokens_per_second": 886.918
    },
    {
      "epoch": 1.5670962850644428,
      "grad_norm": 0.3999773859977722,
      "learning_rate": 4.354430379746836e-05,
      "loss": 0.0909,
      "num_input_tokens_seen": 2901558,
      "step": 1551,
      "train_runtime": 3271.0838,
      "train_tokens_per_second": 887.033
    },
    {
      "epoch": 1.5681071518827394,
      "grad_norm": 0.3055749833583832,
      "learning_rate": 4.344303797468355e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 2903400,
      "step": 1552,
      "train_runtime": 3272.7857,
      "train_tokens_per_second": 887.134
    },
    {
      "epoch": 1.5691180187010363,
      "grad_norm": 0.3198142349720001,
      "learning_rate": 4.334177215189873e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 2904874,
      "step": 1553,
      "train_runtime": 3274.3998,
      "train_tokens_per_second": 887.147
    },
    {
      "epoch": 1.570128885519333,
      "grad_norm": 0.38498926162719727,
      "learning_rate": 4.324050632911393e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 2906934,
      "step": 1554,
      "train_runtime": 3276.2922,
      "train_tokens_per_second": 887.263
    },
    {
      "epoch": 1.5711397523376296,
      "grad_norm": 0.4477058947086334,
      "learning_rate": 4.3139240506329114e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 2908718,
      "step": 1555,
      "train_runtime": 3278.0016,
      "train_tokens_per_second": 887.345
    },
    {
      "epoch": 1.5721506191559262,
      "grad_norm": 0.4519939124584198,
      "learning_rate": 4.3037974683544305e-05,
      "loss": 0.1103,
      "num_input_tokens_seen": 2910224,
      "step": 1556,
      "train_runtime": 3279.6009,
      "train_tokens_per_second": 887.371
    },
    {
      "epoch": 1.573161485974223,
      "grad_norm": 0.2098495066165924,
      "learning_rate": 4.2936708860759496e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 2911796,
      "step": 1557,
      "train_runtime": 3281.301,
      "train_tokens_per_second": 887.391
    },
    {
      "epoch": 1.5741723527925195,
      "grad_norm": 0.36339613795280457,
      "learning_rate": 4.283544303797469e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 2914342,
      "step": 1558,
      "train_runtime": 3283.508,
      "train_tokens_per_second": 887.57
    },
    {
      "epoch": 1.5751832196108162,
      "grad_norm": 0.4097069203853607,
      "learning_rate": 4.273417721518988e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 2915786,
      "step": 1559,
      "train_runtime": 3285.0914,
      "train_tokens_per_second": 887.581
    },
    {
      "epoch": 1.5761940864291129,
      "grad_norm": 0.26405400037765503,
      "learning_rate": 4.263291139240507e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 2917450,
      "step": 1560,
      "train_runtime": 3286.7227,
      "train_tokens_per_second": 887.647
    },
    {
      "epoch": 1.5772049532474095,
      "grad_norm": 0.4004761874675751,
      "learning_rate": 4.253164556962025e-05,
      "loss": 0.1261,
      "num_input_tokens_seen": 2919334,
      "step": 1561,
      "train_runtime": 3290.5112,
      "train_tokens_per_second": 887.198
    },
    {
      "epoch": 1.5782158200657064,
      "grad_norm": 0.367236465215683,
      "learning_rate": 4.243037974683545e-05,
      "loss": 0.0542,
      "num_input_tokens_seen": 2920826,
      "step": 1562,
      "train_runtime": 3292.1237,
      "train_tokens_per_second": 887.216
    },
    {
      "epoch": 1.579226686884003,
      "grad_norm": 0.37024685740470886,
      "learning_rate": 4.2329113924050634e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 2923084,
      "step": 1563,
      "train_runtime": 3294.0702,
      "train_tokens_per_second": 887.378
    },
    {
      "epoch": 1.5802375537022997,
      "grad_norm": 0.43871355056762695,
      "learning_rate": 4.2227848101265825e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 2925074,
      "step": 1564,
      "train_runtime": 3295.8527,
      "train_tokens_per_second": 887.501
    },
    {
      "epoch": 1.5812484205205966,
      "grad_norm": 0.3516930043697357,
      "learning_rate": 4.2126582278481016e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 2926876,
      "step": 1565,
      "train_runtime": 3297.569,
      "train_tokens_per_second": 887.586
    },
    {
      "epoch": 1.5822592873388932,
      "grad_norm": 0.34721431136131287,
      "learning_rate": 4.202531645569621e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 2928626,
      "step": 1566,
      "train_runtime": 3299.2715,
      "train_tokens_per_second": 887.658
    },
    {
      "epoch": 1.5832701541571899,
      "grad_norm": 0.46014919877052307,
      "learning_rate": 4.192405063291139e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 2930188,
      "step": 1567,
      "train_runtime": 3300.9259,
      "train_tokens_per_second": 887.687
    },
    {
      "epoch": 1.5842810209754865,
      "grad_norm": 0.2739830017089844,
      "learning_rate": 4.182278481012659e-05,
      "loss": 0.0173,
      "num_input_tokens_seen": 2932382,
      "step": 1568,
      "train_runtime": 3302.8715,
      "train_tokens_per_second": 887.828
    },
    {
      "epoch": 1.5852918877937832,
      "grad_norm": 0.28977423906326294,
      "learning_rate": 4.172151898734177e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 2934348,
      "step": 1569,
      "train_runtime": 3304.7439,
      "train_tokens_per_second": 887.92
    },
    {
      "epoch": 1.5863027546120798,
      "grad_norm": 0.29666587710380554,
      "learning_rate": 4.162025316455696e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 2935992,
      "step": 1570,
      "train_runtime": 3306.3523,
      "train_tokens_per_second": 887.985
    },
    {
      "epoch": 1.5873136214303765,
      "grad_norm": 0.24439549446105957,
      "learning_rate": 4.1518987341772154e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 2937884,
      "step": 1571,
      "train_runtime": 3308.1229,
      "train_tokens_per_second": 888.082
    },
    {
      "epoch": 1.5883244882486731,
      "grad_norm": 0.2889177203178406,
      "learning_rate": 4.1417721518987345e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 2939926,
      "step": 1572,
      "train_runtime": 3309.9791,
      "train_tokens_per_second": 888.201
    },
    {
      "epoch": 1.5893353550669698,
      "grad_norm": 0.301922082901001,
      "learning_rate": 4.1316455696202536e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 2941746,
      "step": 1573,
      "train_runtime": 3311.6266,
      "train_tokens_per_second": 888.308
    },
    {
      "epoch": 1.5903462218852666,
      "grad_norm": 0.33724361658096313,
      "learning_rate": 4.121518987341773e-05,
      "loss": 0.0364,
      "num_input_tokens_seen": 2943372,
      "step": 1574,
      "train_runtime": 3313.3024,
      "train_tokens_per_second": 888.35
    },
    {
      "epoch": 1.5913570887035633,
      "grad_norm": 0.32761749625205994,
      "learning_rate": 4.111392405063291e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 2945414,
      "step": 1575,
      "train_runtime": 3315.0984,
      "train_tokens_per_second": 888.485
    },
    {
      "epoch": 1.59236795552186,
      "grad_norm": 0.3484243154525757,
      "learning_rate": 4.101265822784811e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 2947230,
      "step": 1576,
      "train_runtime": 3316.7599,
      "train_tokens_per_second": 888.587
    },
    {
      "epoch": 1.5933788223401568,
      "grad_norm": 0.239505335688591,
      "learning_rate": 4.091139240506329e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 2949028,
      "step": 1577,
      "train_runtime": 3318.4555,
      "train_tokens_per_second": 888.675
    },
    {
      "epoch": 1.5943896891584535,
      "grad_norm": 0.6944891214370728,
      "learning_rate": 4.081012658227848e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 2950658,
      "step": 1578,
      "train_runtime": 3320.0662,
      "train_tokens_per_second": 888.735
    },
    {
      "epoch": 1.5954005559767501,
      "grad_norm": 0.3439199924468994,
      "learning_rate": 4.0708860759493674e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 2952280,
      "step": 1579,
      "train_runtime": 3321.745,
      "train_tokens_per_second": 888.774
    },
    {
      "epoch": 1.5964114227950468,
      "grad_norm": 0.36311614513397217,
      "learning_rate": 4.0607594936708865e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 2953958,
      "step": 1580,
      "train_runtime": 3323.4213,
      "train_tokens_per_second": 888.83
    },
    {
      "epoch": 1.5974222896133434,
      "grad_norm": 0.35376057028770447,
      "learning_rate": 4.050632911392405e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 2955546,
      "step": 1581,
      "train_runtime": 3325.052,
      "train_tokens_per_second": 888.872
    },
    {
      "epoch": 1.59843315643164,
      "grad_norm": 0.3227643668651581,
      "learning_rate": 4.040506329113925e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 2957638,
      "step": 1582,
      "train_runtime": 3326.9713,
      "train_tokens_per_second": 888.988
    },
    {
      "epoch": 1.5994440232499367,
      "grad_norm": 0.26405656337738037,
      "learning_rate": 4.030379746835443e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 2959866,
      "step": 1583,
      "train_runtime": 3328.8558,
      "train_tokens_per_second": 889.154
    },
    {
      "epoch": 1.6004548900682334,
      "grad_norm": 0.40279439091682434,
      "learning_rate": 4.020253164556962e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 2961808,
      "step": 1584,
      "train_runtime": 3330.637,
      "train_tokens_per_second": 889.262
    },
    {
      "epoch": 1.6004548900682334,
      "eval_loss": 0.20690308511257172,
      "eval_runtime": 61.4454,
      "eval_samples_per_second": 14.305,
      "eval_steps_per_second": 7.161,
      "num_input_tokens_seen": 2961808,
      "step": 1584
    },
    {
      "epoch": 1.60146575688653,
      "grad_norm": 0.21898916363716125,
      "learning_rate": 4.010126582278481e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 2963660,
      "step": 1585,
      "train_runtime": 3393.812,
      "train_tokens_per_second": 873.254
    },
    {
      "epoch": 1.602476623704827,
      "grad_norm": 0.36059755086898804,
      "learning_rate": 4e-05,
      "loss": 0.0928,
      "num_input_tokens_seen": 2965198,
      "step": 1586,
      "train_runtime": 3395.4389,
      "train_tokens_per_second": 873.289
    },
    {
      "epoch": 1.6034874905231236,
      "grad_norm": 0.2097867876291275,
      "learning_rate": 3.9898734177215194e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 2966894,
      "step": 1587,
      "train_runtime": 3397.1445,
      "train_tokens_per_second": 873.349
    },
    {
      "epoch": 1.6044983573414204,
      "grad_norm": 0.433549165725708,
      "learning_rate": 3.9797468354430385e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 2968904,
      "step": 1588,
      "train_runtime": 3398.9986,
      "train_tokens_per_second": 873.464
    },
    {
      "epoch": 1.605509224159717,
      "grad_norm": 0.24861283600330353,
      "learning_rate": 3.969620253164557e-05,
      "loss": 0.0224,
      "num_input_tokens_seen": 2971010,
      "step": 1589,
      "train_runtime": 3400.846,
      "train_tokens_per_second": 873.609
    },
    {
      "epoch": 1.6065200909780137,
      "grad_norm": 0.3574238717556,
      "learning_rate": 3.959493670886077e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 2972712,
      "step": 1590,
      "train_runtime": 3402.5278,
      "train_tokens_per_second": 873.677
    },
    {
      "epoch": 1.6075309577963104,
      "grad_norm": 0.37372520565986633,
      "learning_rate": 3.949367088607595e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 2974944,
      "step": 1591,
      "train_runtime": 3406.6108,
      "train_tokens_per_second": 873.286
    },
    {
      "epoch": 1.608541824614607,
      "grad_norm": 0.22096741199493408,
      "learning_rate": 3.939240506329114e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 2976462,
      "step": 1592,
      "train_runtime": 3408.2282,
      "train_tokens_per_second": 873.317
    },
    {
      "epoch": 1.6095526914329037,
      "grad_norm": 0.5310118794441223,
      "learning_rate": 3.929113924050633e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 2977774,
      "step": 1593,
      "train_runtime": 3409.8306,
      "train_tokens_per_second": 873.291
    },
    {
      "epoch": 1.6105635582512003,
      "grad_norm": 0.5043955445289612,
      "learning_rate": 3.918987341772152e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 2979276,
      "step": 1594,
      "train_runtime": 3411.4498,
      "train_tokens_per_second": 873.317
    },
    {
      "epoch": 1.611574425069497,
      "grad_norm": 0.33924058079719543,
      "learning_rate": 3.908860759493671e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 2980924,
      "step": 1595,
      "train_runtime": 3413.1497,
      "train_tokens_per_second": 873.365
    },
    {
      "epoch": 1.6125852918877936,
      "grad_norm": 0.41132983565330505,
      "learning_rate": 3.8987341772151905e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 2982850,
      "step": 1596,
      "train_runtime": 3414.8769,
      "train_tokens_per_second": 873.487
    },
    {
      "epoch": 1.6135961587060905,
      "grad_norm": 0.23638512194156647,
      "learning_rate": 3.888607594936709e-05,
      "loss": 0.098,
      "num_input_tokens_seen": 2985090,
      "step": 1597,
      "train_runtime": 3416.8191,
      "train_tokens_per_second": 873.646
    },
    {
      "epoch": 1.6146070255243872,
      "grad_norm": 0.2883054316043854,
      "learning_rate": 3.878481012658228e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 2986740,
      "step": 1598,
      "train_runtime": 3418.4258,
      "train_tokens_per_second": 873.718
    },
    {
      "epoch": 1.6156178923426838,
      "grad_norm": 0.5699235796928406,
      "learning_rate": 3.868354430379747e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 2988010,
      "step": 1599,
      "train_runtime": 3420.015,
      "train_tokens_per_second": 873.683
    },
    {
      "epoch": 1.6166287591609807,
      "grad_norm": 0.3784661293029785,
      "learning_rate": 3.858227848101266e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 2989514,
      "step": 1600,
      "train_runtime": 3421.6311,
      "train_tokens_per_second": 873.71
    },
    {
      "epoch": 1.6176396259792774,
      "grad_norm": 0.5482260584831238,
      "learning_rate": 3.848101265822785e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 2991438,
      "step": 1601,
      "train_runtime": 3423.4669,
      "train_tokens_per_second": 873.804
    },
    {
      "epoch": 1.618650492797574,
      "grad_norm": 0.2891421318054199,
      "learning_rate": 3.837974683544304e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 2993592,
      "step": 1602,
      "train_runtime": 3425.4128,
      "train_tokens_per_second": 873.936
    },
    {
      "epoch": 1.6196613596158707,
      "grad_norm": 0.3944542109966278,
      "learning_rate": 3.827848101265823e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 2995730,
      "step": 1603,
      "train_runtime": 3427.2325,
      "train_tokens_per_second": 874.096
    },
    {
      "epoch": 1.6206722264341673,
      "grad_norm": 0.25274398922920227,
      "learning_rate": 3.8177215189873425e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 2997638,
      "step": 1604,
      "train_runtime": 3429.0067,
      "train_tokens_per_second": 874.2
    },
    {
      "epoch": 1.621683093252464,
      "grad_norm": 0.2859501838684082,
      "learning_rate": 3.807594936708861e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 2999940,
      "step": 1605,
      "train_runtime": 3430.9704,
      "train_tokens_per_second": 874.371
    },
    {
      "epoch": 1.6226939600707606,
      "grad_norm": 0.3490632474422455,
      "learning_rate": 3.79746835443038e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 3001676,
      "step": 1606,
      "train_runtime": 3432.7221,
      "train_tokens_per_second": 874.43
    },
    {
      "epoch": 1.6237048268890573,
      "grad_norm": 0.22489693760871887,
      "learning_rate": 3.787341772151899e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 3003440,
      "step": 1607,
      "train_runtime": 3434.3838,
      "train_tokens_per_second": 874.521
    },
    {
      "epoch": 1.624715693707354,
      "grad_norm": 0.33829188346862793,
      "learning_rate": 3.777215189873418e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 3005078,
      "step": 1608,
      "train_runtime": 3436.1076,
      "train_tokens_per_second": 874.559
    },
    {
      "epoch": 1.6257265605256508,
      "grad_norm": 0.3595799505710602,
      "learning_rate": 3.7670886075949366e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 3007042,
      "step": 1609,
      "train_runtime": 3437.8351,
      "train_tokens_per_second": 874.691
    },
    {
      "epoch": 1.6267374273439474,
      "grad_norm": 0.40967512130737305,
      "learning_rate": 3.756962025316456e-05,
      "loss": 0.0536,
      "num_input_tokens_seen": 3008866,
      "step": 1610,
      "train_runtime": 3439.5656,
      "train_tokens_per_second": 874.781
    },
    {
      "epoch": 1.627748294162244,
      "grad_norm": 0.20662987232208252,
      "learning_rate": 3.746835443037975e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 3010694,
      "step": 1611,
      "train_runtime": 3441.2932,
      "train_tokens_per_second": 874.873
    },
    {
      "epoch": 1.628759160980541,
      "grad_norm": 0.47322145104408264,
      "learning_rate": 3.736708860759494e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 3012564,
      "step": 1612,
      "train_runtime": 3443.0161,
      "train_tokens_per_second": 874.978
    },
    {
      "epoch": 1.6297700277988376,
      "grad_norm": 0.4481852352619171,
      "learning_rate": 3.726582278481013e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 3014556,
      "step": 1613,
      "train_runtime": 3444.8321,
      "train_tokens_per_second": 875.095
    },
    {
      "epoch": 1.6307808946171343,
      "grad_norm": 0.29072868824005127,
      "learning_rate": 3.716455696202532e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 3016728,
      "step": 1614,
      "train_runtime": 3446.7291,
      "train_tokens_per_second": 875.244
    },
    {
      "epoch": 1.631791761435431,
      "grad_norm": 0.43090128898620605,
      "learning_rate": 3.706329113924051e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 3018388,
      "step": 1615,
      "train_runtime": 3448.4321,
      "train_tokens_per_second": 875.293
    },
    {
      "epoch": 1.6328026282537276,
      "grad_norm": 0.2544103264808655,
      "learning_rate": 3.69620253164557e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 3020116,
      "step": 1616,
      "train_runtime": 3450.198,
      "train_tokens_per_second": 875.346
    },
    {
      "epoch": 1.6338134950720242,
      "grad_norm": 0.36950311064720154,
      "learning_rate": 3.6860759493670886e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 3022582,
      "step": 1617,
      "train_runtime": 3452.2543,
      "train_tokens_per_second": 875.539
    },
    {
      "epoch": 1.6348243618903209,
      "grad_norm": 0.5384958982467651,
      "learning_rate": 3.675949367088608e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 3024602,
      "step": 1618,
      "train_runtime": 3454.0514,
      "train_tokens_per_second": 875.668
    },
    {
      "epoch": 1.6358352287086175,
      "grad_norm": 0.3118353486061096,
      "learning_rate": 3.665822784810127e-05,
      "loss": 0.0377,
      "num_input_tokens_seen": 3026142,
      "step": 1619,
      "train_runtime": 3455.7186,
      "train_tokens_per_second": 875.691
    },
    {
      "epoch": 1.6368460955269142,
      "grad_norm": 0.2323567271232605,
      "learning_rate": 3.655696202531646e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 3028408,
      "step": 1620,
      "train_runtime": 3457.7347,
      "train_tokens_per_second": 875.836
    },
    {
      "epoch": 1.637856962345211,
      "grad_norm": 0.2999313175678253,
      "learning_rate": 3.645569620253165e-05,
      "loss": 0.0919,
      "num_input_tokens_seen": 3029928,
      "step": 1621,
      "train_runtime": 3461.5257,
      "train_tokens_per_second": 875.316
    },
    {
      "epoch": 1.6388678291635077,
      "grad_norm": 0.2249026745557785,
      "learning_rate": 3.635443037974683e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 3031656,
      "step": 1622,
      "train_runtime": 3463.177,
      "train_tokens_per_second": 875.397
    },
    {
      "epoch": 1.6398786959818044,
      "grad_norm": 0.3893710672855377,
      "learning_rate": 3.6253164556962024e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 3033868,
      "step": 1623,
      "train_runtime": 3465.1688,
      "train_tokens_per_second": 875.533
    },
    {
      "epoch": 1.6408895628001012,
      "grad_norm": 0.2950393855571747,
      "learning_rate": 3.6151898734177215e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 3035604,
      "step": 1624,
      "train_runtime": 3466.8271,
      "train_tokens_per_second": 875.614
    },
    {
      "epoch": 1.6419004296183979,
      "grad_norm": 0.3039438724517822,
      "learning_rate": 3.6050632911392405e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 3037318,
      "step": 1625,
      "train_runtime": 3468.6247,
      "train_tokens_per_second": 875.655
    },
    {
      "epoch": 1.6429112964366945,
      "grad_norm": 0.31106412410736084,
      "learning_rate": 3.5949367088607596e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 3038774,
      "step": 1626,
      "train_runtime": 3470.2353,
      "train_tokens_per_second": 875.668
    },
    {
      "epoch": 1.6439221632549912,
      "grad_norm": 0.31181344389915466,
      "learning_rate": 3.584810126582279e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 3041112,
      "step": 1627,
      "train_runtime": 3472.2348,
      "train_tokens_per_second": 875.837
    },
    {
      "epoch": 1.6449330300732878,
      "grad_norm": 0.245428204536438,
      "learning_rate": 3.574683544303797e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 3042706,
      "step": 1628,
      "train_runtime": 3473.8686,
      "train_tokens_per_second": 875.884
    },
    {
      "epoch": 1.6459438968915845,
      "grad_norm": 0.5006813406944275,
      "learning_rate": 3.564556962025317e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 3044244,
      "step": 1629,
      "train_runtime": 3475.5517,
      "train_tokens_per_second": 875.902
    },
    {
      "epoch": 1.6469547637098811,
      "grad_norm": 0.5197868943214417,
      "learning_rate": 3.554430379746835e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 3045864,
      "step": 1630,
      "train_runtime": 3477.1808,
      "train_tokens_per_second": 875.958
    },
    {
      "epoch": 1.6479656305281778,
      "grad_norm": 0.472901314496994,
      "learning_rate": 3.5443037974683544e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 3047418,
      "step": 1631,
      "train_runtime": 3478.8007,
      "train_tokens_per_second": 875.997
    },
    {
      "epoch": 1.6489764973464744,
      "grad_norm": 0.286668062210083,
      "learning_rate": 3.5341772151898735e-05,
      "loss": 0.1391,
      "num_input_tokens_seen": 3049356,
      "step": 1632,
      "train_runtime": 3480.5941,
      "train_tokens_per_second": 876.102
    },
    {
      "epoch": 1.6499873641647713,
      "grad_norm": 0.37653443217277527,
      "learning_rate": 3.5240506329113925e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 3051262,
      "step": 1633,
      "train_runtime": 3482.4096,
      "train_tokens_per_second": 876.193
    },
    {
      "epoch": 1.650998230983068,
      "grad_norm": 0.29574844241142273,
      "learning_rate": 3.5139240506329116e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 3053406,
      "step": 1634,
      "train_runtime": 3484.3254,
      "train_tokens_per_second": 876.326
    },
    {
      "epoch": 1.6520090978013646,
      "grad_norm": 0.3396419584751129,
      "learning_rate": 3.503797468354431e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 3055568,
      "step": 1635,
      "train_runtime": 3486.2441,
      "train_tokens_per_second": 876.464
    },
    {
      "epoch": 1.6530199646196615,
      "grad_norm": 0.2615917921066284,
      "learning_rate": 3.493670886075949e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 3057390,
      "step": 1636,
      "train_runtime": 3488.0114,
      "train_tokens_per_second": 876.542
    },
    {
      "epoch": 1.6540308314379581,
      "grad_norm": 0.29763129353523254,
      "learning_rate": 3.483544303797469e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 3059462,
      "step": 1637,
      "train_runtime": 3489.8733,
      "train_tokens_per_second": 876.668
    },
    {
      "epoch": 1.6550416982562548,
      "grad_norm": 0.34447604417800903,
      "learning_rate": 3.473417721518987e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 3061326,
      "step": 1638,
      "train_runtime": 3491.6033,
      "train_tokens_per_second": 876.768
    },
    {
      "epoch": 1.6560525650745515,
      "grad_norm": 0.3206090033054352,
      "learning_rate": 3.4632911392405064e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 3062970,
      "step": 1639,
      "train_runtime": 3493.2486,
      "train_tokens_per_second": 876.826
    },
    {
      "epoch": 1.657063431892848,
      "grad_norm": 0.30783480405807495,
      "learning_rate": 3.4531645569620255e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 3064720,
      "step": 1640,
      "train_runtime": 3494.9723,
      "train_tokens_per_second": 876.894
    },
    {
      "epoch": 1.6580742987111448,
      "grad_norm": 0.3695422112941742,
      "learning_rate": 3.4430379746835445e-05,
      "loss": 0.1878,
      "num_input_tokens_seen": 3066876,
      "step": 1641,
      "train_runtime": 3496.7743,
      "train_tokens_per_second": 877.059
    },
    {
      "epoch": 1.6590851655294414,
      "grad_norm": 0.29870784282684326,
      "learning_rate": 3.432911392405063e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 3068808,
      "step": 1642,
      "train_runtime": 3498.5693,
      "train_tokens_per_second": 877.161
    },
    {
      "epoch": 1.660096032347738,
      "grad_norm": 0.22628048062324524,
      "learning_rate": 3.422784810126583e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 3070700,
      "step": 1643,
      "train_runtime": 3500.2827,
      "train_tokens_per_second": 877.272
    },
    {
      "epoch": 1.6611068991660347,
      "grad_norm": 0.31168633699417114,
      "learning_rate": 3.412658227848101e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 3072532,
      "step": 1644,
      "train_runtime": 3502.0445,
      "train_tokens_per_second": 877.354
    },
    {
      "epoch": 1.6621177659843316,
      "grad_norm": 0.381340891122818,
      "learning_rate": 3.40253164556962e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 3074744,
      "step": 1645,
      "train_runtime": 3503.9945,
      "train_tokens_per_second": 877.497
    },
    {
      "epoch": 1.6631286328026282,
      "grad_norm": 0.39735448360443115,
      "learning_rate": 3.392405063291139e-05,
      "loss": 0.1051,
      "num_input_tokens_seen": 3076964,
      "step": 1646,
      "train_runtime": 3505.912,
      "train_tokens_per_second": 877.65
    },
    {
      "epoch": 1.664139499620925,
      "grad_norm": 0.3131244480609894,
      "learning_rate": 3.3822784810126584e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 3078936,
      "step": 1647,
      "train_runtime": 3507.6954,
      "train_tokens_per_second": 877.766
    },
    {
      "epoch": 1.6651503664392218,
      "grad_norm": 0.3398909866809845,
      "learning_rate": 3.3721518987341775e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 3080974,
      "step": 1648,
      "train_runtime": 3509.5487,
      "train_tokens_per_second": 877.883
    },
    {
      "epoch": 1.6661612332575184,
      "grad_norm": 0.31063729524612427,
      "learning_rate": 3.3620253164556965e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 3082698,
      "step": 1649,
      "train_runtime": 3511.2723,
      "train_tokens_per_second": 877.943
    },
    {
      "epoch": 1.667172100075815,
      "grad_norm": 0.29326948523521423,
      "learning_rate": 3.351898734177215e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 3084766,
      "step": 1650,
      "train_runtime": 3513.0733,
      "train_tokens_per_second": 878.082
    },
    {
      "epoch": 1.6681829668941117,
      "grad_norm": 0.3226855993270874,
      "learning_rate": 3.341772151898735e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 3086578,
      "step": 1651,
      "train_runtime": 3516.59,
      "train_tokens_per_second": 877.719
    },
    {
      "epoch": 1.6691938337124084,
      "grad_norm": 0.429928183555603,
      "learning_rate": 3.331645569620253e-05,
      "loss": 0.0788,
      "num_input_tokens_seen": 3088930,
      "step": 1652,
      "train_runtime": 3518.6015,
      "train_tokens_per_second": 877.886
    },
    {
      "epoch": 1.670204700530705,
      "grad_norm": 0.2601509392261505,
      "learning_rate": 3.321518987341772e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 3091036,
      "step": 1653,
      "train_runtime": 3520.4146,
      "train_tokens_per_second": 878.032
    },
    {
      "epoch": 1.6712155673490017,
      "grad_norm": 0.22332744300365448,
      "learning_rate": 3.311392405063291e-05,
      "loss": 0.0211,
      "num_input_tokens_seen": 3093124,
      "step": 1654,
      "train_runtime": 3522.3461,
      "train_tokens_per_second": 878.143
    },
    {
      "epoch": 1.6722264341672983,
      "grad_norm": 0.455143004655838,
      "learning_rate": 3.3012658227848104e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 3094950,
      "step": 1655,
      "train_runtime": 3524.0557,
      "train_tokens_per_second": 878.235
    },
    {
      "epoch": 1.6732373009855952,
      "grad_norm": 0.2553136646747589,
      "learning_rate": 3.291139240506329e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 3096976,
      "step": 1656,
      "train_runtime": 3525.8481,
      "train_tokens_per_second": 878.363
    },
    {
      "epoch": 1.6742481678038919,
      "grad_norm": 0.29013291001319885,
      "learning_rate": 3.2810126582278485e-05,
      "loss": 0.0216,
      "num_input_tokens_seen": 3098344,
      "step": 1657,
      "train_runtime": 3527.4394,
      "train_tokens_per_second": 878.355
    },
    {
      "epoch": 1.6752590346221885,
      "grad_norm": 0.2900941073894501,
      "learning_rate": 3.270886075949367e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 3099804,
      "step": 1658,
      "train_runtime": 3529.0384,
      "train_tokens_per_second": 878.371
    },
    {
      "epoch": 1.6762699014404854,
      "grad_norm": 0.35698002576828003,
      "learning_rate": 3.260759493670886e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 3101562,
      "step": 1659,
      "train_runtime": 3530.7335,
      "train_tokens_per_second": 878.447
    },
    {
      "epoch": 1.677280768258782,
      "grad_norm": 0.3415001630783081,
      "learning_rate": 3.250632911392405e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 3103098,
      "step": 1660,
      "train_runtime": 3532.3562,
      "train_tokens_per_second": 878.478
    },
    {
      "epoch": 1.6782916350770787,
      "grad_norm": 0.22493651509284973,
      "learning_rate": 3.240506329113924e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 3104696,
      "step": 1661,
      "train_runtime": 3534.0299,
      "train_tokens_per_second": 878.514
    },
    {
      "epoch": 1.6793025018953753,
      "grad_norm": 0.34866082668304443,
      "learning_rate": 3.230379746835443e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 3106364,
      "step": 1662,
      "train_runtime": 3535.7287,
      "train_tokens_per_second": 878.564
    },
    {
      "epoch": 1.680313368713672,
      "grad_norm": 0.36344191431999207,
      "learning_rate": 3.2202531645569624e-05,
      "loss": 0.1163,
      "num_input_tokens_seen": 3108372,
      "step": 1663,
      "train_runtime": 3537.4666,
      "train_tokens_per_second": 878.7
    },
    {
      "epoch": 1.6813242355319686,
      "grad_norm": 0.2658107280731201,
      "learning_rate": 3.210126582278481e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 3110926,
      "step": 1664,
      "train_runtime": 3539.5363,
      "train_tokens_per_second": 878.908
    },
    {
      "epoch": 1.6823351023502653,
      "grad_norm": 0.24368809163570404,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.1506,
      "num_input_tokens_seen": 3112924,
      "step": 1665,
      "train_runtime": 3541.3191,
      "train_tokens_per_second": 879.03
    },
    {
      "epoch": 1.683345969168562,
      "grad_norm": 0.40462726354599,
      "learning_rate": 3.189873417721519e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 3114486,
      "step": 1666,
      "train_runtime": 3542.9466,
      "train_tokens_per_second": 879.067
    },
    {
      "epoch": 1.6843568359868586,
      "grad_norm": 0.29145365953445435,
      "learning_rate": 3.179746835443038e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 3115978,
      "step": 1667,
      "train_runtime": 3544.6292,
      "train_tokens_per_second": 879.07
    },
    {
      "epoch": 1.6853677028051555,
      "grad_norm": 0.49398040771484375,
      "learning_rate": 3.169620253164557e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 3117472,
      "step": 1668,
      "train_runtime": 3546.2591,
      "train_tokens_per_second": 879.087
    },
    {
      "epoch": 1.6863785696234521,
      "grad_norm": 0.28147852420806885,
      "learning_rate": 3.159493670886076e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 3119424,
      "step": 1669,
      "train_runtime": 3548.1174,
      "train_tokens_per_second": 879.177
    },
    {
      "epoch": 1.6873894364417488,
      "grad_norm": 0.2428324669599533,
      "learning_rate": 3.1493670886075946e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 3121726,
      "step": 1670,
      "train_runtime": 3550.011,
      "train_tokens_per_second": 879.357
    },
    {
      "epoch": 1.6884003032600456,
      "grad_norm": 0.4500986337661743,
      "learning_rate": 3.1392405063291144e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 3123334,
      "step": 1671,
      "train_runtime": 3551.6968,
      "train_tokens_per_second": 879.392
    },
    {
      "epoch": 1.6894111700783423,
      "grad_norm": 0.4447426199913025,
      "learning_rate": 3.129113924050633e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 3125578,
      "step": 1672,
      "train_runtime": 3553.6451,
      "train_tokens_per_second": 879.541
    },
    {
      "epoch": 1.690422036896639,
      "grad_norm": 0.2644793689250946,
      "learning_rate": 3.118987341772152e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 3127446,
      "step": 1673,
      "train_runtime": 3555.392,
      "train_tokens_per_second": 879.635
    },
    {
      "epoch": 1.6914329037149356,
      "grad_norm": 0.2933402955532074,
      "learning_rate": 3.108860759493671e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 3129356,
      "step": 1674,
      "train_runtime": 3557.1717,
      "train_tokens_per_second": 879.731
    },
    {
      "epoch": 1.6924437705332323,
      "grad_norm": 0.33312034606933594,
      "learning_rate": 3.09873417721519e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 3131180,
      "step": 1675,
      "train_runtime": 3558.8963,
      "train_tokens_per_second": 879.818
    },
    {
      "epoch": 1.693454637351529,
      "grad_norm": 0.42872354388237,
      "learning_rate": 3.088607594936709e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 3132998,
      "step": 1676,
      "train_runtime": 3560.6798,
      "train_tokens_per_second": 879.888
    },
    {
      "epoch": 1.6944655041698256,
      "grad_norm": 0.4059085547924042,
      "learning_rate": 3.078481012658228e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 3134908,
      "step": 1677,
      "train_runtime": 3562.5173,
      "train_tokens_per_second": 879.97
    },
    {
      "epoch": 1.6954763709881222,
      "grad_norm": 0.34580790996551514,
      "learning_rate": 3.0683544303797466e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 3136666,
      "step": 1678,
      "train_runtime": 3564.2872,
      "train_tokens_per_second": 880.026
    },
    {
      "epoch": 1.6964872378064189,
      "grad_norm": 0.3064793050289154,
      "learning_rate": 3.0582278481012664e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 3138340,
      "step": 1679,
      "train_runtime": 3565.9111,
      "train_tokens_per_second": 880.095
    },
    {
      "epoch": 1.6974981046247157,
      "grad_norm": 0.32855427265167236,
      "learning_rate": 3.0481012658227848e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 3140538,
      "step": 1680,
      "train_runtime": 3567.7794,
      "train_tokens_per_second": 880.25
    },
    {
      "epoch": 1.6985089714430124,
      "grad_norm": 0.17733171582221985,
      "learning_rate": 3.0379746835443042e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 3142338,
      "step": 1681,
      "train_runtime": 3571.6203,
      "train_tokens_per_second": 879.807
    },
    {
      "epoch": 1.699519838261309,
      "grad_norm": 0.3169107735157013,
      "learning_rate": 3.0278481012658226e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 3144178,
      "step": 1682,
      "train_runtime": 3573.4073,
      "train_tokens_per_second": 879.882
    },
    {
      "epoch": 1.700530705079606,
      "grad_norm": 0.5541936159133911,
      "learning_rate": 3.017721518987342e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 3145768,
      "step": 1683,
      "train_runtime": 3575.0448,
      "train_tokens_per_second": 879.924
    },
    {
      "epoch": 1.7015415718979026,
      "grad_norm": 0.385830283164978,
      "learning_rate": 3.0075949367088608e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 3147908,
      "step": 1684,
      "train_runtime": 3576.9055,
      "train_tokens_per_second": 880.065
    },
    {
      "epoch": 1.7025524387161992,
      "grad_norm": 0.3147771954536438,
      "learning_rate": 2.99746835443038e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 3149528,
      "step": 1685,
      "train_runtime": 3578.5851,
      "train_tokens_per_second": 880.104
    },
    {
      "epoch": 1.7035633055344959,
      "grad_norm": 0.35411569476127625,
      "learning_rate": 2.9873417721518986e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 3151662,
      "step": 1686,
      "train_runtime": 3580.4425,
      "train_tokens_per_second": 880.244
    },
    {
      "epoch": 1.7045741723527925,
      "grad_norm": 0.3230752646923065,
      "learning_rate": 2.977215189873418e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 3153480,
      "step": 1687,
      "train_runtime": 3582.2102,
      "train_tokens_per_second": 880.317
    },
    {
      "epoch": 1.7055850391710892,
      "grad_norm": 0.3497144877910614,
      "learning_rate": 2.9670886075949368e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 3155566,
      "step": 1688,
      "train_runtime": 3584.0104,
      "train_tokens_per_second": 880.457
    },
    {
      "epoch": 1.7065959059893858,
      "grad_norm": 0.3007948696613312,
      "learning_rate": 2.956962025316456e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 3157276,
      "step": 1689,
      "train_runtime": 3585.713,
      "train_tokens_per_second": 880.516
    },
    {
      "epoch": 1.7076067728076825,
      "grad_norm": 0.24831940233707428,
      "learning_rate": 2.9468354430379746e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 3158966,
      "step": 1690,
      "train_runtime": 3587.4078,
      "train_tokens_per_second": 880.571
    },
    {
      "epoch": 1.7086176396259791,
      "grad_norm": 0.22481276094913483,
      "learning_rate": 2.936708860759494e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 3161430,
      "step": 1691,
      "train_runtime": 3589.4596,
      "train_tokens_per_second": 880.754
    },
    {
      "epoch": 1.709628506444276,
      "grad_norm": 0.4430962800979614,
      "learning_rate": 2.9265822784810128e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 3163380,
      "step": 1692,
      "train_runtime": 3591.215,
      "train_tokens_per_second": 880.866
    },
    {
      "epoch": 1.7106393732625726,
      "grad_norm": 0.4412935674190521,
      "learning_rate": 2.916455696202532e-05,
      "loss": 0.093,
      "num_input_tokens_seen": 3164830,
      "step": 1693,
      "train_runtime": 3592.8208,
      "train_tokens_per_second": 880.876
    },
    {
      "epoch": 1.7116502400808693,
      "grad_norm": 0.3547500967979431,
      "learning_rate": 2.9063291139240506e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 3166754,
      "step": 1694,
      "train_runtime": 3594.6097,
      "train_tokens_per_second": 880.973
    },
    {
      "epoch": 1.7126611068991662,
      "grad_norm": 0.27725890278816223,
      "learning_rate": 2.89620253164557e-05,
      "loss": 0.1469,
      "num_input_tokens_seen": 3168720,
      "step": 1695,
      "train_runtime": 3596.458,
      "train_tokens_per_second": 881.067
    },
    {
      "epoch": 1.7136719737174628,
      "grad_norm": 0.3641546368598938,
      "learning_rate": 2.8860759493670884e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 3170374,
      "step": 1696,
      "train_runtime": 3598.1465,
      "train_tokens_per_second": 881.113
    },
    {
      "epoch": 1.7146828405357595,
      "grad_norm": 0.31772106885910034,
      "learning_rate": 2.875949367088608e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 3172188,
      "step": 1697,
      "train_runtime": 3599.8613,
      "train_tokens_per_second": 881.197
    },
    {
      "epoch": 1.7156937073540561,
      "grad_norm": 0.2380431741476059,
      "learning_rate": 2.8658227848101266e-05,
      "loss": 0.1429,
      "num_input_tokens_seen": 3174010,
      "step": 1698,
      "train_runtime": 3601.633,
      "train_tokens_per_second": 881.27
    },
    {
      "epoch": 1.7167045741723528,
      "grad_norm": 0.2958558201789856,
      "learning_rate": 2.8556962025316457e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 3175792,
      "step": 1699,
      "train_runtime": 3603.3646,
      "train_tokens_per_second": 881.341
    },
    {
      "epoch": 1.7177154409906494,
      "grad_norm": 0.33768099546432495,
      "learning_rate": 2.8455696202531644e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 3177884,
      "step": 1700,
      "train_runtime": 3605.164,
      "train_tokens_per_second": 881.481
    },
    {
      "epoch": 1.718726307808946,
      "grad_norm": 0.5989857316017151,
      "learning_rate": 2.835443037974684e-05,
      "loss": 0.1356,
      "num_input_tokens_seen": 3179780,
      "step": 1701,
      "train_runtime": 3606.9538,
      "train_tokens_per_second": 881.569
    },
    {
      "epoch": 1.7197371746272427,
      "grad_norm": 0.33277639746665955,
      "learning_rate": 2.8253164556962026e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 3181572,
      "step": 1702,
      "train_runtime": 3608.6584,
      "train_tokens_per_second": 881.65
    },
    {
      "epoch": 1.7207480414455394,
      "grad_norm": 0.2743334174156189,
      "learning_rate": 2.8151898734177217e-05,
      "loss": 0.1222,
      "num_input_tokens_seen": 3183630,
      "step": 1703,
      "train_runtime": 3610.523,
      "train_tokens_per_second": 881.764
    },
    {
      "epoch": 1.7217589082638363,
      "grad_norm": 0.44329899549484253,
      "learning_rate": 2.8050632911392404e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 3185760,
      "step": 1704,
      "train_runtime": 3612.3822,
      "train_tokens_per_second": 881.9
    },
    {
      "epoch": 1.722769775082133,
      "grad_norm": 0.4088926613330841,
      "learning_rate": 2.79493670886076e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 3187358,
      "step": 1705,
      "train_runtime": 3614.0081,
      "train_tokens_per_second": 881.945
    },
    {
      "epoch": 1.7237806419004296,
      "grad_norm": 0.25346213579177856,
      "learning_rate": 2.7848101265822786e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 3189602,
      "step": 1706,
      "train_runtime": 3615.9007,
      "train_tokens_per_second": 882.104
    },
    {
      "epoch": 1.7247915087187264,
      "grad_norm": 0.3881666362285614,
      "learning_rate": 2.7746835443037977e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 3191628,
      "step": 1707,
      "train_runtime": 3617.7394,
      "train_tokens_per_second": 882.216
    },
    {
      "epoch": 1.725802375537023,
      "grad_norm": 0.19983576238155365,
      "learning_rate": 2.7645569620253164e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 3193318,
      "step": 1708,
      "train_runtime": 3619.4224,
      "train_tokens_per_second": 882.273
    },
    {
      "epoch": 1.7268132423553197,
      "grad_norm": 0.34657493233680725,
      "learning_rate": 2.754430379746836e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 3195264,
      "step": 1709,
      "train_runtime": 3621.202,
      "train_tokens_per_second": 882.377
    },
    {
      "epoch": 1.7278241091736164,
      "grad_norm": 0.36328038573265076,
      "learning_rate": 2.7443037974683546e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 3196688,
      "step": 1710,
      "train_runtime": 3622.8077,
      "train_tokens_per_second": 882.379
    },
    {
      "epoch": 1.728834975991913,
      "grad_norm": 0.2944445312023163,
      "learning_rate": 2.7341772151898737e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 3198552,
      "step": 1711,
      "train_runtime": 3626.5823,
      "train_tokens_per_second": 881.974
    },
    {
      "epoch": 1.7298458428102097,
      "grad_norm": 0.3125305771827698,
      "learning_rate": 2.7240506329113924e-05,
      "loss": 0.1438,
      "num_input_tokens_seen": 3200708,
      "step": 1712,
      "train_runtime": 3628.3926,
      "train_tokens_per_second": 882.128
    },
    {
      "epoch": 1.7308567096285064,
      "grad_norm": 0.5570347905158997,
      "learning_rate": 2.7139240506329115e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 3202646,
      "step": 1713,
      "train_runtime": 3630.2888,
      "train_tokens_per_second": 882.201
    },
    {
      "epoch": 1.731867576446803,
      "grad_norm": 0.4130195081233978,
      "learning_rate": 2.7037974683544302e-05,
      "loss": 0.1418,
      "num_input_tokens_seen": 3204522,
      "step": 1714,
      "train_runtime": 3632.0503,
      "train_tokens_per_second": 882.29
    },
    {
      "epoch": 1.7328784432650999,
      "grad_norm": 0.41316428780555725,
      "learning_rate": 2.6936708860759497e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 3206614,
      "step": 1715,
      "train_runtime": 3633.8169,
      "train_tokens_per_second": 882.437
    },
    {
      "epoch": 1.7338893100833965,
      "grad_norm": 0.2011561542749405,
      "learning_rate": 2.6835443037974684e-05,
      "loss": 0.1147,
      "num_input_tokens_seen": 3208340,
      "step": 1716,
      "train_runtime": 3635.5185,
      "train_tokens_per_second": 882.499
    },
    {
      "epoch": 1.7349001769016932,
      "grad_norm": 0.32938793301582336,
      "learning_rate": 2.6734177215189875e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 3209928,
      "step": 1717,
      "train_runtime": 3637.2106,
      "train_tokens_per_second": 882.525
    },
    {
      "epoch": 1.73591104371999,
      "grad_norm": 0.3284657299518585,
      "learning_rate": 2.6632911392405062e-05,
      "loss": 0.1219,
      "num_input_tokens_seen": 3211956,
      "step": 1718,
      "train_runtime": 3639.0588,
      "train_tokens_per_second": 882.634
    },
    {
      "epoch": 1.7369219105382867,
      "grad_norm": 0.2739722728729248,
      "learning_rate": 2.6531645569620257e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 3213430,
      "step": 1719,
      "train_runtime": 3640.678,
      "train_tokens_per_second": 882.646
    },
    {
      "epoch": 1.7379327773565834,
      "grad_norm": 0.36229974031448364,
      "learning_rate": 2.6430379746835444e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 3214796,
      "step": 1720,
      "train_runtime": 3642.278,
      "train_tokens_per_second": 882.633
    },
    {
      "epoch": 1.73894364417488,
      "grad_norm": 0.3124796152114868,
      "learning_rate": 2.6329113924050635e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 3216342,
      "step": 1721,
      "train_runtime": 3643.9185,
      "train_tokens_per_second": 882.66
    },
    {
      "epoch": 1.7399545109931767,
      "grad_norm": 0.39590394496917725,
      "learning_rate": 2.6227848101265822e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 3218220,
      "step": 1722,
      "train_runtime": 3645.689,
      "train_tokens_per_second": 882.747
    },
    {
      "epoch": 1.7409653778114733,
      "grad_norm": 0.5818724632263184,
      "learning_rate": 2.6126582278481017e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 3219820,
      "step": 1723,
      "train_runtime": 3647.3174,
      "train_tokens_per_second": 882.791
    },
    {
      "epoch": 1.74197624462977,
      "grad_norm": 0.4091465473175049,
      "learning_rate": 2.6025316455696204e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 3221266,
      "step": 1724,
      "train_runtime": 3648.9129,
      "train_tokens_per_second": 882.802
    },
    {
      "epoch": 1.7429871114480666,
      "grad_norm": 0.3920089304447174,
      "learning_rate": 2.5924050632911395e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 3222628,
      "step": 1725,
      "train_runtime": 3650.4884,
      "train_tokens_per_second": 882.794
    },
    {
      "epoch": 1.7439979782663633,
      "grad_norm": 0.2700601816177368,
      "learning_rate": 2.5822784810126582e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 3224814,
      "step": 1726,
      "train_runtime": 3652.4073,
      "train_tokens_per_second": 882.928
    },
    {
      "epoch": 1.7450088450846601,
      "grad_norm": 0.31126463413238525,
      "learning_rate": 2.5721518987341777e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 3226534,
      "step": 1727,
      "train_runtime": 3654.0474,
      "train_tokens_per_second": 883.003
    },
    {
      "epoch": 1.7460197119029568,
      "grad_norm": 0.20121997594833374,
      "learning_rate": 2.562025316455696e-05,
      "loss": 0.0697,
      "num_input_tokens_seen": 3228730,
      "step": 1728,
      "train_runtime": 3656.025,
      "train_tokens_per_second": 883.126
    },
    {
      "epoch": 1.7470305787212534,
      "grad_norm": 0.34647297859191895,
      "learning_rate": 2.5518987341772155e-05,
      "loss": 0.086,
      "num_input_tokens_seen": 3230610,
      "step": 1729,
      "train_runtime": 3657.7975,
      "train_tokens_per_second": 883.212
    },
    {
      "epoch": 1.7480414455395503,
      "grad_norm": 0.30007243156433105,
      "learning_rate": 2.5417721518987342e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 3232674,
      "step": 1730,
      "train_runtime": 3659.6546,
      "train_tokens_per_second": 883.328
    },
    {
      "epoch": 1.749052312357847,
      "grad_norm": 0.3535133898258209,
      "learning_rate": 2.5316455696202533e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 3234534,
      "step": 1731,
      "train_runtime": 3661.3835,
      "train_tokens_per_second": 883.419
    },
    {
      "epoch": 1.7500631791761436,
      "grad_norm": 0.28902703523635864,
      "learning_rate": 2.521518987341772e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 3236180,
      "step": 1732,
      "train_runtime": 3663.0059,
      "train_tokens_per_second": 883.477
    },
    {
      "epoch": 1.7510740459944403,
      "grad_norm": 0.2519863545894623,
      "learning_rate": 2.5113924050632915e-05,
      "loss": 0.1221,
      "num_input_tokens_seen": 3237998,
      "step": 1733,
      "train_runtime": 3664.7267,
      "train_tokens_per_second": 883.558
    },
    {
      "epoch": 1.752084912812737,
      "grad_norm": 0.1911328136920929,
      "learning_rate": 2.5012658227848102e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 3239724,
      "step": 1734,
      "train_runtime": 3666.3628,
      "train_tokens_per_second": 883.634
    },
    {
      "epoch": 1.7530957796310336,
      "grad_norm": 0.20987670123577118,
      "learning_rate": 2.491139240506329e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 3241578,
      "step": 1735,
      "train_runtime": 3668.13,
      "train_tokens_per_second": 883.714
    },
    {
      "epoch": 1.7541066464493302,
      "grad_norm": 0.4213102459907532,
      "learning_rate": 2.481012658227848e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 3243238,
      "step": 1736,
      "train_runtime": 3669.7713,
      "train_tokens_per_second": 883.771
    },
    {
      "epoch": 1.7551175132676269,
      "grad_norm": 0.3850823938846588,
      "learning_rate": 2.470886075949367e-05,
      "loss": 0.0777,
      "num_input_tokens_seen": 3244444,
      "step": 1737,
      "train_runtime": 3671.3175,
      "train_tokens_per_second": 883.727
    },
    {
      "epoch": 1.7561283800859235,
      "grad_norm": 0.34885308146476746,
      "learning_rate": 2.4607594936708862e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 3246040,
      "step": 1738,
      "train_runtime": 3672.9985,
      "train_tokens_per_second": 883.757
    },
    {
      "epoch": 1.7571392469042204,
      "grad_norm": 0.5272138118743896,
      "learning_rate": 2.450632911392405e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 3247988,
      "step": 1739,
      "train_runtime": 3674.7367,
      "train_tokens_per_second": 883.87
    },
    {
      "epoch": 1.758150113722517,
      "grad_norm": 0.5044370889663696,
      "learning_rate": 2.440506329113924e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 3249736,
      "step": 1740,
      "train_runtime": 3676.4452,
      "train_tokens_per_second": 883.934
    },
    {
      "epoch": 1.7591609805408137,
      "grad_norm": 0.23041538894176483,
      "learning_rate": 2.430379746835443e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 3251864,
      "step": 1741,
      "train_runtime": 3680.0506,
      "train_tokens_per_second": 883.647
    },
    {
      "epoch": 1.7601718473591106,
      "grad_norm": 0.2476688027381897,
      "learning_rate": 2.420253164556962e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 3253816,
      "step": 1742,
      "train_runtime": 3681.8431,
      "train_tokens_per_second": 883.747
    },
    {
      "epoch": 1.7611827141774072,
      "grad_norm": 0.32131579518318176,
      "learning_rate": 2.410126582278481e-05,
      "loss": 0.0918,
      "num_input_tokens_seen": 3255658,
      "step": 1743,
      "train_runtime": 3683.5575,
      "train_tokens_per_second": 883.835
    },
    {
      "epoch": 1.762193580995704,
      "grad_norm": 0.2802947163581848,
      "learning_rate": 2.4e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 3257502,
      "step": 1744,
      "train_runtime": 3685.3504,
      "train_tokens_per_second": 883.906
    },
    {
      "epoch": 1.7632044478140005,
      "grad_norm": 0.3082278370857239,
      "learning_rate": 2.389873417721519e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 3259444,
      "step": 1745,
      "train_runtime": 3687.1265,
      "train_tokens_per_second": 884.007
    },
    {
      "epoch": 1.7642153146322972,
      "grad_norm": 0.23779797554016113,
      "learning_rate": 2.379746835443038e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 3261450,
      "step": 1746,
      "train_runtime": 3688.973,
      "train_tokens_per_second": 884.108
    },
    {
      "epoch": 1.7652261814505938,
      "grad_norm": 0.37954798340797424,
      "learning_rate": 2.369620253164557e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 3263334,
      "step": 1747,
      "train_runtime": 3690.7708,
      "train_tokens_per_second": 884.188
    },
    {
      "epoch": 1.7662370482688905,
      "grad_norm": 0.3391364812850952,
      "learning_rate": 2.359493670886076e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 3265210,
      "step": 1748,
      "train_runtime": 3692.514,
      "train_tokens_per_second": 884.278
    },
    {
      "epoch": 1.7672479150871871,
      "grad_norm": 0.2723386585712433,
      "learning_rate": 2.3493670886075948e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 3267092,
      "step": 1749,
      "train_runtime": 3694.2161,
      "train_tokens_per_second": 884.38
    },
    {
      "epoch": 1.7682587819054838,
      "grad_norm": 0.21548764407634735,
      "learning_rate": 2.339240506329114e-05,
      "loss": 0.0099,
      "num_input_tokens_seen": 3269218,
      "step": 1750,
      "train_runtime": 3696.0707,
      "train_tokens_per_second": 884.512
    },
    {
      "epoch": 1.7692696487237807,
      "grad_norm": 0.3597107529640198,
      "learning_rate": 2.329113924050633e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 3270860,
      "step": 1751,
      "train_runtime": 3697.7877,
      "train_tokens_per_second": 884.545
    },
    {
      "epoch": 1.7702805155420773,
      "grad_norm": 0.3120933175086975,
      "learning_rate": 2.318987341772152e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 3272532,
      "step": 1752,
      "train_runtime": 3699.4245,
      "train_tokens_per_second": 884.606
    },
    {
      "epoch": 1.771291382360374,
      "grad_norm": 0.32579106092453003,
      "learning_rate": 2.3088607594936708e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 3273872,
      "step": 1753,
      "train_runtime": 3701.0004,
      "train_tokens_per_second": 884.591
    },
    {
      "epoch": 1.7723022491786709,
      "grad_norm": 0.3475702702999115,
      "learning_rate": 2.29873417721519e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 3275510,
      "step": 1754,
      "train_runtime": 3702.6831,
      "train_tokens_per_second": 884.631
    },
    {
      "epoch": 1.7733131159969675,
      "grad_norm": 0.37655583024024963,
      "learning_rate": 2.288607594936709e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 3278086,
      "step": 1755,
      "train_runtime": 3704.8254,
      "train_tokens_per_second": 884.815
    },
    {
      "epoch": 1.7743239828152642,
      "grad_norm": 0.3488801121711731,
      "learning_rate": 2.278481012658228e-05,
      "loss": 0.1461,
      "num_input_tokens_seen": 3279882,
      "step": 1756,
      "train_runtime": 3706.5955,
      "train_tokens_per_second": 884.877
    },
    {
      "epoch": 1.7753348496335608,
      "grad_norm": 0.31840255856513977,
      "learning_rate": 2.2683544303797468e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 3281578,
      "step": 1757,
      "train_runtime": 3708.2196,
      "train_tokens_per_second": 884.947
    },
    {
      "epoch": 1.7763457164518575,
      "grad_norm": 0.2555081844329834,
      "learning_rate": 2.258227848101266e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 3283602,
      "step": 1758,
      "train_runtime": 3710.0757,
      "train_tokens_per_second": 885.05
    },
    {
      "epoch": 1.7773565832701541,
      "grad_norm": 0.36034834384918213,
      "learning_rate": 2.248101265822785e-05,
      "loss": 0.1254,
      "num_input_tokens_seen": 3285178,
      "step": 1759,
      "train_runtime": 3711.761,
      "train_tokens_per_second": 885.073
    },
    {
      "epoch": 1.7783674500884508,
      "grad_norm": 0.2103455513715744,
      "learning_rate": 2.2379746835443037e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 3286980,
      "step": 1760,
      "train_runtime": 3713.5712,
      "train_tokens_per_second": 885.126
    },
    {
      "epoch": 1.7793783169067474,
      "grad_norm": 0.34630411863327026,
      "learning_rate": 2.2278481012658228e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 3288718,
      "step": 1761,
      "train_runtime": 3715.2592,
      "train_tokens_per_second": 885.192
    },
    {
      "epoch": 1.780389183725044,
      "grad_norm": 0.2746468782424927,
      "learning_rate": 2.217721518987342e-05,
      "loss": 0.0984,
      "num_input_tokens_seen": 3290670,
      "step": 1762,
      "train_runtime": 3717.0411,
      "train_tokens_per_second": 885.293
    },
    {
      "epoch": 1.781400050543341,
      "grad_norm": 0.5191876888275146,
      "learning_rate": 2.207594936708861e-05,
      "loss": 0.0789,
      "num_input_tokens_seen": 3292230,
      "step": 1763,
      "train_runtime": 3718.6756,
      "train_tokens_per_second": 885.323
    },
    {
      "epoch": 1.7824109173616376,
      "grad_norm": 0.3496468663215637,
      "learning_rate": 2.1974683544303797e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 3293984,
      "step": 1764,
      "train_runtime": 3720.4005,
      "train_tokens_per_second": 885.384
    },
    {
      "epoch": 1.7834217841799342,
      "grad_norm": 0.46741724014282227,
      "learning_rate": 2.1873417721518988e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 3296392,
      "step": 1765,
      "train_runtime": 3722.4133,
      "train_tokens_per_second": 885.552
    },
    {
      "epoch": 1.7844326509982311,
      "grad_norm": 0.33283084630966187,
      "learning_rate": 2.177215189873418e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 3298364,
      "step": 1766,
      "train_runtime": 3724.1851,
      "train_tokens_per_second": 885.661
    },
    {
      "epoch": 1.7854435178165278,
      "grad_norm": 0.3500096797943115,
      "learning_rate": 2.1670886075949366e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 3299992,
      "step": 1767,
      "train_runtime": 3725.8865,
      "train_tokens_per_second": 885.693
    },
    {
      "epoch": 1.7864543846348244,
      "grad_norm": 0.23019525408744812,
      "learning_rate": 2.1569620253164557e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 3301884,
      "step": 1768,
      "train_runtime": 3727.6217,
      "train_tokens_per_second": 885.788
    },
    {
      "epoch": 1.787465251453121,
      "grad_norm": 0.42400750517845154,
      "learning_rate": 2.1468354430379748e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 3303846,
      "step": 1769,
      "train_runtime": 3729.4567,
      "train_tokens_per_second": 885.879
    },
    {
      "epoch": 1.7884761182714177,
      "grad_norm": 0.3044002950191498,
      "learning_rate": 2.136708860759494e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 3305626,
      "step": 1770,
      "train_runtime": 3731.1867,
      "train_tokens_per_second": 885.945
    },
    {
      "epoch": 1.7894869850897144,
      "grad_norm": 0.30707088112831116,
      "learning_rate": 2.1265822784810126e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 3307618,
      "step": 1771,
      "train_runtime": 3734.7376,
      "train_tokens_per_second": 885.636
    },
    {
      "epoch": 1.790497851908011,
      "grad_norm": 0.32700610160827637,
      "learning_rate": 2.1164556962025317e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 3309164,
      "step": 1772,
      "train_runtime": 3736.3737,
      "train_tokens_per_second": 885.662
    },
    {
      "epoch": 1.7915087187263077,
      "grad_norm": 0.27235326170921326,
      "learning_rate": 2.1063291139240508e-05,
      "loss": 0.0986,
      "num_input_tokens_seen": 3311314,
      "step": 1773,
      "train_runtime": 3738.2442,
      "train_tokens_per_second": 885.794
    },
    {
      "epoch": 1.7925195855446043,
      "grad_norm": 0.2995695471763611,
      "learning_rate": 2.0962025316455695e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 3312934,
      "step": 1774,
      "train_runtime": 3739.8736,
      "train_tokens_per_second": 885.841
    },
    {
      "epoch": 1.7935304523629012,
      "grad_norm": 0.5493440628051758,
      "learning_rate": 2.0860759493670886e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 3314462,
      "step": 1775,
      "train_runtime": 3741.4919,
      "train_tokens_per_second": 885.866
    },
    {
      "epoch": 1.7945413191811979,
      "grad_norm": 0.31136009097099304,
      "learning_rate": 2.0759493670886077e-05,
      "loss": 0.0986,
      "num_input_tokens_seen": 3316446,
      "step": 1776,
      "train_runtime": 3743.3178,
      "train_tokens_per_second": 885.964
    },
    {
      "epoch": 1.7955521859994947,
      "grad_norm": 0.3378680348396301,
      "learning_rate": 2.0658227848101268e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 3319010,
      "step": 1777,
      "train_runtime": 3745.3702,
      "train_tokens_per_second": 886.163
    },
    {
      "epoch": 1.7965630528177914,
      "grad_norm": 0.37042444944381714,
      "learning_rate": 2.0556962025316455e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 3320802,
      "step": 1778,
      "train_runtime": 3747.0782,
      "train_tokens_per_second": 886.238
    },
    {
      "epoch": 1.797573919636088,
      "grad_norm": 0.38629478216171265,
      "learning_rate": 2.0455696202531646e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 3322928,
      "step": 1779,
      "train_runtime": 3748.9524,
      "train_tokens_per_second": 886.362
    },
    {
      "epoch": 1.7985847864543847,
      "grad_norm": 0.40462711453437805,
      "learning_rate": 2.0354430379746837e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 3324696,
      "step": 1780,
      "train_runtime": 3750.6599,
      "train_tokens_per_second": 886.43
    },
    {
      "epoch": 1.7995956532726813,
      "grad_norm": 0.302253782749176,
      "learning_rate": 2.0253164556962025e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 3327390,
      "step": 1781,
      "train_runtime": 3752.8492,
      "train_tokens_per_second": 886.63
    },
    {
      "epoch": 1.800606520090978,
      "grad_norm": 0.29382818937301636,
      "learning_rate": 2.0151898734177215e-05,
      "loss": 0.1435,
      "num_input_tokens_seen": 3329486,
      "step": 1782,
      "train_runtime": 3754.7199,
      "train_tokens_per_second": 886.747
    },
    {
      "epoch": 1.800606520090978,
      "eval_loss": 0.20694167912006378,
      "eval_runtime": 61.4529,
      "eval_samples_per_second": 14.304,
      "eval_steps_per_second": 7.16,
      "num_input_tokens_seen": 3329486,
      "step": 1782
    },
    {
      "epoch": 1.8016173869092746,
      "grad_norm": 0.3214450478553772,
      "learning_rate": 2.0050632911392406e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 3331678,
      "step": 1783,
      "train_runtime": 3818.12,
      "train_tokens_per_second": 872.596
    },
    {
      "epoch": 1.8026282537275713,
      "grad_norm": 0.3057961165904999,
      "learning_rate": 1.9949367088607597e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 3333514,
      "step": 1784,
      "train_runtime": 3819.8521,
      "train_tokens_per_second": 872.681
    },
    {
      "epoch": 1.803639120545868,
      "grad_norm": 0.47353193163871765,
      "learning_rate": 1.9848101265822785e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 3334950,
      "step": 1785,
      "train_runtime": 3821.4494,
      "train_tokens_per_second": 872.692
    },
    {
      "epoch": 1.8046499873641648,
      "grad_norm": 0.16872140765190125,
      "learning_rate": 1.9746835443037975e-05,
      "loss": 0.0205,
      "num_input_tokens_seen": 3337092,
      "step": 1786,
      "train_runtime": 3823.3552,
      "train_tokens_per_second": 872.818
    },
    {
      "epoch": 1.8056608541824615,
      "grad_norm": 0.4897010624408722,
      "learning_rate": 1.9645569620253166e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 3338772,
      "step": 1787,
      "train_runtime": 3825.0592,
      "train_tokens_per_second": 872.868
    },
    {
      "epoch": 1.8066717210007581,
      "grad_norm": 0.38819557428359985,
      "learning_rate": 1.9544303797468354e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 3340352,
      "step": 1788,
      "train_runtime": 3826.7388,
      "train_tokens_per_second": 872.898
    },
    {
      "epoch": 1.807682587819055,
      "grad_norm": 0.3572562038898468,
      "learning_rate": 1.9443037974683544e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 3341712,
      "step": 1789,
      "train_runtime": 3828.3062,
      "train_tokens_per_second": 872.896
    },
    {
      "epoch": 1.8086934546373516,
      "grad_norm": 0.2987414300441742,
      "learning_rate": 1.9341772151898735e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 3343778,
      "step": 1790,
      "train_runtime": 3830.2175,
      "train_tokens_per_second": 872.999
    },
    {
      "epoch": 1.8097043214556483,
      "grad_norm": 0.41519400477409363,
      "learning_rate": 1.9240506329113926e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 3345650,
      "step": 1791,
      "train_runtime": 3831.9996,
      "train_tokens_per_second": 873.082
    },
    {
      "epoch": 1.810715188273945,
      "grad_norm": 0.4155493378639221,
      "learning_rate": 1.9139240506329114e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 3347852,
      "step": 1792,
      "train_runtime": 3833.8915,
      "train_tokens_per_second": 873.226
    },
    {
      "epoch": 1.8117260550922416,
      "grad_norm": 0.3513873517513275,
      "learning_rate": 1.9037974683544304e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 3349184,
      "step": 1793,
      "train_runtime": 3835.4717,
      "train_tokens_per_second": 873.213
    },
    {
      "epoch": 1.8127369219105383,
      "grad_norm": 0.3495260179042816,
      "learning_rate": 1.8936708860759495e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 3350614,
      "step": 1794,
      "train_runtime": 3837.0898,
      "train_tokens_per_second": 873.218
    },
    {
      "epoch": 1.813747788728835,
      "grad_norm": 0.20336616039276123,
      "learning_rate": 1.8835443037974683e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 3353054,
      "step": 1795,
      "train_runtime": 3839.1117,
      "train_tokens_per_second": 873.393
    },
    {
      "epoch": 1.8147586555471316,
      "grad_norm": 0.3769696056842804,
      "learning_rate": 1.8734177215189874e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 3354852,
      "step": 1796,
      "train_runtime": 3840.8695,
      "train_tokens_per_second": 873.462
    },
    {
      "epoch": 1.8157695223654282,
      "grad_norm": 0.21609970927238464,
      "learning_rate": 1.8632911392405064e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 3357008,
      "step": 1797,
      "train_runtime": 3842.7192,
      "train_tokens_per_second": 873.602
    },
    {
      "epoch": 1.816780389183725,
      "grad_norm": 0.37451791763305664,
      "learning_rate": 1.8531645569620255e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 3358872,
      "step": 1798,
      "train_runtime": 3844.546,
      "train_tokens_per_second": 873.672
    },
    {
      "epoch": 1.8177912560020217,
      "grad_norm": 0.3367696702480316,
      "learning_rate": 1.8430379746835443e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 3360660,
      "step": 1799,
      "train_runtime": 3846.252,
      "train_tokens_per_second": 873.749
    },
    {
      "epoch": 1.8188021228203184,
      "grad_norm": 0.2870897948741913,
      "learning_rate": 1.8329113924050634e-05,
      "loss": 0.0223,
      "num_input_tokens_seen": 3362738,
      "step": 1800,
      "train_runtime": 3848.1006,
      "train_tokens_per_second": 873.87
    },
    {
      "epoch": 1.8198129896386153,
      "grad_norm": 0.3743988573551178,
      "learning_rate": 1.8227848101265824e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 3364764,
      "step": 1801,
      "train_runtime": 3851.9506,
      "train_tokens_per_second": 873.522
    },
    {
      "epoch": 1.820823856456912,
      "grad_norm": 0.22562061250209808,
      "learning_rate": 1.8126582278481012e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 3366966,
      "step": 1802,
      "train_runtime": 3853.901,
      "train_tokens_per_second": 873.651
    },
    {
      "epoch": 1.8218347232752086,
      "grad_norm": 0.3605365455150604,
      "learning_rate": 1.8025316455696203e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 3368778,
      "step": 1803,
      "train_runtime": 3855.622,
      "train_tokens_per_second": 873.731
    },
    {
      "epoch": 1.8228455900935052,
      "grad_norm": 0.43901732563972473,
      "learning_rate": 1.7924050632911394e-05,
      "loss": 0.1063,
      "num_input_tokens_seen": 3370368,
      "step": 1804,
      "train_runtime": 3857.2,
      "train_tokens_per_second": 873.786
    },
    {
      "epoch": 1.8238564569118019,
      "grad_norm": 0.35778892040252686,
      "learning_rate": 1.7822784810126584e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 3371894,
      "step": 1805,
      "train_runtime": 3858.8239,
      "train_tokens_per_second": 873.814
    },
    {
      "epoch": 1.8248673237300985,
      "grad_norm": 0.3990665674209595,
      "learning_rate": 1.7721518987341772e-05,
      "loss": 0.0982,
      "num_input_tokens_seen": 3373826,
      "step": 1806,
      "train_runtime": 3860.6086,
      "train_tokens_per_second": 873.91
    },
    {
      "epoch": 1.8258781905483952,
      "grad_norm": 0.44509217143058777,
      "learning_rate": 1.7620253164556963e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 3375664,
      "step": 1807,
      "train_runtime": 3862.3432,
      "train_tokens_per_second": 873.994
    },
    {
      "epoch": 1.8268890573666918,
      "grad_norm": 0.3205857574939728,
      "learning_rate": 1.7518987341772154e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 3377254,
      "step": 1808,
      "train_runtime": 3863.9629,
      "train_tokens_per_second": 874.039
    },
    {
      "epoch": 1.8278999241849885,
      "grad_norm": 0.2987648844718933,
      "learning_rate": 1.7417721518987344e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 3379152,
      "step": 1809,
      "train_runtime": 3865.6868,
      "train_tokens_per_second": 874.14
    },
    {
      "epoch": 1.8289107910032854,
      "grad_norm": 0.35453036427497864,
      "learning_rate": 1.7316455696202532e-05,
      "loss": 0.0238,
      "num_input_tokens_seen": 3380618,
      "step": 1810,
      "train_runtime": 3867.2852,
      "train_tokens_per_second": 874.158
    },
    {
      "epoch": 1.829921657821582,
      "grad_norm": 0.2638246417045593,
      "learning_rate": 1.7215189873417723e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 3382000,
      "step": 1811,
      "train_runtime": 3868.875,
      "train_tokens_per_second": 874.156
    },
    {
      "epoch": 1.8309325246398787,
      "grad_norm": 0.28177931904792786,
      "learning_rate": 1.7113924050632914e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 3383976,
      "step": 1812,
      "train_runtime": 3870.602,
      "train_tokens_per_second": 874.276
    },
    {
      "epoch": 1.8319433914581755,
      "grad_norm": 0.3644472658634186,
      "learning_rate": 1.70126582278481e-05,
      "loss": 0.1228,
      "num_input_tokens_seen": 3385922,
      "step": 1813,
      "train_runtime": 3872.3279,
      "train_tokens_per_second": 874.389
    },
    {
      "epoch": 1.8329542582764722,
      "grad_norm": 0.33471110463142395,
      "learning_rate": 1.6911392405063292e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 3387892,
      "step": 1814,
      "train_runtime": 3874.1621,
      "train_tokens_per_second": 874.484
    },
    {
      "epoch": 1.8339651250947688,
      "grad_norm": 0.347408264875412,
      "learning_rate": 1.6810126582278483e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 3389992,
      "step": 1815,
      "train_runtime": 3875.9632,
      "train_tokens_per_second": 874.619
    },
    {
      "epoch": 1.8349759919130655,
      "grad_norm": 0.4522991180419922,
      "learning_rate": 1.6708860759493674e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 3392154,
      "step": 1816,
      "train_runtime": 3877.7827,
      "train_tokens_per_second": 874.766
    },
    {
      "epoch": 1.8359868587313621,
      "grad_norm": 0.3917560279369354,
      "learning_rate": 1.660759493670886e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 3393572,
      "step": 1817,
      "train_runtime": 3879.4085,
      "train_tokens_per_second": 874.765
    },
    {
      "epoch": 1.8369977255496588,
      "grad_norm": 0.36581575870513916,
      "learning_rate": 1.6506329113924052e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 3395228,
      "step": 1818,
      "train_runtime": 3881.0571,
      "train_tokens_per_second": 874.82
    },
    {
      "epoch": 1.8380085923679554,
      "grad_norm": 0.3207392394542694,
      "learning_rate": 1.6405063291139243e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 3396648,
      "step": 1819,
      "train_runtime": 3882.6866,
      "train_tokens_per_second": 874.819
    },
    {
      "epoch": 1.839019459186252,
      "grad_norm": 0.4688638746738434,
      "learning_rate": 1.630379746835443e-05,
      "loss": 0.1553,
      "num_input_tokens_seen": 3398164,
      "step": 1820,
      "train_runtime": 3884.2839,
      "train_tokens_per_second": 874.85
    },
    {
      "epoch": 1.8400303260045487,
      "grad_norm": 0.32755982875823975,
      "learning_rate": 1.620253164556962e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 3399916,
      "step": 1821,
      "train_runtime": 3885.9857,
      "train_tokens_per_second": 874.917
    },
    {
      "epoch": 1.8410411928228456,
      "grad_norm": 0.4200737476348877,
      "learning_rate": 1.6101265822784812e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 3401452,
      "step": 1822,
      "train_runtime": 3887.6163,
      "train_tokens_per_second": 874.945
    },
    {
      "epoch": 1.8420520596411423,
      "grad_norm": 0.3440350592136383,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 3403272,
      "step": 1823,
      "train_runtime": 3889.3348,
      "train_tokens_per_second": 875.027
    },
    {
      "epoch": 1.843062926459439,
      "grad_norm": 0.30478882789611816,
      "learning_rate": 1.589873417721519e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 3404636,
      "step": 1824,
      "train_runtime": 3890.8991,
      "train_tokens_per_second": 875.026
    },
    {
      "epoch": 1.8440737932777358,
      "grad_norm": 0.46187645196914673,
      "learning_rate": 1.579746835443038e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 3406056,
      "step": 1825,
      "train_runtime": 3892.4702,
      "train_tokens_per_second": 875.037
    },
    {
      "epoch": 1.8450846600960324,
      "grad_norm": 0.35042065382003784,
      "learning_rate": 1.5696202531645572e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 3407952,
      "step": 1826,
      "train_runtime": 3894.2944,
      "train_tokens_per_second": 875.114
    },
    {
      "epoch": 1.846095526914329,
      "grad_norm": 0.3137063980102539,
      "learning_rate": 1.559493670886076e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 3409784,
      "step": 1827,
      "train_runtime": 3896.117,
      "train_tokens_per_second": 875.175
    },
    {
      "epoch": 1.8471063937326258,
      "grad_norm": 0.40788382291793823,
      "learning_rate": 1.549367088607595e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 3411602,
      "step": 1828,
      "train_runtime": 3897.8444,
      "train_tokens_per_second": 875.254
    },
    {
      "epoch": 1.8481172605509224,
      "grad_norm": 0.17767295241355896,
      "learning_rate": 1.539240506329114e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 3413314,
      "step": 1829,
      "train_runtime": 3899.4894,
      "train_tokens_per_second": 875.323
    },
    {
      "epoch": 1.849128127369219,
      "grad_norm": 0.4391786754131317,
      "learning_rate": 1.5291139240506332e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 3415466,
      "step": 1830,
      "train_runtime": 3901.4023,
      "train_tokens_per_second": 875.446
    },
    {
      "epoch": 1.8501389941875157,
      "grad_norm": 0.35986626148223877,
      "learning_rate": 1.5189873417721521e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 3417180,
      "step": 1831,
      "train_runtime": 3904.9063,
      "train_tokens_per_second": 875.099
    },
    {
      "epoch": 1.8511498610058124,
      "grad_norm": 0.3671015799045563,
      "learning_rate": 1.508860759493671e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 3418982,
      "step": 1832,
      "train_runtime": 3906.6302,
      "train_tokens_per_second": 875.174
    },
    {
      "epoch": 1.852160727824109,
      "grad_norm": 0.6754432916641235,
      "learning_rate": 1.49873417721519e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 3420756,
      "step": 1833,
      "train_runtime": 3908.3341,
      "train_tokens_per_second": 875.247
    },
    {
      "epoch": 1.8531715946424059,
      "grad_norm": 0.2855415642261505,
      "learning_rate": 1.488607594936709e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 3422560,
      "step": 1834,
      "train_runtime": 3910.0398,
      "train_tokens_per_second": 875.326
    },
    {
      "epoch": 1.8541824614607025,
      "grad_norm": 0.41475895047187805,
      "learning_rate": 1.478481012658228e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 3424078,
      "step": 1835,
      "train_runtime": 3911.6975,
      "train_tokens_per_second": 875.343
    },
    {
      "epoch": 1.8551933282789992,
      "grad_norm": 0.20816388726234436,
      "learning_rate": 1.468354430379747e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 3426512,
      "step": 1836,
      "train_runtime": 3913.7126,
      "train_tokens_per_second": 875.514
    },
    {
      "epoch": 1.856204195097296,
      "grad_norm": 0.46875783801078796,
      "learning_rate": 1.458227848101266e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 3428360,
      "step": 1837,
      "train_runtime": 3915.4882,
      "train_tokens_per_second": 875.589
    },
    {
      "epoch": 1.8572150619155927,
      "grad_norm": 0.3428696393966675,
      "learning_rate": 1.448101265822785e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 3430342,
      "step": 1838,
      "train_runtime": 3917.2267,
      "train_tokens_per_second": 875.707
    },
    {
      "epoch": 1.8582259287338894,
      "grad_norm": 0.2921421229839325,
      "learning_rate": 1.437974683544304e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 3431844,
      "step": 1839,
      "train_runtime": 3918.8142,
      "train_tokens_per_second": 875.735
    },
    {
      "epoch": 1.859236795552186,
      "grad_norm": 0.41684088110923767,
      "learning_rate": 1.4278481012658228e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 3433804,
      "step": 1840,
      "train_runtime": 3920.5941,
      "train_tokens_per_second": 875.838
    },
    {
      "epoch": 1.8602476623704827,
      "grad_norm": 0.3576996624469757,
      "learning_rate": 1.417721518987342e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 3435836,
      "step": 1841,
      "train_runtime": 3922.3962,
      "train_tokens_per_second": 875.953
    },
    {
      "epoch": 1.8612585291887793,
      "grad_norm": 0.3274343013763428,
      "learning_rate": 1.4075949367088608e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 3437454,
      "step": 1842,
      "train_runtime": 3924.0883,
      "train_tokens_per_second": 875.988
    },
    {
      "epoch": 1.862269396007076,
      "grad_norm": 0.40019598603248596,
      "learning_rate": 1.39746835443038e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 3439068,
      "step": 1843,
      "train_runtime": 3925.7205,
      "train_tokens_per_second": 876.035
    },
    {
      "epoch": 1.8632802628253726,
      "grad_norm": 0.45412638783454895,
      "learning_rate": 1.3873417721518988e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 3440584,
      "step": 1844,
      "train_runtime": 3927.3415,
      "train_tokens_per_second": 876.059
    },
    {
      "epoch": 1.8642911296436695,
      "grad_norm": 0.3301598131656647,
      "learning_rate": 1.377215189873418e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 3442392,
      "step": 1845,
      "train_runtime": 3929.0573,
      "train_tokens_per_second": 876.137
    },
    {
      "epoch": 1.8653019964619661,
      "grad_norm": 0.5462703108787537,
      "learning_rate": 1.3670886075949368e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 3444012,
      "step": 1846,
      "train_runtime": 3930.6795,
      "train_tokens_per_second": 876.187
    },
    {
      "epoch": 1.8663128632802628,
      "grad_norm": 0.37690606713294983,
      "learning_rate": 1.3569620253164557e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 3445536,
      "step": 1847,
      "train_runtime": 3932.3011,
      "train_tokens_per_second": 876.214
    },
    {
      "epoch": 1.8673237300985597,
      "grad_norm": 0.257944792509079,
      "learning_rate": 1.3468354430379748e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 3447234,
      "step": 1848,
      "train_runtime": 3933.9451,
      "train_tokens_per_second": 876.279
    },
    {
      "epoch": 1.8683345969168563,
      "grad_norm": 0.31285640597343445,
      "learning_rate": 1.3367088607594937e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 3449022,
      "step": 1849,
      "train_runtime": 3935.6335,
      "train_tokens_per_second": 876.358
    },
    {
      "epoch": 1.869345463735153,
      "grad_norm": 0.32242077589035034,
      "learning_rate": 1.3265822784810128e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 3450846,
      "step": 1850,
      "train_runtime": 3937.3614,
      "train_tokens_per_second": 876.436
    },
    {
      "epoch": 1.8703563305534496,
      "grad_norm": 0.281527578830719,
      "learning_rate": 1.3164556962025317e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 3452606,
      "step": 1851,
      "train_runtime": 3939.0014,
      "train_tokens_per_second": 876.518
    },
    {
      "epoch": 1.8713671973717463,
      "grad_norm": 0.22572745382785797,
      "learning_rate": 1.3063291139240508e-05,
      "loss": 0.0144,
      "num_input_tokens_seen": 3454210,
      "step": 1852,
      "train_runtime": 3940.6835,
      "train_tokens_per_second": 876.551
    },
    {
      "epoch": 1.872378064190043,
      "grad_norm": 0.32201293110847473,
      "learning_rate": 1.2962025316455697e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 3455988,
      "step": 1853,
      "train_runtime": 3942.4027,
      "train_tokens_per_second": 876.62
    },
    {
      "epoch": 1.8733889310083396,
      "grad_norm": 0.4775518476963043,
      "learning_rate": 1.2860759493670888e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 3457496,
      "step": 1854,
      "train_runtime": 3943.9925,
      "train_tokens_per_second": 876.649
    },
    {
      "epoch": 1.8743997978266362,
      "grad_norm": 0.21644511818885803,
      "learning_rate": 1.2759493670886077e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 3459732,
      "step": 1855,
      "train_runtime": 3945.9906,
      "train_tokens_per_second": 876.771
    },
    {
      "epoch": 1.875410664644933,
      "grad_norm": 0.29097041487693787,
      "learning_rate": 1.2658227848101267e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 3461420,
      "step": 1856,
      "train_runtime": 3947.691,
      "train_tokens_per_second": 876.821
    },
    {
      "epoch": 1.8764215314632298,
      "grad_norm": 0.2558102309703827,
      "learning_rate": 1.2556962025316457e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 3463444,
      "step": 1857,
      "train_runtime": 3949.5483,
      "train_tokens_per_second": 876.922
    },
    {
      "epoch": 1.8774323982815264,
      "grad_norm": 0.3625556230545044,
      "learning_rate": 1.2455696202531645e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 3465548,
      "step": 1858,
      "train_runtime": 3951.4562,
      "train_tokens_per_second": 877.031
    },
    {
      "epoch": 1.878443265099823,
      "grad_norm": 0.28841790556907654,
      "learning_rate": 1.2354430379746836e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 3467678,
      "step": 1859,
      "train_runtime": 3953.3142,
      "train_tokens_per_second": 877.157
    },
    {
      "epoch": 1.87945413191812,
      "grad_norm": 0.34299448132514954,
      "learning_rate": 1.2253164556962025e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 3469522,
      "step": 1860,
      "train_runtime": 3955.088,
      "train_tokens_per_second": 877.23
    },
    {
      "epoch": 1.8804649987364166,
      "grad_norm": 0.3639189898967743,
      "learning_rate": 1.2151898734177216e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 3471412,
      "step": 1861,
      "train_runtime": 3958.8851,
      "train_tokens_per_second": 876.866
    },
    {
      "epoch": 1.8814758655547132,
      "grad_norm": 0.45584601163864136,
      "learning_rate": 1.2050632911392405e-05,
      "loss": 0.0928,
      "num_input_tokens_seen": 3473686,
      "step": 1862,
      "train_runtime": 3960.8914,
      "train_tokens_per_second": 876.996
    },
    {
      "epoch": 1.88248673237301,
      "grad_norm": 0.28255757689476013,
      "learning_rate": 1.1949367088607596e-05,
      "loss": 0.1055,
      "num_input_tokens_seen": 3475678,
      "step": 1863,
      "train_runtime": 3962.6863,
      "train_tokens_per_second": 877.101
    },
    {
      "epoch": 1.8834975991913065,
      "grad_norm": 0.39519962668418884,
      "learning_rate": 1.1848101265822785e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 3477542,
      "step": 1864,
      "train_runtime": 3964.341,
      "train_tokens_per_second": 877.206
    },
    {
      "epoch": 1.8845084660096032,
      "grad_norm": 0.18378497660160065,
      "learning_rate": 1.1746835443037974e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 3479142,
      "step": 1865,
      "train_runtime": 3966.0296,
      "train_tokens_per_second": 877.236
    },
    {
      "epoch": 1.8855193328278999,
      "grad_norm": 0.21587584912776947,
      "learning_rate": 1.1645569620253165e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 3481038,
      "step": 1866,
      "train_runtime": 3967.7558,
      "train_tokens_per_second": 877.332
    },
    {
      "epoch": 1.8865301996461965,
      "grad_norm": 0.41051149368286133,
      "learning_rate": 1.1544303797468354e-05,
      "loss": 0.0752,
      "num_input_tokens_seen": 3482548,
      "step": 1867,
      "train_runtime": 3969.3621,
      "train_tokens_per_second": 877.357
    },
    {
      "epoch": 1.8875410664644932,
      "grad_norm": 0.34189385175704956,
      "learning_rate": 1.1443037974683545e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 3484614,
      "step": 1868,
      "train_runtime": 3971.1844,
      "train_tokens_per_second": 877.475
    },
    {
      "epoch": 1.88855193328279,
      "grad_norm": 0.2599343955516815,
      "learning_rate": 1.1341772151898734e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 3486578,
      "step": 1869,
      "train_runtime": 3972.9682,
      "train_tokens_per_second": 877.575
    },
    {
      "epoch": 1.8895628001010867,
      "grad_norm": 0.3474741280078888,
      "learning_rate": 1.1240506329113925e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 3488246,
      "step": 1870,
      "train_runtime": 3974.5922,
      "train_tokens_per_second": 877.636
    },
    {
      "epoch": 1.8905736669193833,
      "grad_norm": 0.354998379945755,
      "learning_rate": 1.1139240506329114e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 3490076,
      "step": 1871,
      "train_runtime": 3976.316,
      "train_tokens_per_second": 877.716
    },
    {
      "epoch": 1.8915845337376802,
      "grad_norm": 0.3715677261352539,
      "learning_rate": 1.1037974683544305e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 3492288,
      "step": 1872,
      "train_runtime": 3978.2722,
      "train_tokens_per_second": 877.84
    },
    {
      "epoch": 1.8925954005559769,
      "grad_norm": 0.3217340409755707,
      "learning_rate": 1.0936708860759494e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 3494018,
      "step": 1873,
      "train_runtime": 3979.9176,
      "train_tokens_per_second": 877.912
    },
    {
      "epoch": 1.8936062673742735,
      "grad_norm": 0.2525309920310974,
      "learning_rate": 1.0835443037974683e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 3495862,
      "step": 1874,
      "train_runtime": 3981.7067,
      "train_tokens_per_second": 877.981
    },
    {
      "epoch": 1.8946171341925702,
      "grad_norm": 0.2678094804286957,
      "learning_rate": 1.0734177215189874e-05,
      "loss": 0.1625,
      "num_input_tokens_seen": 3497378,
      "step": 1875,
      "train_runtime": 3983.3387,
      "train_tokens_per_second": 878.002
    },
    {
      "epoch": 1.8956280010108668,
      "grad_norm": 0.38275614380836487,
      "learning_rate": 1.0632911392405063e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 3498964,
      "step": 1876,
      "train_runtime": 3985.0118,
      "train_tokens_per_second": 878.031
    },
    {
      "epoch": 1.8966388678291635,
      "grad_norm": 0.2158384770154953,
      "learning_rate": 1.0531645569620254e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 3500910,
      "step": 1877,
      "train_runtime": 3986.8023,
      "train_tokens_per_second": 878.125
    },
    {
      "epoch": 1.8976497346474601,
      "grad_norm": 0.5794618725776672,
      "learning_rate": 1.0430379746835443e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 3502370,
      "step": 1878,
      "train_runtime": 3988.4101,
      "train_tokens_per_second": 878.137
    },
    {
      "epoch": 1.8986606014657568,
      "grad_norm": 0.31127601861953735,
      "learning_rate": 1.0329113924050634e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 3504322,
      "step": 1879,
      "train_runtime": 3990.1957,
      "train_tokens_per_second": 878.233
    },
    {
      "epoch": 1.8996714682840534,
      "grad_norm": 0.3494141101837158,
      "learning_rate": 1.0227848101265823e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 3506208,
      "step": 1880,
      "train_runtime": 3991.9845,
      "train_tokens_per_second": 878.312
    },
    {
      "epoch": 1.9006823351023503,
      "grad_norm": 0.22893235087394714,
      "learning_rate": 1.0126582278481012e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 3508008,
      "step": 1881,
      "train_runtime": 3993.6982,
      "train_tokens_per_second": 878.386
    },
    {
      "epoch": 1.901693201920647,
      "grad_norm": 0.37710627913475037,
      "learning_rate": 1.0025316455696203e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 3510068,
      "step": 1882,
      "train_runtime": 3995.5624,
      "train_tokens_per_second": 878.492
    },
    {
      "epoch": 1.9027040687389436,
      "grad_norm": 0.30796605348587036,
      "learning_rate": 9.924050632911392e-06,
      "loss": 0.0956,
      "num_input_tokens_seen": 3512120,
      "step": 1883,
      "train_runtime": 3997.3825,
      "train_tokens_per_second": 878.605
    },
    {
      "epoch": 1.9037149355572405,
      "grad_norm": 0.19698543846607208,
      "learning_rate": 9.822784810126583e-06,
      "loss": 0.0373,
      "num_input_tokens_seen": 3513688,
      "step": 1884,
      "train_runtime": 3999.0201,
      "train_tokens_per_second": 878.637
    },
    {
      "epoch": 1.9047258023755371,
      "grad_norm": 0.3822472095489502,
      "learning_rate": 9.721518987341772e-06,
      "loss": 0.106,
      "num_input_tokens_seen": 3515550,
      "step": 1885,
      "train_runtime": 4000.8138,
      "train_tokens_per_second": 878.709
    },
    {
      "epoch": 1.9057366691938338,
      "grad_norm": 0.362237811088562,
      "learning_rate": 9.620253164556963e-06,
      "loss": 0.1226,
      "num_input_tokens_seen": 3517484,
      "step": 1886,
      "train_runtime": 4002.6698,
      "train_tokens_per_second": 878.784
    },
    {
      "epoch": 1.9067475360121304,
      "grad_norm": 0.4247080981731415,
      "learning_rate": 9.518987341772152e-06,
      "loss": 0.1002,
      "num_input_tokens_seen": 3519024,
      "step": 1887,
      "train_runtime": 4004.3403,
      "train_tokens_per_second": 878.802
    },
    {
      "epoch": 1.907758402830427,
      "grad_norm": 0.33333954215049744,
      "learning_rate": 9.417721518987341e-06,
      "loss": 0.1222,
      "num_input_tokens_seen": 3520706,
      "step": 1888,
      "train_runtime": 4005.9704,
      "train_tokens_per_second": 878.865
    },
    {
      "epoch": 1.9087692696487237,
      "grad_norm": 0.4789365530014038,
      "learning_rate": 9.316455696202532e-06,
      "loss": 0.1789,
      "num_input_tokens_seen": 3522746,
      "step": 1889,
      "train_runtime": 4007.8431,
      "train_tokens_per_second": 878.963
    },
    {
      "epoch": 1.9097801364670204,
      "grad_norm": 0.34310978651046753,
      "learning_rate": 9.215189873417721e-06,
      "loss": 0.0462,
      "num_input_tokens_seen": 3524824,
      "step": 1890,
      "train_runtime": 4009.7533,
      "train_tokens_per_second": 879.063
    },
    {
      "epoch": 1.910791003285317,
      "grad_norm": 0.31715911626815796,
      "learning_rate": 9.113924050632912e-06,
      "loss": 0.0603,
      "num_input_tokens_seen": 3526718,
      "step": 1891,
      "train_runtime": 4013.5824,
      "train_tokens_per_second": 878.696
    },
    {
      "epoch": 1.9118018701036137,
      "grad_norm": 0.3918403685092926,
      "learning_rate": 9.012658227848101e-06,
      "loss": 0.0488,
      "num_input_tokens_seen": 3528542,
      "step": 1892,
      "train_runtime": 4015.368,
      "train_tokens_per_second": 878.759
    },
    {
      "epoch": 1.9128127369219106,
      "grad_norm": 0.4037635624408722,
      "learning_rate": 8.911392405063292e-06,
      "loss": 0.0704,
      "num_input_tokens_seen": 3530010,
      "step": 1893,
      "train_runtime": 4016.9815,
      "train_tokens_per_second": 878.772
    },
    {
      "epoch": 1.9138236037402072,
      "grad_norm": 0.25474458932876587,
      "learning_rate": 8.810126582278481e-06,
      "loss": 0.0843,
      "num_input_tokens_seen": 3531598,
      "step": 1894,
      "train_runtime": 4018.6069,
      "train_tokens_per_second": 878.812
    },
    {
      "epoch": 1.9148344705585039,
      "grad_norm": 0.28693681955337524,
      "learning_rate": 8.708860759493672e-06,
      "loss": 0.0828,
      "num_input_tokens_seen": 3533364,
      "step": 1895,
      "train_runtime": 4020.3776,
      "train_tokens_per_second": 878.864
    },
    {
      "epoch": 1.9158453373768007,
      "grad_norm": 0.35959458351135254,
      "learning_rate": 8.607594936708861e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 3535146,
      "step": 1896,
      "train_runtime": 4022.0371,
      "train_tokens_per_second": 878.944
    },
    {
      "epoch": 1.9168562041950974,
      "grad_norm": 0.44091129302978516,
      "learning_rate": 8.50632911392405e-06,
      "loss": 0.1202,
      "num_input_tokens_seen": 3537024,
      "step": 1897,
      "train_runtime": 4023.8337,
      "train_tokens_per_second": 879.018
    },
    {
      "epoch": 1.917867071013394,
      "grad_norm": 0.4475288391113281,
      "learning_rate": 8.405063291139241e-06,
      "loss": 0.0818,
      "num_input_tokens_seen": 3538894,
      "step": 1898,
      "train_runtime": 4025.6565,
      "train_tokens_per_second": 879.085
    },
    {
      "epoch": 1.9188779378316907,
      "grad_norm": 0.36203065514564514,
      "learning_rate": 8.30379746835443e-06,
      "loss": 0.0476,
      "num_input_tokens_seen": 3541112,
      "step": 1899,
      "train_runtime": 4027.5907,
      "train_tokens_per_second": 879.213
    },
    {
      "epoch": 1.9198888046499873,
      "grad_norm": 0.27448907494544983,
      "learning_rate": 8.202531645569621e-06,
      "loss": 0.0956,
      "num_input_tokens_seen": 3543566,
      "step": 1900,
      "train_runtime": 4029.6121,
      "train_tokens_per_second": 879.381
    },
    {
      "epoch": 1.920899671468284,
      "grad_norm": 0.26250508427619934,
      "learning_rate": 8.10126582278481e-06,
      "loss": 0.0818,
      "num_input_tokens_seen": 3545538,
      "step": 1901,
      "train_runtime": 4031.4416,
      "train_tokens_per_second": 879.472
    },
    {
      "epoch": 1.9219105382865806,
      "grad_norm": 0.4708884656429291,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 3546782,
      "step": 1902,
      "train_runtime": 4032.9791,
      "train_tokens_per_second": 879.445
    },
    {
      "epoch": 1.9229214051048773,
      "grad_norm": 0.7044501304626465,
      "learning_rate": 7.89873417721519e-06,
      "loss": 0.0739,
      "num_input_tokens_seen": 3548922,
      "step": 1903,
      "train_runtime": 4034.8493,
      "train_tokens_per_second": 879.567
    },
    {
      "epoch": 1.9239322719231742,
      "grad_norm": 0.3273887038230896,
      "learning_rate": 7.79746835443038e-06,
      "loss": 0.2211,
      "num_input_tokens_seen": 3550554,
      "step": 1904,
      "train_runtime": 4036.5337,
      "train_tokens_per_second": 879.605
    },
    {
      "epoch": 1.9249431387414708,
      "grad_norm": 0.37105241417884827,
      "learning_rate": 7.69620253164557e-06,
      "loss": 0.075,
      "num_input_tokens_seen": 3552364,
      "step": 1905,
      "train_runtime": 4038.292,
      "train_tokens_per_second": 879.67
    },
    {
      "epoch": 1.9259540055597675,
      "grad_norm": 0.4764083921909332,
      "learning_rate": 7.5949367088607605e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 3553992,
      "step": 1906,
      "train_runtime": 4040.0242,
      "train_tokens_per_second": 879.696
    },
    {
      "epoch": 1.9269648723780644,
      "grad_norm": 0.22545349597930908,
      "learning_rate": 7.49367088607595e-06,
      "loss": 0.0792,
      "num_input_tokens_seen": 3555994,
      "step": 1907,
      "train_runtime": 4041.8702,
      "train_tokens_per_second": 879.789
    },
    {
      "epoch": 1.927975739196361,
      "grad_norm": 0.21613523364067078,
      "learning_rate": 7.39240506329114e-06,
      "loss": 0.0392,
      "num_input_tokens_seen": 3557394,
      "step": 1908,
      "train_runtime": 4043.4633,
      "train_tokens_per_second": 879.789
    },
    {
      "epoch": 1.9289866060146577,
      "grad_norm": 0.3875751495361328,
      "learning_rate": 7.29113924050633e-06,
      "loss": 0.0777,
      "num_input_tokens_seen": 3560020,
      "step": 1909,
      "train_runtime": 4045.6219,
      "train_tokens_per_second": 879.969
    },
    {
      "epoch": 1.9299974728329543,
      "grad_norm": 0.24568264186382294,
      "learning_rate": 7.18987341772152e-06,
      "loss": 0.1334,
      "num_input_tokens_seen": 3562156,
      "step": 1910,
      "train_runtime": 4047.4314,
      "train_tokens_per_second": 880.103
    },
    {
      "epoch": 1.931008339651251,
      "grad_norm": 0.2617509365081787,
      "learning_rate": 7.08860759493671e-06,
      "loss": 0.1945,
      "num_input_tokens_seen": 3563986,
      "step": 1911,
      "train_runtime": 4049.1581,
      "train_tokens_per_second": 880.18
    },
    {
      "epoch": 1.9320192064695476,
      "grad_norm": 0.23431828618049622,
      "learning_rate": 6.9873417721519e-06,
      "loss": 0.027,
      "num_input_tokens_seen": 3566308,
      "step": 1912,
      "train_runtime": 4051.1941,
      "train_tokens_per_second": 880.31
    },
    {
      "epoch": 1.9330300732878443,
      "grad_norm": 0.36068475246429443,
      "learning_rate": 6.88607594936709e-06,
      "loss": 0.0826,
      "num_input_tokens_seen": 3568304,
      "step": 1913,
      "train_runtime": 4052.9919,
      "train_tokens_per_second": 880.412
    },
    {
      "epoch": 1.934040940106141,
      "grad_norm": 0.3782718777656555,
      "learning_rate": 6.784810126582279e-06,
      "loss": 0.1143,
      "num_input_tokens_seen": 3570556,
      "step": 1914,
      "train_runtime": 4055.0012,
      "train_tokens_per_second": 880.531
    },
    {
      "epoch": 1.9350518069244376,
      "grad_norm": 0.28000250458717346,
      "learning_rate": 6.683544303797469e-06,
      "loss": 0.09,
      "num_input_tokens_seen": 3572374,
      "step": 1915,
      "train_runtime": 4056.6524,
      "train_tokens_per_second": 880.621
    },
    {
      "epoch": 1.9360626737427344,
      "grad_norm": 0.3298400342464447,
      "learning_rate": 6.582278481012659e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 3574458,
      "step": 1916,
      "train_runtime": 4058.4978,
      "train_tokens_per_second": 880.734
    },
    {
      "epoch": 1.937073540561031,
      "grad_norm": 0.23989160358905792,
      "learning_rate": 6.481012658227849e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 3575954,
      "step": 1917,
      "train_runtime": 4060.1131,
      "train_tokens_per_second": 880.752
    },
    {
      "epoch": 1.9380844073793277,
      "grad_norm": 0.3262454569339752,
      "learning_rate": 6.379746835443039e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 3578000,
      "step": 1918,
      "train_runtime": 4061.9716,
      "train_tokens_per_second": 880.853
    },
    {
      "epoch": 1.9390952741976246,
      "grad_norm": 0.28369343280792236,
      "learning_rate": 6.278481012658229e-06,
      "loss": 0.0817,
      "num_input_tokens_seen": 3580114,
      "step": 1919,
      "train_runtime": 4063.834,
      "train_tokens_per_second": 880.97
    },
    {
      "epoch": 1.9401061410159213,
      "grad_norm": 0.24143604934215546,
      "learning_rate": 6.177215189873418e-06,
      "loss": 0.0345,
      "num_input_tokens_seen": 3582430,
      "step": 1920,
      "train_runtime": 4065.7856,
      "train_tokens_per_second": 881.116
    },
    {
      "epoch": 1.941117007834218,
      "grad_norm": 0.3521853983402252,
      "learning_rate": 6.075949367088608e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 3584138,
      "step": 1921,
      "train_runtime": 4069.4925,
      "train_tokens_per_second": 880.733
    },
    {
      "epoch": 1.9421278746525146,
      "grad_norm": 0.43395107984542847,
      "learning_rate": 5.974683544303798e-06,
      "loss": 0.17,
      "num_input_tokens_seen": 3585998,
      "step": 1922,
      "train_runtime": 4071.2663,
      "train_tokens_per_second": 880.807
    },
    {
      "epoch": 1.9431387414708112,
      "grad_norm": 0.23371769487857819,
      "learning_rate": 5.873417721518987e-06,
      "loss": 0.0624,
      "num_input_tokens_seen": 3588064,
      "step": 1923,
      "train_runtime": 4073.0557,
      "train_tokens_per_second": 880.927
    },
    {
      "epoch": 1.9441496082891079,
      "grad_norm": 0.49193838238716125,
      "learning_rate": 5.772151898734177e-06,
      "loss": 0.0673,
      "num_input_tokens_seen": 3590246,
      "step": 1924,
      "train_runtime": 4074.9349,
      "train_tokens_per_second": 881.056
    },
    {
      "epoch": 1.9451604751074045,
      "grad_norm": 0.3861614465713501,
      "learning_rate": 5.670886075949367e-06,
      "loss": 0.1078,
      "num_input_tokens_seen": 3591950,
      "step": 1925,
      "train_runtime": 4076.6286,
      "train_tokens_per_second": 881.108
    },
    {
      "epoch": 1.9461713419257012,
      "grad_norm": 0.22905173897743225,
      "learning_rate": 5.569620253164557e-06,
      "loss": 0.0623,
      "num_input_tokens_seen": 3593526,
      "step": 1926,
      "train_runtime": 4078.3093,
      "train_tokens_per_second": 881.131
    },
    {
      "epoch": 1.9471822087439978,
      "grad_norm": 0.33775705099105835,
      "learning_rate": 5.468354430379747e-06,
      "loss": 0.0639,
      "num_input_tokens_seen": 3594912,
      "step": 1927,
      "train_runtime": 4079.9028,
      "train_tokens_per_second": 881.127
    },
    {
      "epoch": 1.9481930755622947,
      "grad_norm": 0.23181428015232086,
      "learning_rate": 5.367088607594937e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 3596888,
      "step": 1928,
      "train_runtime": 4081.6845,
      "train_tokens_per_second": 881.226
    },
    {
      "epoch": 1.9492039423805914,
      "grad_norm": 0.313874214887619,
      "learning_rate": 5.265822784810127e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 3598914,
      "step": 1929,
      "train_runtime": 4083.5516,
      "train_tokens_per_second": 881.32
    },
    {
      "epoch": 1.950214809198888,
      "grad_norm": 0.20736163854599,
      "learning_rate": 5.164556962025317e-06,
      "loss": 0.0261,
      "num_input_tokens_seen": 3600660,
      "step": 1930,
      "train_runtime": 4085.2556,
      "train_tokens_per_second": 881.379
    },
    {
      "epoch": 1.9512256760171849,
      "grad_norm": 0.2333381623029709,
      "learning_rate": 5.063291139240506e-06,
      "loss": 0.0581,
      "num_input_tokens_seen": 3602410,
      "step": 1931,
      "train_runtime": 4086.9643,
      "train_tokens_per_second": 881.439
    },
    {
      "epoch": 1.9522365428354815,
      "grad_norm": 0.2814578711986542,
      "learning_rate": 4.962025316455696e-06,
      "loss": 0.0584,
      "num_input_tokens_seen": 3603912,
      "step": 1932,
      "train_runtime": 4088.5787,
      "train_tokens_per_second": 881.458
    },
    {
      "epoch": 1.9532474096537782,
      "grad_norm": 0.24945557117462158,
      "learning_rate": 4.860759493670886e-06,
      "loss": 0.0943,
      "num_input_tokens_seen": 3605632,
      "step": 1933,
      "train_runtime": 4090.2269,
      "train_tokens_per_second": 881.524
    },
    {
      "epoch": 1.9542582764720748,
      "grad_norm": 0.2942098379135132,
      "learning_rate": 4.759493670886076e-06,
      "loss": 0.0861,
      "num_input_tokens_seen": 3607248,
      "step": 1934,
      "train_runtime": 4091.8579,
      "train_tokens_per_second": 881.567
    },
    {
      "epoch": 1.9552691432903715,
      "grad_norm": 0.2916257977485657,
      "learning_rate": 4.658227848101266e-06,
      "loss": 0.0192,
      "num_input_tokens_seen": 3609602,
      "step": 1935,
      "train_runtime": 4093.8655,
      "train_tokens_per_second": 881.71
    },
    {
      "epoch": 1.9562800101086681,
      "grad_norm": 0.2910366356372833,
      "learning_rate": 4.556962025316456e-06,
      "loss": 0.0753,
      "num_input_tokens_seen": 3611586,
      "step": 1936,
      "train_runtime": 4095.5503,
      "train_tokens_per_second": 881.832
    },
    {
      "epoch": 1.9572908769269648,
      "grad_norm": 0.2816304862499237,
      "learning_rate": 4.455696202531646e-06,
      "loss": 0.0829,
      "num_input_tokens_seen": 3613422,
      "step": 1937,
      "train_runtime": 4097.2667,
      "train_tokens_per_second": 881.91
    },
    {
      "epoch": 1.9583017437452614,
      "grad_norm": 0.6069521903991699,
      "learning_rate": 4.354430379746836e-06,
      "loss": 0.1285,
      "num_input_tokens_seen": 3615344,
      "step": 1938,
      "train_runtime": 4098.9904,
      "train_tokens_per_second": 882.008
    },
    {
      "epoch": 1.959312610563558,
      "grad_norm": 0.30993974208831787,
      "learning_rate": 4.253164556962025e-06,
      "loss": 0.0374,
      "num_input_tokens_seen": 3617668,
      "step": 1939,
      "train_runtime": 4100.9925,
      "train_tokens_per_second": 882.145
    },
    {
      "epoch": 1.960323477381855,
      "grad_norm": 0.3111766278743744,
      "learning_rate": 4.151898734177215e-06,
      "loss": 0.0367,
      "num_input_tokens_seen": 3619994,
      "step": 1940,
      "train_runtime": 4102.9379,
      "train_tokens_per_second": 882.293
    },
    {
      "epoch": 1.9613343442001516,
      "grad_norm": 0.25089362263679504,
      "learning_rate": 4.050632911392405e-06,
      "loss": 0.0452,
      "num_input_tokens_seen": 3621498,
      "step": 1941,
      "train_runtime": 4104.615,
      "train_tokens_per_second": 882.299
    },
    {
      "epoch": 1.9623452110184483,
      "grad_norm": 0.4375602602958679,
      "learning_rate": 3.949367088607595e-06,
      "loss": 0.0825,
      "num_input_tokens_seen": 3623018,
      "step": 1942,
      "train_runtime": 4106.2249,
      "train_tokens_per_second": 882.323
    },
    {
      "epoch": 1.9633560778367452,
      "grad_norm": 0.305571049451828,
      "learning_rate": 3.848101265822785e-06,
      "loss": 0.189,
      "num_input_tokens_seen": 3624918,
      "step": 1943,
      "train_runtime": 4107.9957,
      "train_tokens_per_second": 882.406
    },
    {
      "epoch": 1.9643669446550418,
      "grad_norm": 0.189753919839859,
      "learning_rate": 3.746835443037975e-06,
      "loss": 0.0161,
      "num_input_tokens_seen": 3626724,
      "step": 1944,
      "train_runtime": 4109.762,
      "train_tokens_per_second": 882.466
    },
    {
      "epoch": 1.9653778114733385,
      "grad_norm": 0.2611081600189209,
      "learning_rate": 3.645569620253165e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 3628918,
      "step": 1945,
      "train_runtime": 4111.6253,
      "train_tokens_per_second": 882.599
    },
    {
      "epoch": 1.966388678291635,
      "grad_norm": 0.5160294771194458,
      "learning_rate": 3.544303797468355e-06,
      "loss": 0.0625,
      "num_input_tokens_seen": 3630596,
      "step": 1946,
      "train_runtime": 4113.3186,
      "train_tokens_per_second": 882.644
    },
    {
      "epoch": 1.9673995451099318,
      "grad_norm": 0.21070647239685059,
      "learning_rate": 3.443037974683545e-06,
      "loss": 0.0314,
      "num_input_tokens_seen": 3632930,
      "step": 1947,
      "train_runtime": 4115.2625,
      "train_tokens_per_second": 882.794
    },
    {
      "epoch": 1.9684104119282284,
      "grad_norm": 0.38716843724250793,
      "learning_rate": 3.3417721518987344e-06,
      "loss": 0.0306,
      "num_input_tokens_seen": 3634688,
      "step": 1948,
      "train_runtime": 4116.9753,
      "train_tokens_per_second": 882.854
    },
    {
      "epoch": 1.969421278746525,
      "grad_norm": 0.33610138297080994,
      "learning_rate": 3.2405063291139244e-06,
      "loss": 0.0394,
      "num_input_tokens_seen": 3636356,
      "step": 1949,
      "train_runtime": 4118.6084,
      "train_tokens_per_second": 882.909
    },
    {
      "epoch": 1.9704321455648217,
      "grad_norm": 0.2751634418964386,
      "learning_rate": 3.1392405063291144e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 3638022,
      "step": 1950,
      "train_runtime": 4120.2459,
      "train_tokens_per_second": 882.962
    },
    {
      "epoch": 1.9714430123831184,
      "grad_norm": 0.3494411110877991,
      "learning_rate": 3.037974683544304e-06,
      "loss": 0.1783,
      "num_input_tokens_seen": 3640064,
      "step": 1951,
      "train_runtime": 4123.8275,
      "train_tokens_per_second": 882.691
    },
    {
      "epoch": 1.9724538792014152,
      "grad_norm": 0.278598427772522,
      "learning_rate": 2.9367088607594935e-06,
      "loss": 0.1578,
      "num_input_tokens_seen": 3642054,
      "step": 1952,
      "train_runtime": 4125.6131,
      "train_tokens_per_second": 882.791
    },
    {
      "epoch": 1.973464746019712,
      "grad_norm": 0.2557959258556366,
      "learning_rate": 2.8354430379746835e-06,
      "loss": 0.0403,
      "num_input_tokens_seen": 3644092,
      "step": 1953,
      "train_runtime": 4127.4745,
      "train_tokens_per_second": 882.887
    },
    {
      "epoch": 1.9744756128380085,
      "grad_norm": 0.5219696760177612,
      "learning_rate": 2.7341772151898735e-06,
      "loss": 0.0724,
      "num_input_tokens_seen": 3646030,
      "step": 1954,
      "train_runtime": 4129.2676,
      "train_tokens_per_second": 882.973
    },
    {
      "epoch": 1.9754864796563054,
      "grad_norm": 0.16795364022254944,
      "learning_rate": 2.6329113924050635e-06,
      "loss": 0.082,
      "num_input_tokens_seen": 3648064,
      "step": 1955,
      "train_runtime": 4131.1025,
      "train_tokens_per_second": 883.073
    },
    {
      "epoch": 1.976497346474602,
      "grad_norm": 0.3070884644985199,
      "learning_rate": 2.531645569620253e-06,
      "loss": 0.1373,
      "num_input_tokens_seen": 3649644,
      "step": 1956,
      "train_runtime": 4132.731,
      "train_tokens_per_second": 883.107
    },
    {
      "epoch": 1.9775082132928987,
      "grad_norm": 0.37285706400871277,
      "learning_rate": 2.430379746835443e-06,
      "loss": 0.1259,
      "num_input_tokens_seen": 3651276,
      "step": 1957,
      "train_runtime": 4134.3503,
      "train_tokens_per_second": 883.156
    },
    {
      "epoch": 1.9785190801111954,
      "grad_norm": 0.3180710971355438,
      "learning_rate": 2.329113924050633e-06,
      "loss": 0.0952,
      "num_input_tokens_seen": 3652972,
      "step": 1958,
      "train_runtime": 4136.0556,
      "train_tokens_per_second": 883.202
    },
    {
      "epoch": 1.979529946929492,
      "grad_norm": 0.3023328185081482,
      "learning_rate": 2.227848101265823e-06,
      "loss": 0.0269,
      "num_input_tokens_seen": 3654638,
      "step": 1959,
      "train_runtime": 4137.6924,
      "train_tokens_per_second": 883.255
    },
    {
      "epoch": 1.9805408137477887,
      "grad_norm": 0.26938310265541077,
      "learning_rate": 2.1265822784810126e-06,
      "loss": 0.0985,
      "num_input_tokens_seen": 3656578,
      "step": 1960,
      "train_runtime": 4139.4251,
      "train_tokens_per_second": 883.354
    },
    {
      "epoch": 1.9815516805660853,
      "grad_norm": 0.31211763620376587,
      "learning_rate": 2.0253164556962026e-06,
      "loss": 0.0916,
      "num_input_tokens_seen": 3658518,
      "step": 1961,
      "train_runtime": 4141.2147,
      "train_tokens_per_second": 883.441
    },
    {
      "epoch": 1.982562547384382,
      "grad_norm": 0.31595489382743835,
      "learning_rate": 1.9240506329113926e-06,
      "loss": 0.0505,
      "num_input_tokens_seen": 3660208,
      "step": 1962,
      "train_runtime": 4142.8399,
      "train_tokens_per_second": 883.502
    },
    {
      "epoch": 1.9835734142026786,
      "grad_norm": 0.2708406150341034,
      "learning_rate": 1.8227848101265824e-06,
      "loss": 0.1296,
      "num_input_tokens_seen": 3661700,
      "step": 1963,
      "train_runtime": 4144.4475,
      "train_tokens_per_second": 883.519
    },
    {
      "epoch": 1.9845842810209755,
      "grad_norm": 0.2860327363014221,
      "learning_rate": 1.7215189873417724e-06,
      "loss": 0.0329,
      "num_input_tokens_seen": 3663864,
      "step": 1964,
      "train_runtime": 4146.3123,
      "train_tokens_per_second": 883.644
    },
    {
      "epoch": 1.9855951478392722,
      "grad_norm": 0.3053610622882843,
      "learning_rate": 1.6202531645569622e-06,
      "loss": 0.0283,
      "num_input_tokens_seen": 3665944,
      "step": 1965,
      "train_runtime": 4148.2215,
      "train_tokens_per_second": 883.739
    },
    {
      "epoch": 1.986606014657569,
      "grad_norm": 0.32138562202453613,
      "learning_rate": 1.518987341772152e-06,
      "loss": 0.1864,
      "num_input_tokens_seen": 3667618,
      "step": 1966,
      "train_runtime": 4149.9149,
      "train_tokens_per_second": 883.781
    },
    {
      "epoch": 1.9876168814758657,
      "grad_norm": 0.28214770555496216,
      "learning_rate": 1.4177215189873418e-06,
      "loss": 0.0554,
      "num_input_tokens_seen": 3669234,
      "step": 1967,
      "train_runtime": 4151.5454,
      "train_tokens_per_second": 883.824
    },
    {
      "epoch": 1.9886277482941623,
      "grad_norm": 0.2837907671928406,
      "learning_rate": 1.3164556962025317e-06,
      "loss": 0.0279,
      "num_input_tokens_seen": 3671044,
      "step": 1968,
      "train_runtime": 4153.2579,
      "train_tokens_per_second": 883.895
    },
    {
      "epoch": 1.989638615112459,
      "grad_norm": 0.22075307369232178,
      "learning_rate": 1.2151898734177215e-06,
      "loss": 0.0838,
      "num_input_tokens_seen": 3672762,
      "step": 1969,
      "train_runtime": 4154.9586,
      "train_tokens_per_second": 883.947
    },
    {
      "epoch": 1.9906494819307556,
      "grad_norm": 0.46182411909103394,
      "learning_rate": 1.1139240506329115e-06,
      "loss": 0.1019,
      "num_input_tokens_seen": 3674320,
      "step": 1970,
      "train_runtime": 4156.5486,
      "train_tokens_per_second": 883.983
    },
    {
      "epoch": 1.9916603487490523,
      "grad_norm": 0.5559793710708618,
      "learning_rate": 1.0126582278481013e-06,
      "loss": 0.0564,
      "num_input_tokens_seen": 3675884,
      "step": 1971,
      "train_runtime": 4158.1782,
      "train_tokens_per_second": 884.013
    },
    {
      "epoch": 1.992671215567349,
      "grad_norm": 0.2674468457698822,
      "learning_rate": 9.113924050632912e-07,
      "loss": 0.0762,
      "num_input_tokens_seen": 3677952,
      "step": 1972,
      "train_runtime": 4159.9284,
      "train_tokens_per_second": 884.138
    },
    {
      "epoch": 1.9936820823856456,
      "grad_norm": 0.4438008964061737,
      "learning_rate": 8.101265822784811e-07,
      "loss": 0.1007,
      "num_input_tokens_seen": 3679434,
      "step": 1973,
      "train_runtime": 4161.5398,
      "train_tokens_per_second": 884.152
    },
    {
      "epoch": 1.9946929492039422,
      "grad_norm": 0.27451828122138977,
      "learning_rate": 7.088607594936709e-07,
      "loss": 0.1345,
      "num_input_tokens_seen": 3681674,
      "step": 1974,
      "train_runtime": 4163.4179,
      "train_tokens_per_second": 884.291
    },
    {
      "epoch": 1.9957038160222391,
      "grad_norm": 0.3934115469455719,
      "learning_rate": 6.075949367088608e-07,
      "loss": 0.0663,
      "num_input_tokens_seen": 3683144,
      "step": 1975,
      "train_runtime": 4165.0573,
      "train_tokens_per_second": 884.296
    },
    {
      "epoch": 1.9967146828405358,
      "grad_norm": 0.3689475357532501,
      "learning_rate": 5.063291139240507e-07,
      "loss": 0.0331,
      "num_input_tokens_seen": 3684728,
      "step": 1976,
      "train_runtime": 4166.6966,
      "train_tokens_per_second": 884.328
    },
    {
      "epoch": 1.9977255496588324,
      "grad_norm": 0.33448612689971924,
      "learning_rate": 4.0506329113924055e-07,
      "loss": 0.1883,
      "num_input_tokens_seen": 3686432,
      "step": 1977,
      "train_runtime": 4168.3366,
      "train_tokens_per_second": 884.389
    },
    {
      "epoch": 1.9987364164771293,
      "grad_norm": 0.3961541950702667,
      "learning_rate": 3.037974683544304e-07,
      "loss": 0.0497,
      "num_input_tokens_seen": 3687956,
      "step": 1978,
      "train_runtime": 4169.961,
      "train_tokens_per_second": 884.41
    },
    {
      "epoch": 1.999747283295426,
      "grad_norm": 0.24878139793872833,
      "learning_rate": 2.0253164556962027e-07,
      "loss": 0.0542,
      "num_input_tokens_seen": 3690326,
      "step": 1979,
      "train_runtime": 4171.9135,
      "train_tokens_per_second": 884.564
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8398122191429138,
      "learning_rate": 1.0126582278481014e-07,
      "loss": 0.1104,
      "num_input_tokens_seen": 3690454,
      "step": 1980,
      "train_runtime": 4172.3104,
      "train_tokens_per_second": 884.511
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.20445670187473297,
      "eval_runtime": 61.4995,
      "eval_samples_per_second": 14.293,
      "eval_steps_per_second": 7.155,
      "num_input_tokens_seen": 3690454,
      "step": 1980
    }
  ],
  "logging_steps": 1,
  "max_steps": 1980,
  "num_input_tokens_seen": 3690454,
  "num_train_epochs": 2,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.141614475047936e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}